The Peter Attia Drive - #295 ‒ Roadway death and injury： why everyone should care and what you can do to reduce risk ｜ Mark Rosekind, Ph.D.

00:00:00.000 Hey, everyone. Welcome to the Drive podcast. I'm your host, Peter Atiyah. This podcast,

00:00:16.540 my website, and my weekly newsletter all focus on the goal of translating the science of longevity

00:00:21.520 into something accessible for everyone. Our goal is to provide the best content in health and

00:00:26.740 wellness, and we've established a great team of analysts to make this happen. It is extremely

00:00:31.660 important to me to provide all of this content without relying on paid ads. To do this, our work

00:00:36.960 is made entirely possible by our members, and in return, we offer exclusive member-only content

00:00:42.700 and benefits above and beyond what is available for free. If you want to take your knowledge of

00:00:47.940 this space to the next level, it's our goal to ensure members get back much more than the price

00:00:53.200 of a subscription. If you want to learn more about the benefits of our premium membership,

00:00:58.040 head over to peteratiyahmd.com forward slash subscribe. My guest this week is Mark Rosekind.

00:01:06.500 Mark is a safety, sleep, fatigue, and policy leader with more than 30 years of experience

00:01:11.320 enacting strategic, practical, and effective database solutions that enhance safety and

00:01:16.320 health in complex environments. He was the chief safety innovation officer at Zoox,

00:01:20.940 an Amazon-owned autonomous mobile company from 2017 to 2022. He was also appointed the distinguished

00:01:27.920 policy scholar in the Department of Health Policy and Management at the Johns Hopkins Bloomberg School

00:01:32.860 of Public Health between 2020 and 2022. Previously, he was appointed by President Obama as the 15th

00:01:39.540 administrator of the National Highway Traffic Safety Administration. Before becoming the NHTSA

00:01:45.560 administrator, he was appointed by President Obama and served as the 40th member

00:01:49.720 of the National Transportation Safety Board, the NSTB, which you no doubt will recognize is the

00:01:56.120 organization that is always investigating plane crashes, train crashes, and other disasters.

00:02:02.240 Additionally, Mark previously directed the Fatigue Countermeasures Program at NASA Ames Research Center

00:02:08.140 and was the chief of aviation operations branch in the flight management and human factors division.

00:02:13.640 He earned his bachelor's degree from Stanford and his master's and doctorate from Yale University and

00:02:20.380 completed postdoctoral training at Brown University Medical School. In looking at what we internally call

00:02:27.740 the death bars, which you've likely heard me talk about, which we use to identify what are the threats

00:02:35.260 to our lifespan, you may recall that while the four horsemen generally get the lion's share of our

00:02:42.420 attention, there's always this pesky fifth cause of death, which is deaths due to accidents. And while we

00:02:50.340 typically speak about one subset of those, which are the accidents that are due to falls because they

00:02:56.500 disproportionately increase later in life, there's one cause that seems relatively consistent throughout

00:03:04.420 life. And that is accidental deaths due to transport. And so I wanted to do a deep dive into this topic

00:03:12.900 because frankly, when I consider my own mortality over the next decade, this occupies a disproportionate

00:03:20.820 share of what might account for my relative risk of death. And I know that for many of you listening,

00:03:26.520 that is also true. In this conversation, we obviously talk about Mark's background, which is quite unique

00:03:32.400 and how it led him to be an expert in this. We look at the statistics of car crashes and how that's

00:03:36.640 changed over time. We talk about the groups that are most at risk and the locations where most of

00:03:42.460 these incidents take place. We then look at various things that can increase the risk, such as obviously

00:03:48.180 being on your phone and being distracted, the role of alcohol and cannabis, sleep deprivation and

00:03:52.700 drowsiness, speed and weather. We talk about autonomous vehicles and new safety technology. And we talk about

00:03:59.720 what pedestrians need to be aware of and what resources are available for people to learn more,

00:04:04.600 especially parents. And this is something I'm thinking a lot about as my daughter is on the cusp

00:04:09.300 of beginning to drive. So without further delay, please enjoy my conversation with Mark Rosekite.

00:04:20.180 Well, Mark, thank you so much for joining me here today. This is certainly a topic that I don't think

00:04:26.060 gets enough attention given the consequences of it. It's also a topic that I think maybe people don't

00:04:33.320 necessarily appreciate the frequency of such interactions. And I guess one of the things I'm

00:04:38.140 hoping to understand today is how much of the nature of what we're going to discuss today is under our

00:04:44.480 control. I know that as I think about the things that are a threat to our lifespan, a number of these

00:04:50.560 things are kind of out of our control, but many of them are actually in our control. Certain diseases,

00:04:55.800 for example, like cardiovascular disease are almost entirely within our control and genes play a

00:05:00.420 role, but your ability to sort of go above and beyond the genetic hand you're dealt is significant.

00:05:06.780 But here, when it comes to accidental deaths, and as a subset of that, the role that fatigue plays in

00:05:12.120 that, I want to really explore this. But I think before we do, I think it's important that people really

00:05:17.000 get a sense of you and your background. When we decided we wanted to spend time on this topic,

00:05:23.520 it felt almost too good to be true that we discovered you and your work. The way things

00:05:28.980 sort of work, Mark, is basically, we sit around and brainstorm ideas that we want to cover. And

00:05:34.160 then we go out and look for an expert. That's 99% of our podcasts work in that way. I remember when

00:05:40.280 Nick came to me and said, okay, you know, this is what we found. And I was like, wow, that seems

00:05:45.120 amazing. We're going to really answer a lot of questions I've always had. So tell us a little bit

00:05:49.600 about your background. Because the first thing that stood out to me was how at a very young age,

00:05:54.420 you lost your father. Yeah, let's start there, right? Long time ago, but it still is challenging

00:06:00.040 to talk about, actually. Let me just begin by thanking you for making this topic part of your

00:06:05.960 discussion in your podcast. You already said it. This is so prevalent in everybody's life,

00:06:11.440 just being on the road. Everyone's a pedestrian at some point. We're all in vehicles moving around.

00:06:16.060 And yet, we have come to accept the carnage in ways that should just be unacceptable in our society.

00:06:22.800 So without bringing more focus, including what you can control, like, what should I be doing to be

00:06:27.900 safer versus things that I can't? Just a critical conversation. I really appreciate you taking the

00:06:33.560 time to do this. And I paused for a moment because it actually was not until my Senate confirmation

00:06:39.040 hearing to be administrator of NHTSA that I really talked publicly about this very much.

00:06:43.140 But my father was a San Francisco motorcycle policeman. He was chasing a traffic violator and

00:06:49.200 somebody ran a red, hit him, and he was killed in the line of duty. Actually, just over my shoulder,

00:06:54.700 that's a shadow box with his badge and purple heart. He was 30 years old. My brother and I,

00:06:59.760 I was three and a half. My brother was two, raised by a single parent. I'm okay talking about this to

00:07:04.840 start with because it just points out this is personal for me. And I've told people, it's like,

00:07:09.660 I don't wear that as a badge on my shoulder every time we have the conversation, but it is clearly

00:07:14.840 foundational to sort of what has put me on a lifelong path of pursuing safety and how to make

00:07:20.300 people's lives safer. And one of the things I often talk about, I'm sure we'll get to it, is

00:07:25.040 if you want to know how many lives were lost on our roads, make sure you know the exact number

00:07:28.980 because every one of those numbers is a father or mother or sister or brother or some relative or

00:07:35.920 one of your neighbors, etc. Those are real people and we cannot bring them back. And I think that

00:07:41.960 too often when we start talking about all the statistics, we walk right by the human part of

00:07:47.320 this and that gets lost. And really, that's where we should always start. Yeah, there's a quote that I'm

00:07:53.280 not remembering exactly, but it speaks to the idea that a million people is a statistic, but a small

00:07:59.380 number of people is a story. And I agree with that completely. I mean, I think we can get very numb to

00:08:04.760 what those numbers mean. I don't know if you've seen it, but there's a series on Netflix right now

00:08:09.580 that takes a lot of old footage from World War II and basically does some remarkable technology

00:08:17.840 application where it puts it back into color and makes it really remarkable. And so it's a six-part

00:08:23.540 series on World War II that is, like I think anybody who's watched it has shared my reaction to it,

00:08:29.680 which is it's stunning. But what's hard to fathom as you go through this is the loss of life. You

00:08:36.740 hear about 60 million or 80 million, I don't even remember, people lost their lives globally.

00:08:42.240 Obviously, we're aware of 6 million people being exterminated in concentration camps. And yet I

00:08:48.580 realized as I came to the end of that, I don't even know what that means because I've never seen

00:08:52.640 6 million people with my eyes. I've never seen 1 million people with my eyes. So it's very difficult

00:09:00.720 to explain those things. And what I think this series does very well is what you've done, which is

00:09:07.700 you get a few stories. You get a few stories that are very representative of the horrific nature of

00:09:15.960 what happened. And then at least you have some semblance of saying, wow, now imagine multiplying that

00:09:21.000 by a thousand or a million. And the numbers numb you. As you just said, it's like, I don't know what

00:09:26.720 that really means. So it numbs you. Very often in talks, I would actually start with what we call in

00:09:32.320 the business bent metal. And so, you know, having been at the NTSB, I'll literally take photos from

00:09:37.640 investigations minus the humans. But it's like, this was Dawn at 20 years old. She not only lost her life,

00:09:45.540 but the people in this minivan you see, four out of the five died. And the only one who didn't was a

00:09:50.620 child in a car. I mean, that makes it real for people. And what you hope is people translate

00:09:55.940 that to this could be you or a partner or your kids or your neighbor, you know, people that matter.

00:10:03.080 And there's this huge gap between those numbers you're talking about, which most people have no

00:10:07.240 concept. And the fact that when somebody actually in your circle loses their life in some kind of

00:10:12.820 crash, that makes it very personal. And again, you can't bring them back. There's no coming back from

00:10:17.820 that. Well, I definitely want to get into these statistics, both in terms of people in cars,

00:10:24.360 people as pedestrians, cyclists, et cetera. But before we do, I do want to come back a little

00:10:27.700 bit to your story, because I want people to kind of understand your training and what took you to

00:10:32.820 Yale. You did your PhD there, what you studied and how that ultimately kind of led to what you're

00:10:36.180 doing now. So tell us as much or as little as you think is necessary for listeners to kind of get a

00:10:41.120 sense of the trajectory you've taken to where you are today.

00:10:44.160 I'll give you the arc and it's more of a zigzag. Whatever you want to go in depth, let me know,

00:10:49.460 because as always, there are interesting stories along the way. I'm trained as a scientist. I was

00:10:54.840 actually a pre-med at Stanford and had the incredibly good fortune to take a course called

00:10:59.480 Sleep and Dreams when I was a sophomore. And it was taught by William Dement, MD, PhD, a professor in

00:11:06.420 the medical school. And Dr. Dement was part of the team that discovered REM sleep. And some believe he

00:11:12.160 was actually the guy who coined the term REM sleep. And as you know, even though we all as a human

00:11:17.060 race have been sleeping since calling out a primordial slime, we only knew about non-REM and

00:11:22.300 REM in the mid-50s. And so Dr. Dement came to Stanford and started really one of the first sleep

00:11:27.720 centers and was teaching this undergraduate course. And it was just fascinating. He was this passionate,

00:11:33.640 charismatic professor who was engaging. And here's this medical school professor teaching an

00:11:38.040 undergraduate course, which by the way, at the time, the two most popular courses were Sleep and

00:11:42.200 Dreams and Human Sex, another medical school professor, you know, it's kind of like for

00:11:46.100 undergraduates, what else is there? It was just fascinating. But what was really brilliant about

00:11:50.460 what Bill did is he actually offered a couple of courses. You could take a course and become a TA

00:11:55.300 for Sleep and Dreams, or you could take another course and learn where to put electrodes and how to

00:11:59.440 score sleep and actually get involved in research. So the summer of my sophomore year,

00:12:03.760 rising junior, I actually got signed up to be a research assistant, staying up all night in a

00:12:09.560 laboratory. Think about it, that as an undergraduate was an incredible experience. You'll decide if we

00:12:14.760 talk about this later, but everything got canceled that summer except one project, which was studying

00:12:18.920 the effects of the waterbed surface on Sleep. It was the 70s, and I actually met my wife during that

00:12:24.320 study. It's worked out quite nicely. But it was a way to get involved in research as an undergraduate

00:12:28.920 that you don't usually get the opportunity. So that kind of changed everything. And frankly,

00:12:34.080 when I graduated, Dr. DeMent made an offer for me to stay and run research projects. And the reason

00:12:39.340 I bring that up is because if you knew where the electrodes went, and this is a K-complex, 0.79

00:12:43.140 a sleep spindle, here's how you score sleep, you could run projects. And so I did that for a few

00:12:47.600 years, and I tell people that was like doing a postdoc before going to graduate school.

00:12:51.720 My direct supervisor was Dr. DeMent. Okay, so I did three years of doing it. And that's why when I was

00:12:57.640 ready, I actually applied to medical school, but changed because at the time, there was no sleep

00:13:02.200 specialty. There was no sleep medicine fellowship. There was no training in any of that. Dr. DeMent

00:13:07.920 was pushing to make that real. And so I decided if I wanted to spend a career looking at sleep,

00:13:13.060 it was go get a PhD. And so as you mentioned, I ended up at Yale, great academic research program,

00:13:18.820 but could also get clinical training. So if I wanted to deal with humans and projects and stuff,

00:13:23.680 I could get my training in clinical psychology, but still do research.

00:13:26.760 The interesting spin actually after that is, and I appreciate the chance to talk a little more

00:13:32.520 about this than just dates. But when I was finishing my PhD, somebody I knew from Stanford,

00:13:38.720 Mary Karskadon, who had been at Stanford, and she and Bill actually created what is now the gold 0.59

00:13:44.280 standard for objectively measuring sleepiness called the multiple sleep latency test.

00:13:48.860 Nobody ever thought Mary would leave Stanford. And she was about to start a new position at the

00:13:53.440 Brown Medical School, an assistant professor there. She had one technician ready to go. And she called

00:13:58.320 me up and said, Mark, would you like to come and do a postdoc? So there were three of us getting her

00:14:02.460 program started at Brown. It was an incredible experience. The plan was to stay, go on the faculty,

00:14:08.160 et cetera. And instead, I ended up going back and working for Bill, running a human research program

00:14:13.660 for a while. So, you know, I'm hardcore academic sleep. But part of my job was to get new projects

00:14:19.800 going. And so I got engaged with NASA Ames Group, which is one of the NASA facilities out here in

00:14:25.140 Mountain View, California. And they were doing fatigue jet lag research, but didn't have many

00:14:30.160 actual sleep people helping them do that. Some chronobiology circadian people. But it was fascinating

00:14:35.000 because they were doing a study that required recording EEG in a cockpit. And they weren't really sure how to do

00:14:40.400 that. So part of my job was to actually help them problem solve that so we could record pilots' brain

00:14:45.360 and eye movement activity in an ongoing way during flights. Fascinating. And I mentioned that because

00:14:51.040 that was a transition really out of the very specific academic kind of environment to NASA.

00:14:56.520 And so I ended up being recruited to work there and directed the program at the NASA Ames Fatigue

00:15:01.860 Countermeasures Program for seven years. And it was fantastic. And I think it emphasizes what has become

00:15:07.320 sort of for me, you know, again, not just personal, but a clear focus of my career. And that is

00:15:12.860 the application of the science into real world application. And so that's really been a force

00:15:18.640 for me throughout my career, hence the safety emphasis. Great sleep science, but how do you use that

00:15:24.220 to help people every day? Whether that's driving a car or flying a space shuttle, what do you do to make

00:15:29.220 that better? I did that for seven years. That was commercial and military pilots, astronauts,

00:15:35.040 controllers at Johnson Space Center, et cetera. Fantastic, incredible work. Love to talk about

00:15:40.460 that too, I hope. And then I started my own company, which broadened it from aerospace to basically

00:15:45.120 everybody. So when I had my own company, we worked with folks in all modes of transportation, basically

00:15:50.340 all over the world in healthcare, energy, military operations, you name it, was fantastic.

00:15:56.660 The next part that was so interesting, part of the zag, was I had the opportunity to become a board

00:16:02.460 member at the National Transportation Safety Board. So there are five members that are there.

00:16:06.440 These are the kind of positions where they call you, you don't actually submit a resume.

00:16:10.080 I had done some work at NASA helping them identify fatigue in a DC-8 crash in Guantanamo Bay,

00:16:15.140 before anybody knew that there was a naval air station there, a DC-8 crash. And the NASA group,

00:16:20.100 we helped them at the NDSP define the methodology to investigate fatigue. And they ended up identifying

00:16:25.820 fatigue as the probable cause in that particular crash. And that has become the methodology that they use.

00:16:32.460 So 10 years later, it was amazing to basically get a phone call and say,

00:16:36.700 would you like to be considered for a board member position? So I was a board member at the NDSP for

00:16:41.540 five years, launched on seven crashes, and sat through about 50 investigations that we voted on.

00:16:47.700 And I was ready to stay for a second five-year term when I got a call to become the head of NHTSA,

00:16:52.920 which is the National Highway Traffic Safety Administration. And that's the administrator

00:16:56.640 administrator of the organization within the Department of Transportation that is responsible

00:17:01.500 for all car and vehicle safety, regulation, and enforcement. And I was there for just a couple

00:17:07.660 of years because that one is tied to the president, basically. Unbelievable experience. Again, if you

00:17:12.320 think about what we just talked about, which is how do you take the science and data and actually

00:17:16.320 apply it to make things safer, better, was an incredible experience there. And when I left Washington,

00:17:21.120 I actually came back and worked as the chief safety innovation officer at an autonomous vehicle

00:17:25.000 company, which is like a way at the other end of the continuum that's been fascinating as well.

00:17:30.700 Wow. It's interesting because there's this marriage between two things that you're talking about,

00:17:36.980 right? Which is one could have stayed entirely within the world of sleep and done obviously very

00:17:42.900 interesting work. And you are fortunate to have been at Stanford, which in many ways certainly was the

00:17:48.680 epicenter of sleep research, at least in the US and potentially in the world. And then you've also

00:17:54.940 got this interest in safety and crashes, right? Obviously the NSTB. So we're going to talk about

00:18:00.920 both of these things today. I think I'd like to just start probably on the automotive safety side of

00:18:08.540 things. And I think what will come out of that is the role that fatigue plays. And then we can

00:18:13.800 certainly talk about that. I said earlier, we're going to put some statistics out there.

00:18:18.040 Can you give me a sense of what is the risk for death or injury that somehow touches the road?

00:18:27.620 So that means, again, you're a passenger in a vehicle, you're driving a vehicle, you're a pedestrian 0.99

00:18:32.120 struck by a vehicle, you're a cyclist struck by a vehicle. However you organize that, Mark, can you give

00:18:37.280 me a sense of what that looked like in 1950-ish versus 1970-ish versus 2000-ish versus today? Or give us a

00:18:46.680 sense of what that looks like?

00:18:48.740 Let's start with the final numbers from 2021, which is the last year we actually have complete data.

00:18:56.460 42,929 people lost their lives on our roadways. And you were just saying it says drivers...

00:19:03.360 Gee, gee, say that again. How big?

00:19:05.500 42,929 people in 2020. That is 118 people every single day. And good for you. It's like, 1.00

00:19:14.740 give me that number again. This is what I was telling you. It's like, you should know the exact

00:19:17.780 number because those are people, individuals. We can't bring them back. That's 118 every day.

00:19:23.000 And a lot of times people often, it's like, so how come I don't hear more about this? Or why are we

00:19:27.140 fighting? Like pandemic meant it was all hands on deck. Let's go get this. Like, how can we put up

00:19:31.880 with this? When you think about it, these happen geographically separated. So these are happening all

00:19:37.880 over the country. And very often it could be an individual in one of those vehicles. So it may affect

00:19:43.560 your family or your community. But very often most of these go unreported in the general media or

00:19:49.220 visibility for our society. 118 lives every single day. Just to put it in context, along with that,

00:19:57.280 we have about two and a half million injuries, which are everything from slight to very serious

00:20:02.780 life-changing injuries. And those are in the context of six plus million crashes every single year.

00:20:10.320 And I won't go into this in any depth, but just to put it on a global scale, it's about 1.4 million

00:20:17.100 people globally every year. And that's about 3,700 people every single day. But we're going to focus

00:20:24.900 just on the United States. And I think your question is really important because we know

00:20:30.080 over the years, things have come down very significantly with all kinds of different things I'm sure that

00:20:35.700 we're going to talk about. But it used to be like literally 100 years ago in 1923, it used to be

00:20:41.980 about 18.5 deaths per 100 million miles. Now we're down to about 1.5. So we often talk about vehicle

00:20:50.620 miles traveled or VMT. And that's important just because of the number of miles driven and the number

00:20:56.440 of people that are out there doing. There's a lot of different ways to cut this. And by the way,

00:21:01.160 I'm sure we'll go into some detail, but at NHTSA.gov, NHTSA keeps all kinds of data on this. You can look

00:21:09.080 up any year you want. And it's segmented by ages and geography states. You know, I mean, it's like the

00:21:16.140 level of detail is unbelievable what's out there. I will just comment that for a lot of people, when they

00:21:21.040 look at the data, the reason I cite 21, it's the last year that we have actual final data. So we can talk

00:21:27.560 about what we know from 2023. But those are estimates just done on a quarterly basis. And I'm sure we'll

00:21:33.760 talk about it. But we get those and they adjust up and down a bit. But it's the big numbers that really

00:21:38.260 matter.

00:21:39.400 There are two things that just jump out at me when you rattle off those stats, Mark. The first is I'm just

00:21:45.040 doing really quick math in my head, both on the global count and on the U.S. count. And this is

00:21:50.600 approximately one tenth the mortality for cancer. In the U.S., it's a little bit better than one

00:21:57.660 tenth. But globally, it's actually slightly worse than one tenth. A, that's incredible. It might also 0.95

00:22:03.400 say that in the U.S., we are probably safer on average than globally on an individual basis. So

00:22:09.500 not sure if that's true, but that would be my first thought. The second thing that just kind of jumps out

00:22:15.400 at me when you talk about this is something I remember somebody saying once that I think is

00:22:20.780 absolutely true. If you read about a death in the newspaper, it's because the manner in which it

00:22:26.820 happened was so unexpected or is somehow so horrific to us. And that's why you don't read about people

00:22:32.680 having heart attacks. I mean, let's not lose sight of the fact that that's the leading cause of death.

00:22:37.400 And that occurs 20 times more frequently than what we're talking about. And I don't know the last time

00:22:42.680 I read about somebody having a heart attack for the sake of having a heart attack and dying,

00:22:46.540 if it's somebody famous or whatever. But common things just generally don't get reported on. And

00:22:52.200 I think sadly, that's probably why we are a little bit numb to what's going on here, even though I would

00:22:58.900 argue that there's a difference, which is those of us who are driving by and see this carnage are kind

00:23:07.120 of left a bit visibly shaken by it. And I find it difficult to get information. I'll give you one

00:23:12.660 example. I live in Texas and we have big roads out here that are big and fast. And I think a lot of

00:23:19.380 them are not necessarily set up as the safest roads. There's a particular road near where we live,

00:23:25.000 where I believe there's probably a fatal accident on this road three or four times a year. And yet,

00:23:32.100 even after it happens, I'll go and do a quick search to see if I can get more information. And it's

00:23:36.940 not readily available. It's not entirely obvious what just happened. So if that's happening,

00:23:42.660 under my nose where I'm seeing the accident or, you know, seeing the aftermath of it within 25

00:23:47.660 minutes or something, I can completely appreciate why most of these deaths go unnoticed by all those

00:23:55.300 except for the people directly impacted by the relationship. That's the numbness. They're off the

00:24:01.500 radar. They're not visible, even locally to many people. So that's why I often cite NTSB statistics or

00:24:07.560 experiences because those are usually not necessarily mass casually, but there's usually

00:24:12.240 a lot of people involved. There have to be fatalities and they're very visible. As you

00:24:16.680 mentioned, very often it's because they're visible to the entire country, you know, and they get

00:24:21.480 invested. It takes a year, 18 months to investigate those, et cetera. But as you're pointing out, if it's

00:24:25.620 not in the local police blotter that it happened, there's almost never follow-up that actually say,

00:24:30.960 oh, by the way, you know, what we reported on last month, here's what actually happened there.

00:24:35.020 And I just mentioned that because I tell you from the NTSB, when you investigate these,

00:24:39.340 it's always a chain of events. It's never just one thing. And it takes a while to figure that out.

00:24:44.920 And just to be clear, a lot of that sometimes is just the local resources. Just to be clear,

00:24:50.220 you know, in the last few years, traffic enforcement resources and local police departments have gone

00:24:55.500 down. So as you were just saying, they may go and run to the scene when they've got to help people

00:25:01.120 in that emergency situation. But the ongoing investigation may be checking a box on a form.

00:25:07.220 So it stays below the radar for most people, unless you're in the circle that that person

00:25:12.560 touches your life. Yeah. And I struggle with this a lot, Mark, because I feel like there's this

00:25:17.920 enormous missed opportunity. So I look at an intersection that's no more than three miles

00:25:22.700 from our house. We haven't lived here for four years. I have already seen three fatal accidents

00:25:31.480 at that intersection in four years. And I can't find any really good information about the chain

00:25:37.320 of events and what happened. I have some sense a little bit. But to me, it's like, why aren't there

00:25:43.920 four minute videos being made that explain every one of these as a here's what not to do.

00:25:50.720 Here are the warning signs. At this time of day, when the light is this way, this is a very easy

00:25:57.640 mistake to make. And the other thing we don't really have a sense of, Mark, is the near misses.

00:26:01.480 If three accidents have resulted in fatalities, and by the way, I'm not saying three people have

00:26:06.660 died. I'm seeing three fatal accidents. It's probably six or seven people have died there in

00:26:11.140 three years. But I mean, you can tell me, Mark, but what would be your guess as to how many

00:26:16.840 accidents have occurred there or near misses occurred there that could have resulted in

00:26:22.440 fatalities? Do you have a sense of how you could even estimate that?

00:26:25.700 You could. And just two things about this. One, you're describing in what we call in human factors

00:26:31.200 that safety pyramid. Crashes are at the top. Near misses are right underneath that, but a bigger

00:26:36.840 layer. And the layer underneath that are errors. And this is why a lot of people where you see

00:26:41.080 proactive safety in aviation, they do a lot of work trying to capture those errors, knowing that

00:26:46.420 they lead to near misses. And when near misses get more visible, you see the panic in aviation right

00:26:52.480 now because they're seeing more critical near miss incidents because those are the precursors to the

00:26:57.060 crashes that occur, to your point. So you could calculate that. And I got to tell you, make a list

00:27:02.400 because out of this conversation, there's some concrete things that either you or I need to pursue.

00:27:06.720 The one you just mentioned is we got to make these more visible rather than just the family and

00:27:12.260 community feeling it. It's your point, which is even in a four minute video, you could capture what

00:27:17.900 happened and what was learned from that. And that could be enough for people to say, okay, I got to pay

00:27:22.920 attention to that because I go through that intersection every day twice. Yeah, exactly. Four times a day, I am

00:27:27.900 driving through that place that is the Bermuda triangle of death. And once a year, there's going to be a

00:27:35.880 death or series of deaths, probably five times a day, there's a near miss, probably 60 times a day,

00:27:44.180 there's an error that could have even been a near miss. And yet I don't actually know the predisposing

00:27:49.860 factors. I'm kind of extrapolating and making up from the data out there that I know loosely that

00:27:56.060 we'll talk about today. But yes, this is very troubling. So I think we've kind of hopefully made

00:28:00.100 the case for why this matters. Just a couple of other questions before we get into some of the

00:28:03.540 specifics. And I won't expect you to rattle off with the same precision. But can you give

00:28:10.040 some commentary on what these numbers would have looked like 20 years ago, 40 years ago, 60 years

00:28:16.500 ago? I mean, I have to think that with airbags and seatbelts and better cars, better car technology,

00:28:23.260 in terms of the collapsibility of cars, that things have gotten better. Like, is it easier to just start

00:28:29.400 in the 1950s? And presumably, it was just mass carnage? What's interesting is, as cars just

00:28:35.140 became more prevalent, that's when the numbers went up. And yes, in the 50s, you saw huge numbers,

00:28:40.360 which were actually part of what initiated efforts to establish the National Highway Traffic Safety

00:28:45.880 Administration, which was what had responsibility for creating federal motor vehicle safety standards,

00:28:51.400 like crash testing, all of those. So if you go again, 100 years back, we're probably over 90%

00:28:59.020 reduced to where we are today. Probably in the 50s and 60s are where you saw the peak of those

00:29:03.940 numbers, and they've been coming down. I would just say, because we're talking about this in the

00:29:08.520 statistical sense, what's interesting is we have really population level statistics we talk about,

00:29:14.020 42,929. What's interesting is we get these estimates every quarter, and then it takes a while

00:29:19.860 to finalize it. Again, we don't need to get into it, but literally NHTSA collects reports from every

00:29:25.740 police department in the country to come up with these numbers. So it takes a long time to collect

00:29:30.340 all that data, review it, finalize it, et cetera. But I say that because the estimates often prompt

00:29:36.340 people to say, oh, we're down 3% this quarter from last quarter. I think we should celebrate those lives

00:29:42.780 that have been saved, that are still with us. At the same time, that 2%, 3%, 4% up or down,

00:29:49.820 within the context of the overall population numbers, we haven't budged much.

00:29:54.040 And I'm sure we're going to talk about it, but I got to give you one example.

00:29:57.540 You know, a lot of people think, you know, we've cured drunk driving. It's been around forever.

00:30:01.840 We know about it. We've got all these things we can do. When you look at the 42,929, the top three

00:30:08.380 causes in there, impaired driving related to drunk driving, that has stayed about 30% for 20 years.

00:30:17.700 So 30% of the lives lost are due to that for about 20 years. So the absolute number has come down,

00:30:23.340 but the percentage has stayed about the same for 20 years. And then again, on that list right below

00:30:29.220 that is speeding. And now we've got number three, that's distraction. And I'm sure we'll be talking

00:30:34.240 about all of those. But again, I think to your point, it's both the large population numbers,

00:30:39.380 50s, we started seeing that peak. Now, again, it's down to about 1.5 per 100 million miles that we see.

00:30:45.560 But I do warn people is, again, when you start looking at just the quarterly to quarterly estimate

00:30:51.140 up or down by one or 2%, that will belie what is actually the larger population, things that aren't

00:30:57.140 changing dramatically enough if we're ever going to actually say get to zero deaths on our roadways.

00:31:02.760 And Mark, you mentioned that alcohol contributed to about 30% of those fatalities. What is the

00:31:08.120 contribution of speed and distraction approximately?

00:31:10.340 So speed's over 20%. I want to say somewhere between 20% and 25%. Distraction is very difficult,

00:31:18.420 right? I just actually called a friend of mine. There's this guy, Larry Blinko, who's a great

00:31:24.040 statistician at NHTSA. He's starting to think it could actually be up to 30%. And part of it,

00:31:29.800 you got to realize these are not individual numbers. So somebody drinking could be on their phone,

00:31:34.140 they could be going too fast. Somebody who's on their phone could be going, right? It's like

00:31:37.740 they all mix and match, but they're pretty big numbers. So it's at least a third, which again,

00:31:43.520 takes you into like 33%, 35% range for the alcohol impaired region. Speedings in the, again,

00:31:49.300 20% to 25% probably. And the distraction used to be considered somewhere in the 12%. But I think

00:31:55.060 Larry's new data is going to suggest it's probably much higher, maybe close to a third with some of

00:31:58.940 them. And especially if we talk about it, distraction is more than just your phone that people do in

00:32:04.320 their vehicle. So when you look at all of it, just like impairment is more than alcohol now.

00:32:08.400 Yeah, that's a great question. So to be clear, I said alcohol, but really impairment is that broader

00:32:13.500 category of which alcohol is probably still the most prevalent. In fact, at NHTSA, it's so appropriate

00:32:20.000 for this conversation, but impairment was always three Ds, drunk, drugged, distracted. And when I was

00:32:27.780 there, it's like, and I credit my son actually, it's like, where's the fourth D dad? Drowsy. Because really

00:32:35.400 any one of those would be enough to impair your driving ability. So now impairment, you do have

00:32:40.380 to think broadly that it could be any one of those. To your point, alcohol is still the number one.

00:32:45.360 Yeah. Well, there's so much I want to talk about. I'm just trying to think about how to,

00:32:47.980 how to structure us so I don't miss anything. Let's talk a little bit about, let's go deeper into this

00:32:53.700 contributing factor thing. Okay. So, um, the most obvious thing in the distraction realm is the

00:32:59.700 advent of the phone. So prior to, I don't know, early two thousands phones were not readily used.

00:33:08.440 Can we appreciate, or is there an appreciable signal in the data that suggests that the downward

00:33:15.940 trajectory of mortality has been slowed or in any way altered with the introduction of mobile phones

00:33:24.260 20 to 25 years ago? I think the way we would be able to characterize it is that we stopped seeing

00:33:29.820 the decrease that we had been seeing. And I say that because as you know, two things, one is it's

00:33:35.640 really hard to measure the distraction numbers. I mean, it's just so difficult. Same thing with

00:33:40.740 drowsy sleepiness and stuff. And the other part is just the evolution of the phones where first it was

00:33:45.680 just, you know, on a phone call, if you will, but now it's texting and looking up stuff on the,

00:33:49.760 I mean, it's just so diverse, the activities that you could be engaged in. And so again, I think that

00:33:54.520 maybe the way to think about it is more that we were making progress in a bunch of ways that

00:33:57.980 flattened or maybe got a little bit worse. And we may not be able to put all of our quote variants

00:34:02.920 appropriate that to specifically just phones. But I would also say it took to a new level,

00:34:09.180 the distractions you could get in your car. Because some of them have been around for a long

00:34:13.400 time playing with the radio. We can talk about buttons versus touchscreens, et cetera. Kids in

00:34:18.740 the backseat, those kinds of distractions have been around forever. But again, I think as those were

00:34:23.200 getting more controlled, you saw the numbers come down. Phones, I think, again, we could look at it more

00:34:28.920 as a flattening out, probably. Has there been any significant change in the past 40-ish years with

00:34:37.440 respect to the demographics of the drivers at fault in these crashes? In other words, are we seeing a

00:34:46.520 shift to younger people, to older people, anything that you can point to that there's a causal explanation?

00:34:52.620 So there are two groups that seem to be most at risk. Those are the 16 to 17-year-olds

00:34:59.100 and about the 65 to 70-plus. And the conjecture around those two, clearly we're talking about

00:35:09.720 people who are just learning to drive and who are at an age that don't have that frontal cortex

00:35:16.020 fully developed. I have a friend of mine, Greg Belenke, always talked about it. It's like there's a

00:35:20.320 hole in their head right there. So we're letting those people behind a couple tons of metal who

00:35:25.000 would have very little experience in just learning stuff. And at the other end, you've got people

00:35:28.740 getting older who actually, some of it may be aging-specific effects. The other is more just how much

00:35:34.400 they're actually driving and experience and that sort of thing. But those are the two main age groups

00:35:39.820 that we see that are mostly affected. More men than women die in these crashes. So many questions with

00:35:46.720 that. Some of it's segmented. Some of it, to your point, the causal or contrivative parts of that

00:35:51.140 would just be us with hypotheses about what's actually causing those differences and things.

00:35:56.700 What was the rationale for letting people drive at the age of 16 when, if you think about other

00:36:03.440 things that are mandated by age, you have an age at which you can join the military. You have an age

00:36:09.560 at which you can vote. You have an age at which you can drink alcohol, purchase firearms.

00:36:14.580 So there are various things that seem tethered to age, but driving is the youngest. I've always

00:36:21.620 wondered, I didn't get my driver's license until I was like almost 18 because I viewed it as a badge

00:36:28.500 of honor to ride my bike and take the bus everywhere. And I didn't want to be lazy. Like I was a weird

00:36:32.240 kid in that way. So I personally can't relate to what it's like to be a 15-year-old who's dying to get

00:36:37.140 his license. But tell me a little bit about that process and how that came to be. And for example,

00:36:41.780 given the stats you've just shared, why that age hasn't been pushed up?

00:36:46.760 I don't know you. I don't blow smoke. And so I don't mind telling you. Nobody's asked me that

00:36:51.580 question before. I'm going to go look that up now because nobody's really had a discussion around

00:36:56.100 that. The discussion is all about our education system, driver's ed. In so many other areas,

00:37:02.180 we've got recurrent training. I mean, you have that in medicine, you have that for professional

00:37:05.900 drivers, pilots, et cetera. How they actually came up with that age and came up with what are

00:37:11.420 we going to do around that age to actually prepare these people for lifelong driving experience? I'm

00:37:17.640 going to go look that one up. I don't know. That's the first time somebody's actually asked

00:37:21.040 me that. But it's fascinating because I would also say part of your point is we've actually not

00:37:26.340 gone back to question whether we need to change that or not, which, as you know, comes up all the time.

00:37:31.640 I can give you an example. I could pontificate and say, look, I mean, kids were working jobs.

00:37:36.740 They needed to be able to get there and dah, dah, dah. And it's like, okay, maybe all those

00:37:39.380 things are true. Is that true today? I don't know. The other thing that has always struck me,

00:37:44.720 one thing you should know about me, Mark, I do love driving and I love driving race cars.

00:37:49.220 So I love all things motorsport related. And I love, I love drifting, like doing all of these things.

00:37:55.460 And one of the things I've been trying to encourage my wife and daughter, my daughter's 15.

00:38:01.100 We're coming into this discussion and I've been trying to organize a course for my daughter and

00:38:07.060 some of her friends where we get a group of really good driving instructors on a 20 acre skid pad to

00:38:13.720 really teach them high end driving skills. The stuff I certainly didn't learn when I was young,

00:38:19.840 but the things that I've learned driving a race car, which is everything not to do,

00:38:25.700 like your natural inclination when this happens is going to be to do this and you will spin the car.

00:38:33.200 And if you're lucky, nothing else will happen. If you're unlucky, you'll hit something else. And

00:38:38.560 if you do this, you'll actually flip the car. And I don't believe that you can just academically

00:38:44.920 out learn that you have to just do the reps. You have to be on the track in the car doing it over and

00:38:51.480 over again. So again, I'm guessing that there has been some calculation that has said we can't

00:38:57.740 justify putting those resources into mandatory driver education, right? It's just people have

00:39:02.940 decided that we just can't request that kids learn that.

00:39:07.680 I don't think so. I think it's the first part of your question, which is, I don't think people

00:39:11.560 have questioned from the beginning. Why did we even start there? Are we preparing these kids well

00:39:17.120 enough for this life experience of driving? And how do we revisit that in our knowledgeable,

00:39:24.100 technology-driven society? How do we actually go and upgrade that to something that could actually

00:39:29.080 save their lives and the people around them better? I don't think those questions have been asked.

00:39:33.760 And I'm with you. I think the intellectual academic part of lessons, great. But unless you do the muscle

00:39:40.320 memory part of the behavioral piece of actually experiencing it, I don't think there's any question.

00:39:44.940 And I don't mind telling you, since you're familiar with F1, you know Jean-Todd. And I've gotten to know

00:39:49.500 him over the years. He's got this high-level panel. And what's been fascinating, he is now the UN's

00:39:55.040 sort of ambassador for global road safety. Oh, I didn't know that. And he has taken that on from

00:40:00.800 not only his F1 days, but his Ferrari days, etc. But he's taken this on with a passion. How do we take

00:40:07.760 what we've learned there and applied it? And why am I mentioning that? Because you already have the one idea

00:40:11.520 about the videos. This is another one you should write down. Because this is one of those investments,

00:40:16.540 right? If we get these kids early, it's a lifelong investment of the ones who actually learn how to

00:40:21.700 do this and be better at driving than the ones who sit in the course or do it online and never get

00:40:27.560 behind the wheel to know where the signal is. That's not really much education. And again, I think

00:40:32.940 it's more not that the analysis is done and we should write it off. I think it's people haven't asked

00:40:37.300 those questions or taken the time. And I mentioned John just because, you know what? Let's send him

00:40:42.340 a letter. Let's come up with what you're doing. Formulate that and send a letter and say, so

00:40:46.480 somebody at the very least ought to do this and let's study it and see if you can't come up with

00:40:50.780 a course that would make sense. Because what we just talked about is that investment now could be

00:40:56.600 huge in saving lives and costs. Yeah. Let's talk a little bit about the locations. I always tell my

00:41:03.060 patients that there are three areas where you need to really have a heightened sense of awareness

00:41:07.600 to protect yourself. And feel free to just correct anything I'm saying that is not 100% correct.

00:41:14.160 Because again, my data is a little bit old, but the three things I say is one, you have to be

00:41:18.660 hypervigilant in intersections. And that's both the standard four-way intersection, but also a T

00:41:23.680 intersection. Like when you're coming out of a mall or something and you're coming out of a gas station

00:41:28.220 or something like that. The second place is the two-way traffic without a median. Devastating,

00:41:35.560 especially as I said, we have these roads here that are 60 mile an hour roads with no median.

00:41:41.500 So personally, those roads scare me, but they're unavoidable where I live.

00:41:45.860 And then I think the third place you have to be aware is on freeways and in particular around exits

00:41:51.760 and on ramps where people are acting sometimes irrationally merging, trying to get off at the

00:41:58.200 last second, trying to get on doing silly things like that. And I basically say, look, if you can

00:42:03.440 harness the power of your attention only selectively while driving, make it those three spots.

00:42:11.720 So what would you say to that, Mark? What would you add to that? And can you comment on what fraction

00:42:17.700 of fatalities are a result of crashes under those three scenarios?

00:42:23.240 Yeah, I think you've actually nailed it. I'm not sure I can give you the exact percentages for each

00:42:28.020 one of those. It's actually in the order of you've got. And all I would do is say, you've got the

00:42:32.000 intersection part, which is any intersection, as you pointed out. The second is beyond just the median

00:42:37.260 in the road, it's any road separation. So that's true, not only of vehicles, but also pedestrians and

00:42:43.100 bicyclists. You would think about the median in the middle, but it's also on the sides.

00:42:47.900 Where unfortunately, the last 10 years, we've seen literally a 50% increase in pedestrian deaths.

00:42:53.520 Again, I'm sure we'll talk more about that. But again, that has to do with what you're describing,

00:42:58.000 not just the median in the road, but let's expand that to separation of the vehicles from,

00:43:04.020 again, the sides where pedestrians and cyclists might be as well. So I think those garner it and

00:43:09.660 you're on and off ramp of any kind of system. Again, with the issue there is the speed differential,

00:43:16.000 the challenge you have there is trying to figure out not just that I have to merge,

00:43:20.620 but the speed you need to merge with traffic going at whatever their flow is in that particular

00:43:25.260 situation. I think the other part, it's great for us to talk about the statistics. You know,

00:43:30.700 we love to hit the bell curves and like these are the highest. But I always point out, we also got to do

00:43:36.040 those edge cases because life still get lost in those. One of my mantras is never again. You know,

00:43:42.940 when those kinds of especially edge cases happen, you need to share that data. Hence,

00:43:46.900 that four minute video is so interesting, whether it's common or edge case. So it doesn't happen

00:43:51.060 again. You should need every intersection in the world or everywhere, you know, to go through that

00:43:55.300 kind of loss of life to decide we should change something. We should do something different here in

00:44:00.340 some way. And unless you make it visible, understanding the causal and contributory

00:44:04.440 factors, you can't make those changes. And just so I understand, Mark, I've seen accidents where

00:44:10.420 a driver makes a slight mistake, but it becomes catastrophic. And I'm always taken back to

00:44:16.860 1994 when my hero, Ayrton Senna died May 1st at Imola. And Max Mosley, who was the head of the

00:44:25.780 FIA at the time? I knew Max. Oh, you did? Wow. Yeah. Max said something very, very insightful

00:44:32.860 at the press conference in the days following Senna's crash, because the press understandably

00:44:39.280 were completely fixated on why he crashed. How did Senna crash? And Mosley said, you're asking the

00:44:47.520 wrong question. He crashed because he is the best driver in the world driving a car at the physical

00:44:55.300 mechanical limit of what it is capable of doing. Crashes are going to occur. The question isn't

00:45:00.780 why did he crash? The question is why did he die? And he really made sure that the sport took a turn

00:45:08.640 at that moment in, we will not tolerate drivers dying. We might not be able to stop the crashes,

00:45:15.820 but there will be no more deaths. And knock on wood, there has only been one death in F1 in the 30

00:45:22.320 years since. And what I find troubling is I'm gathering from you that that analysis isn't being

00:45:31.160 done for the 42 plus thousand people who died in 2021, where someone is saying, what contributed to

00:45:40.620 the death? And what can be learned about making that the ultimate thing that we put a buffer between?

00:45:48.160 For example, when a person makes a mistake, you always want to have more of a buffer for that

00:45:55.580 mistake to not result in the nature of force that could kill them. So for example, in a racetrack,

00:46:01.940 that's the difference between having a bigger runoff than a smaller runoff. That's the difference

00:46:07.060 between having more impact absorbing things in areas where we expect people to potentially go off.

00:46:13.260 And yet my guess is that analysis on the individual basis isn't necessarily being done?

00:46:20.000 That is correct. And I think what you do see are the level of analysis that we can cite,

00:46:26.660 the segmentation, et cetera, that again, are at sort of more population level than into the specifics

00:46:32.300 of a particular crash at that particular site with those particular individuals. Again, I'll cite the

00:46:37.600 NTSB. This is why it takes a year plus to do an investigation, because you're going to look at all of the

00:46:42.500 factors that are involved there, identify both the probable cause as well as contributing factors,

00:46:48.640 and then make recommendations so it doesn't happen again, which again, I hope we talk about this some

00:46:52.640 more, but those investigations are reactive, but they're intended to take information so you can

00:46:57.420 prevent them from reoccurring again in the future. And I think that, again, is what separates NTSB

00:47:02.560 investigations that are so thorough from, again, what we've been talking about at the local

00:47:06.700 police department level. They just don't have the resources or time and other kinds of things to go

00:47:11.620 and do those kinds of analyses. Does NTSB only investigate trains and airplanes and sort of

00:47:18.220 huge things? Like what's the mandate of the NTSB? Because it says transportation, which you would

00:47:22.360 think includes all forms of transportation, but obviously it can't do this type of analysis for

00:47:27.420 cars. Exactly right. The NTSB is required to investigate every aviation crash. So that's the

00:47:33.940 big commercial stuff, but that's also the local general aviation stuff that happens in every community

00:47:38.280 all over the place. But then it investigates specific crashes that occur that have national

00:47:45.320 importance. And so in every mode of transportation, mostly, for example, on the roadway, you see mostly

00:47:50.460 buses and trucks, because those are big. When there are fatalities, there's more involved usually,

00:47:56.360 but they can do single individual crashes as well. So when I was there at the NTSB, we did several

00:48:01.640 related to distraction with folks on their phones, for example. Now, when you go into it, you don't know

00:48:06.340 exactly what you're going to find, but you get enough information to say, we should go after and

00:48:10.000 see this because it would allow us to make recommendations that would have national

00:48:13.340 importance to do that. How did those come to you? Given the sheer volume of these crashes,

00:48:19.360 how would the NTSB make a decision or determination that this car accident is potentially one we're going

00:48:28.880 to put resources towards? I mean, do you have some inkling of what the cause is before? And you think this

00:48:33.860 is basically the illustration case? Let's pull the cover back on the NTSB a little bit to see how

00:48:40.040 this process goes. And it's very much what you're talking about. So we got to do all aviation. What

00:48:46.060 are the high opportunity investigations we could look for in marine, in roadway, in other areas? In fact,

00:48:55.200 people don't realize it, but pipeline and hazardous material also fits under there. So pipeline is a mode

00:49:00.120 of transportation that gets investigated as well. But to your point, what happens is it's like, so

00:49:05.340 what are the hot issues that we think are going on in this industry? And then you wait for calls. And

00:49:09.900 where do those call comes? They come from everywhere. A local PD might call about one that's going on.

00:49:14.420 You could get it from official. The FAA lets you know about a crash of something that's going on

00:49:18.680 somewhere. There's an op center, 24-7 op center at the NTSB. It gets all of those phone calls.

00:49:24.060 And again, to your point, you won't know ahead of time, but if somebody says, well,

00:49:27.240 we think this was involved, we think that was involved, you have to make a decision. Do we go

00:49:31.480 after that and look and then maybe find out it wasn't the case? But if it is, that's the one that

00:49:37.400 you want to be able to go in depth on so you can make recommendations to the entire industry or

00:49:41.800 individuals or whatever you need to do to try and get changes. Now, when we think of the NTSB or when

00:49:46.660 I think of the NTSB, I think of aviation and everybody thinks of the black box. You guys have this

00:49:52.800 amazing data recorder that almost without exception is recoverable and it contains not only what the

00:49:58.780 pilots said up until the moment of impact, but also the telemetry. You actually see this was the

00:50:05.260 position of this aileron. This was the thrust on this engine. This was the yaw in the stick. I mean,

00:50:10.380 you know every detail of it. What do cars have in them that allow the NTSB to do that kind of

00:50:18.500 thorough investigation? I mean, I'm a little embarrassed. I don't know the answer to this,

00:50:22.040 but I assume the car still contains some sort of a black box for telemetry. I mean, I know my

00:50:27.080 racing cars do, but I assume streetcars have some of that. Yes. And just to start, those black boxes

00:50:32.120 are actually orange. Yeah. And that's partly so you can find them, right? Yeah. Yeah. That besides all

00:50:37.100 the other beeping that's going on, it's like sometimes you can just spot it that way. Vehicles do have an

00:50:41.960 EDR, electronic data recorder, and there's actually no reason you would even know that because you don't

00:50:47.200 really have access to it. And what's interesting is most states, another conversation if you choose,

00:50:53.380 but there are some of these things that are controlled at the federal level through NHTSA,

00:50:57.260 Federal Motor Vehicle Standards and other kinds of regulatory authorities that are there. And many

00:51:02.900 of these things though, they're actually controlled at the state level. Okay. So that's kind of important

00:51:07.660 about which things are. So EDRs are where there's a basic federal requirement, but to your point,

00:51:13.900 in aviation, there are hundreds, sometimes now the newer ones, thousands of variables that you can

00:51:18.660 get from those recorders. And the EDR in most vehicles, it's the basics. And there's always this

00:51:24.580 tension between the industry about how much data they want to leave out versus those on the

00:51:29.900 investigation side that no, more is better. So some of these very recently only recorded like

00:51:35.560 30 seconds or a minute worth of data, maybe three minutes back. And then, and this is a debate even in

00:51:41.400 the aviation ones in the old days, they used to record over. For those of us who want to say,

00:51:46.160 how long have they been breaking hard like that? Or how long have they been at that speed for whatever

00:51:50.480 else? In a vehicle on the road, EDR, you may not have a long period of data to track that kind of

00:51:56.360 thing. So it's both the variables that are recorded and the amount of time that it records,

00:52:01.080 and that it's available to you post crash to be able to look at all that and reconstruct what

00:52:05.840 actually happened there. And is that one of the more important data sources you would rely on

00:52:11.180 when the NTSB did indeed come in for auto accidents?

00:52:14.980 Absolutely. That and of course, you talk to everybody. So witnesses not only in the scene,

00:52:20.100 but you talk to family and other people, you know, work, all that kind of thing. And within the fatigue

00:52:25.100 realm, when you look at that, it's literally if this is a trucker who was on the road, you're looking

00:52:29.580 at literally, where's their hotel motel key? And can you actually see that they got in the room? And

00:52:34.700 then we have to look to see if they were on their cell phone or not, where they're actually sleeping

00:52:37.980 when they, you know, it's like, you go with the electronic pieces you can, and then there's all

00:52:42.620 the other human elements of that that you use to try and piece at a minimum, the three days before

00:52:48.280 the crash. And again, you extend that based on what you're finding to go back, which is why I'll just

00:52:53.020 make one thing. I have a few soapboxes I'm sure we'll touch on. One of the hardest things is people

00:52:58.160 speculating in the first 24, 48 hours of what happened. Having launched on seven different

00:53:03.380 investigations. The first one was a Reno air show crash was air races in here. Huge. But what happens

00:53:10.480 is everybody, the media in particular, is in the first 24, 48 hours. So like, here's what caused

00:53:14.820 that. And then before you know it, it's like already established. And I used to go back to the DC and say,

00:53:20.520 okay, so how often do we actually see, you know, that first 24, 48 hours speculation end up in the

00:53:26.240 final report? People couldn't really think of many, maybe a little element of it shows up. The thorough

00:53:31.040 investigation uncovers things you would have never known if you stopped at that first 48 hours.

00:53:35.800 So my biggest thing from a safety standpoint is if you're not careful, you take action based on the

00:53:41.100 speculation. Now you've spent a year doing a bunch of changes and things that may actually not have had

00:53:45.860 any role in the crash that you were investigating. So the speculation can really bite you if you're not

00:53:50.900 careful. Hence the data sources like EDRs are critical. What can you glean about phone use in the car?

00:53:58.180 It's so common now. I sort of have this innate anger when I pass somebody and they're on their

00:54:05.240 phone. And I don't mean like they're talking and the phone is wherever it is. I mean, they're holding

00:54:10.260 their phone and you can see it as you drive past them. I sort of feel like, is this that much different

00:54:16.060 than if you were drinking a beer and I could see the bottle? What is your view on that? And how difficult

00:54:23.180 is it for, let's assume that a crash does not rise to the level of being one of the very, very few that

00:54:29.180 the NTSB looks at, but yet it still becomes a manner of like a criminal prosecution or something

00:54:34.860 like that. I mean, how much data are they able to infer if a person wasn't actually speaking on the

00:54:39.960 phone, which I assume is the easiest thing to figure out from the cell signal.

00:54:43.280 So let's actually start with, if you're going 55 miles per hour and you take five seconds to look

00:54:50.960 at your phone, your eyes off the road, you can travel the distance of a football field.

00:54:57.440 You know, then that's playing with the radio on your phone, whatever it is. Five seconds at 55 is

00:55:02.920 enough to take your football field. And I'm sure we'll talk about this if we do more impairment stuff,

00:55:07.940 but there's a very straightforward, you know, when you're driving three things you need to be

00:55:12.520 taking care of your hands on the wheel, your eyes on the road and your head in the game.

00:55:18.320 So when you talk about distraction, anything that pulls you away from one of those

00:55:21.980 is going to be a problem. And we now know that talking on the phone can degrade your performance

00:55:28.320 equivalent to 0.08 alcohol kind of performance decrements. Again, we don't have to get into all the

00:55:34.340 data, but there's really good data that even hands-free can do that. Because as you know,

00:55:39.520 there's no such thing as multitasking, right? It's switching, et cetera. And so people are like,

00:55:43.480 it's legal. You're like, hands-free, it's okay. It's like, no, it's not. Because if you're engaged

00:55:47.220 in that conversation, then your head's not in this game, whatever. And so again, I think that's one of

00:55:52.760 the challenges you have when you're looking at all this is how diversity are. So to your point, 1.00

00:55:56.980 the NTSB will go and get all the kinds of data. Locally, they're probably never going to get that.

00:56:01.880 So what would you look for? And I often think about this when I do drive by those people,

00:56:06.200 usually because I'm going the speed limit here and they're going less than that in a faster lane

00:56:10.840 because they're sitting there staring at something, right? Is I'm always sitting there thinking,

00:56:14.540 I know enough to basically ask the police department, get those phone records. I want

00:56:18.740 to know if a signal was bouncing off a tower somewhere locally, was that going on? And of course,

00:56:24.020 nowadays we have more cameras everywhere. And so can I show that they actually were on that phone

00:56:29.120 at a certain place, whether it's their video or video from a vehicle you might be in, etc.

00:56:34.080 There are more sources for that kind of information. But to your point, if you didn't

00:56:38.120 know about an EDR or video or that, again, an investigation of the NTSB, you can get a subpoena

00:56:43.840 if the phone company won't give it to you and literally get the records to know if they were

00:56:47.500 on the phone during a certain time. That stuff's available. In most crashes, insurance companies

00:56:52.020 even aren't going to necessarily pursue all of that. But those things would be available depending

00:56:56.320 on what happened for you to, again, like any other investigation at NTSB level, you could go after

00:57:01.300 that stuff to determine what was really there. Just a couple of weeks ago, we're on El Camino here

00:57:07.740 in California. So it's three lanes, 35 miles per hour speed limit. And we're stopped. We had seen a

00:57:14.920 motorcycle officer who had stopped somebody a while back. All of a sudden now he's coming up on the

00:57:19.400 left. We're like in the middle. So there's another vehicle on the left. But the guy's coming up

00:57:23.020 splitting the lane because we're stopped. And he's literally looking in the cars because I can see

00:57:28.120 his helmet turning and looking in the cars. And the guy next to us was on his phone. I mean, I was

00:57:32.080 sitting there just watching him, you know, and the cop came up and looked in there and like start

00:57:36.140 shaking his head. And the guy put it down. And I'm thinking like, that's the difference when you

00:57:40.940 actually have someone looking at you, whether it's video or something else that says you shouldn't

00:57:45.280 be doing that. And you and I know that as soon as the cop went by, he's back on the phone again,

00:57:50.000 right? So Mark, have there been any technological solutions proposed to dramatically lessen the

00:57:58.680 burden of phone use while driving? And again, to be clear, I say this as someone who would

00:58:04.840 be inconvenienced by it myself. I'm constantly using long drives when I take them as a chance

00:58:11.680 to get caught up on phone calls and listen to podcasts and audio books. And even though my hands

00:58:17.840 are on the wheel and my eyes are on the road, there's no doubt, especially in a phone call,

00:58:23.740 it takes my head out of the game a little bit. So again, phone companies and consumers alike would

00:58:32.860 be, I'm sure, opposed to this. But is there no technology solution proposed to make it much

00:58:40.880 more difficult to be distracted while driving?

00:58:43.240 There are. It's interesting because companies did oppose it quite a bit, but there are now,

00:58:49.600 and you should look on your own phone, but there is usually now a button that says don't call. And

00:58:55.740 it has a message. It'll send a text or a voicemail basically and says, I'm driving now, call you later

00:59:01.820 or leave a message or whatever. So there are technological things.

00:59:06.080 Yes, but it still requires the user to take the action. And again, the question is like,

00:59:12.640 how can you make this so that you don't have to opt into it?

00:59:16.620 Right. And those could exist as well. They already do. Phones can now tell when your vehicle's moving.

00:59:22.880 So accelerometers and other sorts of things do that. And I think what you're getting to,

00:59:27.340 which is interesting, is that's a whole nother choice issue about you're moving above a certain

00:59:33.280 speed. So you're out of a parking lot or whatever else. We jam it or we just don't let it happen.

00:59:39.000 That would be an interesting battle, I bet, for some. And that's why it's so interesting about your

00:59:44.040 hypothesis there, because right now you could just make the decision. I'm going to shut it off

00:59:49.700 or I'm going to put it on airplane mode and I'm not going to get anything while I'm here.

00:59:53.560 So when I was at the NTSB, Debbie Herzman was the chairman and we were investigating a couple,

00:59:59.720 as I mentioned, crashes where clearly cell phones were an issue. And we made a recommendation that

01:00:04.740 cell phones should not be used in cars except for emergency situations. And I remember exactly

01:00:10.860 around that time because I used to call Debbie and we would be talking about stuff and she had a long

01:00:17.020 commute and she used to use that to catch up with all kinds of stuff. And then all of a sudden we were

01:00:22.120 doing these investigations and I would keep getting her voicemail. And then as we got closer to our

01:00:27.700 recommendation, et cetera, I said, I know what's going on now. She goes, yeah, I wanted to know

01:00:31.900 what it would be like to shut the phone off or put it in a bag in the back. And you know what?

01:00:37.120 It's really inconvenient, but I'm a better driver when I'm doing this than when I'm doing it

01:00:41.640 the convenient way. So to your point, there's the technological, that's sort of a societal,

01:00:47.600 back to our question, like, do we make that vote and say, no, we're not going to let you do that

01:00:51.120 if you're moving. But you always have the personal choice about if it's important enough to you,

01:00:55.700 you can decide. And the personal choice in some cases could also mean you can differentiate.

01:00:59.860 It's like, I'm on a strange road and I'm a little faster. I got more people in the car. I'm not

01:01:03.220 going to do that as opposed to open roads, speed limit. There's nobody around. Music's not even an

01:01:08.780 issue. Again, you could make more choices when you did that. Yeah. Or do I have the ability to override

01:01:14.180 it? And I mean, this is how I perhaps erroneously justified in my own mind, which is I'm now steeped in

01:01:21.480 the practice of identifying hotspots. And whatever it is I'm doing, if I'm listening to something or

01:01:28.620 if I'm on the phone, I at least convince myself, again, this might not be true, but I convinced

01:01:33.600 myself, okay, you're just going to pay more attention right now because this is the intersection

01:01:37.760 and you're looking both ways, even when you have the right of way or that kind of thing.

01:01:41.480 But yeah, the other example of this would be, again, it would be hideous to look at,

01:01:46.500 but you could easily install a breathalyzer in every vehicle that allows it to not start

01:01:52.780 without a blood alcohol below. You might even make it a more egregious setting than 0.08 and

01:01:58.020 make it 0.05 or something like that. And nobody would want those in their cars because they're

01:02:03.520 so unsightly. I think they do actually have devices like that for people who have been convicted

01:02:07.960 drinking and driving. But when you just look at the numbers of deaths due to intoxication and

01:02:14.560 distraction, it begs the question, right? Like how much would we be willing to be inconvenienced

01:02:20.320 to save, call it 50 lives a day? Because that's about what you would do. You'd save about 50 lives

01:02:26.640 a day if you took those two things off the table. If people didn't have phones in their cars

01:02:33.100 and they couldn't have alcohol in their system when they drove, that's a pretty good estimate,

01:02:38.060 right? About 50 people per day would be alive.

01:02:39.980 A third of 118, so maybe 40. So let's go back a moment. Realize you're different when you were

01:02:47.280 talking about being on the phone in certain places where you're going to pay more attention.

01:02:51.000 You're different in a good way because you're actually cognizant. You're thinking about those

01:02:55.620 things and what risk they would create for you. And again, in your framework, it's like,

01:03:00.700 what can I control? What's out of my control? Well, that's something you have control.

01:03:04.100 You're different than most people because you're even thinking those things.

01:03:07.160 So that actually does make you safer in those situations than someone who isn't even thinking

01:03:11.960 there's an issue with me talking on the phone and playing with my radio. And by the way, I'll do

01:03:16.900 this. It's like that puts you in a different place than those folks. But going to your other,

01:03:21.960 it's not hypothetical anymore. Congress just passed a law that new cars will have to have a technology

01:03:28.600 that can detect whether you're at the 0.08 level or higher. And what's interesting is that

01:03:34.400 technology has actually been in the works. So take it as a model for what you're talking about

01:03:38.060 with phones. But that technology has been in the works for a long time.

01:03:44.040 And it's another, it's called DADS. It's one of them. And it has a free breath sort of analyzer

01:03:48.980 and also stuff that's in the steering wheel that literally is looking at molecules in the air.

01:03:54.700 And it's what you said is it won't start if it hits. It's for everybody.

01:03:58.680 Now, how will it differentiate between someone in the car who's incredibly inebriated and

01:04:06.840 presumably putting particulate matter of ethanol or whatever metabolite is being searched? But the

01:04:12.260 driver, let's assume, is not. The designated driver. I mean, that's why you do that. Right now,

01:04:17.300 it's specific enough with the, and that's why I think the latest has both sensors, the one that's

01:04:21.840 in the wheel as well as the other more general one. And they've done enough research to be able to

01:04:25.960 differentiate those. The last thing I'll just say is what's interesting, that's one form of it.

01:04:30.700 The legislation that just passed is that basically car companies are going to have to provide a

01:04:35.000 technology without, again, it's more of a performance target than it is the mechanism.

01:04:40.040 And so other people could come up with different ways to be able to do that.

01:04:43.360 But I mentioned this because I think you're onto something here, which is it took maybe decades

01:04:49.260 to get this legislation from Congress to say we should have this in every car because of the lives

01:04:55.160 it could save. And part of that was the ongoing debate, sometimes about technical things, more

01:05:00.840 often about what we've started this conversation with, the societal value if we think it's more

01:05:05.700 important to inconvenience you if you're drunk than it is to allow you to kill someone else on the

01:05:10.520 roadway. When did drunk driving fall out of favor? Like, I get the sense that there was a day in

01:05:18.640 which driving drunk was a normal thing to do. Like, nobody cared that you had a drink and got in your

01:05:23.440 car. Clearly, that's not the case today, despite the fact that the numbers are still as high as

01:05:28.100 they are. Was there a particular moment in time in which an accident changed all of that? Or was it

01:05:34.200 just a general sense of awareness?

01:05:37.760 I will apologize to them, but it really was about MAD, Mothers Against Drunk Drivers. And the woman who 1.00

01:05:44.900 started it lost a child in a drunk driving crash.

01:05:48.580 And this is what, in the 80s?

01:05:49.840 Yeah, and that's why I'm trying, I'm kind of appalling, because I think it was late 70s,

01:05:54.180 early 80s. But they really are the first ones to have a, quote, victim group that said this should

01:06:00.860 not be allowable anymore. And so I would say it wasn't so much an individual crash.

01:06:05.700 But they put faces to a set of crashes and a cause.

01:06:10.000 And took it to the statehouse, to Congress. It's like, we cannot let this happen anymore.

01:06:16.200 And people don't know this, but MAD actually provides counseling services and all kinds of

01:06:21.040 other things that they do. But they really are the model, again, a victim group that said,

01:06:26.540 we need this issue in front of our society. It's not okay. And so they're the ones to get 0.08 and

01:06:32.480 all. Again, see, that's one of the things. Every state has to vote for what the blood alcohol level

01:06:36.960 has to be. And so everyone's different. But they're the ones who pushed, it used to be higher,

01:06:42.680 and then finally got to 0.08. They're the ones who basically are the ones who pushed especially

01:06:48.060 hard. Others did as well, but they were right out there in the front for this new legislation

01:06:52.320 regarding the technology. And they're the ones who will make sure that the pressure stays on

01:06:56.780 to see the technology actually integrated within vehicles coming. So I think that's an example of

01:07:02.420 what you're talking about. It could be a crash, high visibility, maybe a celebrity. What do we do about

01:07:07.020 this? Or it can be this new model that was created with MADD, which we see in other arenas as well.

01:07:13.320 Yeah, I always find that the individual cases do more. You're probably familiar with the case of

01:07:17.840 Libby Zion, just for folks listening. So Libby Zion was a woman admitted to a hospital in New York.

01:07:23.880 I can't even remember if it was Presbyterian.

01:07:25.760 I think it was Mount Zion.

01:07:27.100 It was a Mount Sinai. Yeah, you're right. It was Mount Sinai. She was admitted to the ER,

01:07:31.700 a resident who technically probably wasn't even qualified to make a decision that was being made about her

01:07:36.640 care, made a decision to give her, I don't even remember the details, gave her a medication while

01:07:41.500 she was already on an MAOI, and it resulted in her death. And even though I don't really think this

01:07:47.700 was just about fatigue, I think it was more about resident supervision, it became the linchpin case

01:07:53.200 that her father basically took against the medical community about resident work hours. And it might have

01:08:01.040 taken 15 years from her death until the changes that were imposed on resident work hours, which was

01:08:08.220 kind of at the end of my training period. But again, it all came back to this woman's death, wasn't about

01:08:14.820 the million other insane stories that resulted from medical residents being exhausted. So I guess I bring

01:08:23.060 all of this up to say, Mark, have we had that moment yet with distracted driving? Do we have the

01:08:28.820 equivalent of Mothers Against Drunk Driving? Or do we have the equivalent of Libby Zion's death? 0.54

01:08:34.620 Because I have my story. So my story is, I used to work with a guy named Nick Venuto. Nick and I sat

01:08:42.660 opposite each other in our offices. And Nick was a really, really amazing guy. We always say this,

01:08:50.160 but truly an amazing guy, a remarkable cyclist. But he was so committed to his family that he used his

01:08:58.480 commute, his long commute on his bike to do the bulk of his training so that he didn't have to do

01:09:04.780 any cycling during the week and he could just do his main training on the weekends. He had just won

01:09:10.680 one of the most remarkable time trials in Austin, which was the Mount Palomar climb, which is an epic

01:09:16.400 climb. One of the hardest climbs in the United States. Anyway, it was a Tuesday afternoon in May of,

01:09:23.340 I want to say 2010, 2011. We were both just leaving the office together. And I was going to drive home

01:09:31.120 to get on my bike to then ride to where I did my intervals that day. And he was riding home.

01:09:38.040 And I remember thinking there were two places I could go and ride that day. One was up Torrey Pines

01:09:43.640 and the other was on the bike path of 56, which was a bike path that ran along a freeway.

01:09:48.840 That was the road that he always took home. And I was like, maybe I go ride that way and I'll ride

01:09:53.920 out one way with him. I ultimately ended up deciding to go up Torrey Pines that day.

01:09:59.540 That day when he was driving home, a woman in an SUV in the right-hand lane, which is the lane 0.99

01:10:05.640 adjacent to the bike path, where there's a hill that you have to go up. It's about a 12-foot hill

01:10:11.340 at about 40 degrees with a fence, was on her phone. And I think what happened is the car in front of her

01:10:19.640 stopped and she drove up the hill through the fence, hit and killed Nick on a path that I rode on most days. 0.89

01:10:29.160 And I mean, I spent the next two years so goddamn angry that I wanted to start carrying a gun when 0.98

01:10:40.840 I rode my bike to shoot any motorist that got anywhere near me. I mean, that's how pissed I was. 0.98

01:10:49.840 And I thought to myself, how many times does this happen? Like this is one guy that I just happened

01:10:55.840 to know who died, but this is happening constantly. And why are people not more irate? Why is something

01:11:04.520 not being done about this? Forgive my long story, but is there a movement around these? It's not just

01:11:11.680 the cyclists and pedestrians, it's other drivers who are victims of the distracted driver. And by the way,

01:11:17.140 I want to say one more thing. I am a distracted driver sometimes. I could easily have been that woman.

01:11:22.980 There are times when cars in front of me are stopping and I haven't caught it until the last

01:11:28.060 second. And just because it hasn't resulted in an accident, I don't know that it makes me morally

01:11:32.160 any better than that woman. Well, it's personal for you too then, just as we started. Very much so.

01:11:39.240 Again, I paused at the beginning because I don't often talk about my father, but it's personal for a

01:11:44.020 lot of people. But I think what's interesting is the stories actually aren't told enough. That's what

01:11:50.320 we're talking about MAD. They told the stories. This should stop. We have not reached the societal

01:11:56.120 unacceptability, the societal outrage that you just portrayed at the level of we just got to stop it.

01:12:03.860 And that's why I say MAD helped bring the visibility, but the legislation for technology

01:12:07.820 in the vehicle just got passed. It's not even in the cars yet, right? I mean, it's still got to figure out

01:12:13.260 how to do that part. So you asked it earlier. We could do the same thing with distraction.

01:12:17.620 And I think good and bad. The good part is people are asking questions like what you just did.

01:12:23.920 How do we take this to the next step? We have models, alcohol impaired driving. We have ways to go 0.59

01:12:29.380 after this and even do better with technology and things, etc. How do we do that in the realm of

01:12:33.960 distraction? And you've also identified why it's so hard. The bad part is even us people who are trying

01:12:41.100 to be really good, trying to save lives and do the right thing. It's hard. There's still some things

01:12:46.620 where we make choices, even knowing choices like that's not really the best thing to do here.

01:12:51.840 And that's a challenge, which is even really good folks who know what's going on. I have friends of

01:12:56.440 mine in the police department. It's kind of like, I want to make sure I see every one of those police

01:12:59.860 cars stop at the stop sign. No rolling because you're the model for others. It's okay. Watching

01:13:05.680 that motorcycle get nodded head of the guy who was on his phone, etc. It's like we need all the

01:13:11.600 available mechanism strategies we have to change many of these things. And what you're getting to

01:13:16.740 is there are so many of them. We can look at the top three, but that doesn't necessarily get rid of

01:13:22.540 all of them. But I think to your point, and a lot of what you focus on these conversations is,

01:13:27.320 so what are the things that are under your control, though? So for you, if that means watching a little

01:13:31.740 bit more about when you're on those calls, and how you're paying attention to them, hands on or

01:13:36.100 hands off, etc. And that's why I'm part, you're already different because you're already thinking about

01:13:40.340 where those danger zones are. Someone who doesn't even think about that is at a higher risk, okay?

01:13:45.480 And as we talked about it, this is risk management. What's under your control, what's not? It's risk

01:13:50.340 management. And what you choose to actually try and control and take care of others that will still

01:13:55.060 put you at risk in certain situations. Yeah, one of the things I've often wondered, Mark, is there any

01:14:00.180 research into the lives of those who kill other people on the road, and how their lives are forever

01:14:07.900 impacted? Because I have to be honest with you, I think until right now, as I'm telling the story,

01:14:13.040 I've never once given a thought to that woman who killed Nick. Truthfully, I've hated her guts. 0.98

01:14:18.920 But I don't know that that's fair. And I don't know anything about her life today. I don't know how

01:14:23.700 often she thinks about Nick. I don't know what she lives with. But my impression is that there are two

01:14:31.680 deaths at every death, and that a part of the person who killed this person is at risk. And this

01:14:38.220 is a very different type of homicide. This is not first-degree murder. This is involuntary

01:14:42.800 vehicular manslaughter. Nobody wakes up thinking they want to kill someone today in their car.

01:14:48.560 I know your training is in psychology, so I just wonder if there's any aspect of your training that

01:14:54.140 gives you a sense into what those people go through. And by the way, is that something we should be

01:14:57.860 tapping into to help create more of a zeitgeist around this? It's not just the lives of those

01:15:02.500 who die. It's those who live. 25% to 33% of people in a crash are going to have PTSD within 30 days or 0.96

01:15:09.880 longer. There is a literature on that. And I would just say, going from the statistics to the individual,

01:15:15.820 Matt is quite good at actually not only having victims' families talk about what's going on,

01:15:21.880 but they have drivers who have killed someone that are also part of who speaks to folks about what

01:15:28.460 this does to change their life. And sometimes that's jail time. Sometimes even if they don't

01:15:33.300 go to jail, it changes their life forever because they are waking up every morning. Survivor's guilt,

01:15:38.560 anybody who really cares about humanity is going to carry that with them for the rest of their lives.

01:15:43.760 And so, yeah, we see numbers. Actually, again, close to a third, basically, people are going to have

01:15:48.240 PTSD, some aspect of that. And then telling those stories can be very powerful.

01:15:54.180 Before we go into some other things I want to chat about, I want to round out some of the other

01:15:57.480 contributing factors. Where does weather rank on this list? Whether it be rain, snow, ice,

01:16:04.240 sudden darkness, or cloud, how much does that contribute?

01:16:07.340 So I'm going to just drop a small bomb here for a minute about something. And then to answer the

01:16:12.460 question, then we'll come back to it whenever you're ready. But NHTSA did a study, came out in 2015,

01:16:17.940 2018, that showed in the chain of events, the last event before a crash, that 94% of the time,

01:16:25.220 it's a human choice or error. Okay, and that's my language, human choice or error. So then the

01:16:30.980 question is, what are the other 6%? Well, 2% are vehicle defects, 2% are the environment, 2% other.

01:16:38.820 And so to your question, the environment is both the physical environment, potholes,

01:16:45.700 road conditions, etc. And then things like weather.

01:16:49.380 It's virtually nothing.

01:16:50.320 Not a lot. That's right. So that does happen, but it's going to be small compared to the other.

01:16:55.140 And again, we can talk a little more about the 94% because there's been some controversy about

01:16:59.020 that number. But that's a soapbox for me. You got to be ready to get into that one.

01:17:03.660 So let's say more about that then. So where's the controversy around that? Are there people who

01:17:07.880 are arguing that that number is too high, that that's overstating it?

01:17:11.440 What's interesting is nobody's actually addressed the number. And so you'll tell me where you want

01:17:16.980 me to, okay, that's enough of that, Mark. Let's go on to something else. But let me just start.

01:17:21.100 This is why a crash is not an accident. A crash is preventable. An accident implies it's inevitable.

01:17:29.020 And most people don't know this, but the word accident started getting used mostly in

01:17:33.840 occupational environments like the work setting in the 20s and 30s when companies didn't want any

01:17:39.440 responsibility of on-the-job injuries and deaths. Oh, that was an accident. I always say it's like

01:17:44.400 it was an act of God. It was unintentional. Nobody could do whatever. And part of the reason I bring

01:17:49.500 that up is because if you believe in safety, that by investigating, you can determine the causal and

01:17:56.340 contributory factors and then intervene in some way by changing, so you prevent it from

01:18:01.340 reoccurring, you prevent whatever, then that's what safety is about. It's a crash. And if you

01:18:06.940 didn't believe in that and you really think it's an accident that's inevitable, why are you trying to

01:18:11.280 make changes, investigating, et cetera? No. And that's the mission of the NTSB, investigate those

01:18:16.220 crashes to make recommendations so they don't reoccur. Okay? So what's interesting is there's two

01:18:22.000 elements to this. One is, as a NASA scientist, pretty much you could start any paper with 70,

01:18:27.680 because it was so well-established, 70 to 80% of incidents, you know, that pyramid we were talking

01:18:33.660 about, 70, 80% were human factors related. It was just a given after all this research and things

01:18:40.320 that were going on. You know, the Institute of Medicine report, right, to errors human, 100,000

01:18:45.560 medical errors every year due to human, et cetera. And so it always surprised me. It's like 94%

01:18:52.080 shouldn't actually surprise anybody. But there is something which I call safety misdirection.

01:18:59.040 And that's where two things are going on. One, denial. Let's not deny what the causal or

01:19:04.680 contributory factor was. And the other is blame. And so the old version of safety misdirection is,

01:19:11.160 oh, the car companies are just blaming the drivers for stuff. Okay? And they want to sort of avoid,

01:19:17.000 it's kind of like the corporate accident language. Let's get away from that. And so I always point out

01:19:23.040 that what I learned at the NTSB is you can have safety or you can have blame. If you want to blame

01:19:27.760 people or da-da-da-da, that's not going to get you to the safety. The safety part is identifying the

01:19:33.080 causal and contributory factors and then intervene so those things don't happen again. Which, by the way,

01:19:38.760 even if it's the human, that very often, as we've already discussed, there's a different

01:19:43.180 intervention like technology, which our society happens to turn to a lot. When it's something

01:19:47.920 about our human behavior that's hard to change, well, what's the technology we might be able to

01:19:51.980 use that either supports, helps, or just eliminates the ability to do that? So that's one of the things,

01:19:57.680 the controversial part that's been interesting is people saying, well, if you talk about 94%,

01:20:01.240 you're just trying to blame the drivers, et cetera. Now, there are some car companies and some people

01:20:06.340 in the autonomous vehicle space that love to do that. So it's, again, how they use the data,

01:20:12.040 but scientifically, a huge part of this is human choice or error. Then we still have the challenge

01:20:18.540 of then how do we address those to make sure those things don't reoccur again in another place?

01:20:23.460 So, again, it's a little bit of a soapbox for me when I hear about people like, oh, that 94% has

01:20:28.800 been, it's like, no, it's not. Nobody's actually arguing about the number. I think it's more the blame

01:20:33.780 and shame part, which again, doesn't get you safety no matter what. So you can't deny what

01:20:38.640 the sources are and blaming people. In fact, nowadays it's like, oh, you're just blaming

01:20:42.380 the driver. Then they point and they start blaming the car. They blame the software.

01:20:46.360 That's still not going to get things fixed for you. So you're going to save lives.

01:20:50.080 So there you go. I'm with the 94% and think that that other 6% shows that last event in the chain

01:20:56.460 is very small percent. It's going to be weather or potentially defects.

01:20:59.480 I don't know if you will have this statistic, but while we're on the topic of cause and fault,

01:21:07.740 we talk about nearly 50,000 people a year died last year. Do we know how many of those people

01:21:16.480 were the people who were the cause of the mistake that led to the crash versus people who were

01:21:23.880 presumably not doing anything that would have led to a mistake?

01:21:29.280 We don't have that level of detail of those specific crashes that occurred to be able to tell

01:21:36.000 that. The reason I'm asking, by the way, is to sort of make the point that there are two elements to

01:21:41.500 this. There is what can I do as a driver to make sure I'm not the one who makes the mistake?

01:21:48.560 And then what can I do as a driver to make sure that when the other driver makes the mistake,

01:21:54.680 I'm in a position to react better or see their mistake before it's happening.

01:22:00.220 And I think about this a lot as a dad who is starting to talk a lot about this to his daughter.

01:22:08.360 If I'm driving her to school, just ask her. It's a nonstop lecture. It's, hey, Olivia,

01:22:14.680 did you see how I did that? Did you notice why I did that? Why did I look there? Did you see what

01:22:19.900 that driver just did? What I'm assuming is that we have to be vigilant on both fronts. It's not

01:22:27.300 enough to just say, I'm going to drive at a responsible speed and I'm not going to hold my

01:22:33.800 phone and I'm not going to drink. That's a great first step. I don't think it's the complete step.

01:22:39.320 You have to assume this is the mantra I use, Mark. And my wife thinks this is a little grotesque,

01:22:46.460 but I don't care. I say it to her all the time. I say, I want you to imagine that somebody woke up

01:22:52.680 today with the stated purpose of killing you. So they've been handed an envelope. The envelope 0.65

01:22:59.640 contains your name and they're told to kill you today. But here's the catch. They're not allowed to 1.00

01:23:05.520 use a gun. They're not allowed to poison you. They have to do it with their car. So I said,

01:23:11.960 armed with that knowledge, how would you drive different today? It's only one car. You're going

01:23:17.380 to see thousands of cars today. It's only one, but there is someone out here who's trying to kill you. 0.82

01:23:24.080 They're being paid to kill you. What will that knowledge do to your attention? And how will you

01:23:31.840 treat each intersection, each on-ramp, each off-ramp, each T-joint, all of those things that

01:23:38.040 we talk about? I don't know that that's a great way to go living your life. It's a bit morose,

01:23:42.380 but it's the only heuristic I've come up with to help with the other half of the equation.

01:23:48.080 That's why I asked the question, right? I don't know how to quantify the effect

01:23:51.740 size of each of these things. Is it worth the baggage and the overhead, the emotional overhead

01:23:58.080 of playing that game? Could be your life. So yeah, it's worth it. And I said this at the

01:24:05.020 beginning, which is thank you for bringing this topic up. Because at the core, like you say,

01:24:09.820 your mantra of what's under your control, what's out of your control, that's what we're talking about

01:24:15.260 here. But even the things that are out of your control, you can be more vigilant to some of the

01:24:19.900 risk factors or other things that you could do. At the higher conceptual level, I would say,

01:24:24.720 and maybe just a slight sort of reworking exactly what you said. But part of this is

01:24:30.280 situational awareness, SA. I mean, you know this in racing. It's a big thing in aviation.

01:24:36.140 And I think it's what you just literally articulated with your daughter. Did you see how I looked at that?

01:24:41.740 Did you see that car over there doing whatever? So I think that is huge. And the second part is what

01:24:47.740 we would typically frame as defensive driving. It's a little different than what you portrayed,

01:24:53.320 but it is the same thing, actually, which is there are people that are out here that are not making

01:24:57.580 the good choices you are, but who are on their phone or drunk or haven't had enough sleep. You

01:25:02.980 don't have any control over that. But with your situational awareness and driving defensively,

01:25:07.340 there are things you can do. So besides scanning here, looking there, you can come back again.

01:25:12.160 You can look at those intersections where you know things are higher so that if that raises your

01:25:16.020 vigilance, that could be enough for you just to pause at that stop sign longer.

01:25:20.120 And it becomes not even an error, let alone a near miss or worse. And so I think that's what

01:25:26.120 you're portraying. The situational awareness includes following the rules, doing the good.

01:25:31.180 And then the other is the defensive driving part is don't think everyone else is doing the same thing.

01:25:36.340 They're not. And I would just say whether they're intending to get you or not,

01:25:40.920 those bad choices are going on. Those errors are going on. And so, yeah, I think that's actually a

01:25:47.020 great way to approach it. And I think, unfortunately, it's so easy to drive nowadays. A lot

01:25:52.260 of folks, especially with even newer technologies, some people have their heads so far out of the

01:25:57.300 game, they've lost the situational awareness, right? It's all about the music, the conversation,

01:26:01.900 et cetera. So all of that is lost. So, no, I think that's actually when you're doing your other

01:26:07.220 training with the muscle memory stuff, that needs to be in there, too, because part of this is the

01:26:11.160 mental game of making the right choices and knowing that other people won't, then what can you do?

01:26:16.520 Yeah. Let's now talk about some of these more contributing factors in a little bit more

01:26:20.060 detail. And I want to start with substance abuse or substance use or whatever you want to call it.

01:26:24.100 We've talked about alcohol, of course. Give people a sense of what does it take to reach 0.08?

01:26:28.820 Because I suspect it's a lot easier to get to 0.08 than people realize.

01:26:34.740 Yeah. In fact, most people in the safety arena with alcohol like to say that impairment starts with

01:26:40.800 the first drink. And I think that's a great, like, as soon as you pull up alcohol, that's going to

01:26:45.220 start changing your reaction time, thinking everything's going to start changing. I've got

01:26:49.560 a great model to demonstrate this. So I drive a racing simulator and I've done this game where I

01:26:55.740 will have a glass of wine and get in the simulator. And to be completely clear, with a glass of wine,

01:27:01.860 I do not perceive anything in my own level of awareness. I don't feel a buzz. There's nothing like

01:27:08.820 that. However, in the simulator, there's a noticeable difference. Even at a glass of wine,

01:27:14.320 where I'm sure I am below 0.08, because I weigh a lot, and I don't feel buzzed in any way, shape,

01:27:20.340 or form, I don't drive as well. I miss the apex more. I'm more likely to spin. It's a subtle difference,

01:27:28.920 but it's absolutely noticeable in that environment, which is much, much more demanding than just driving

01:27:36.120 home from the restaurant. So, but as a general rule for someone my size, I've never taken a

01:27:41.860 breathalyzer test. So I don't actually even know what 0.08 is. So for you, that's going to be two

01:27:47.320 to three beers and one good shot of hard liquor. Over what period of time? Within one to two hours.

01:27:55.360 Okay. But you're bringing up a great point, which is why I always hesitate to give too much information

01:27:59.500 about that. Because you can go online with some pretty good BAC estimators, because your age,

01:28:06.400 your weight, your health. Exactly. All of those things, whether again, it's less alcohol content

01:28:13.160 in beers and wine, you know, versus again, hard liquor stuff. Over the time course, you were just

01:28:18.060 about, did you eat? And what did you have in your stomach when you read? It's like, there's so much

01:28:22.020 there. I think actually even more important is what you were telling in your story,

01:28:26.020 which is there's very often a disconnect between the subjective experience and the objective one,

01:28:33.380 which is that very often people think they're doing a lot better. We see this parallel in alcohol

01:28:38.040 and sleep loss. I'm doing great. I'm really fine. When if you look at the objective performance

01:28:43.420 measures, they're just doing horribly or they're off, but don't realize how off they are. So that

01:28:49.120 disconnects really important. That's why you say I'm good to go, whether it's not having enough sleep

01:28:53.820 or, you know, I had a couple of glasses of wine, but I'm okay. Had enough to eat and I'm good to go.

01:28:58.320 That disconnect, I think is actually bigger risk than sometimes the exact amount of alcohol you

01:29:03.180 have in your system. Yeah. And I would argue that the sleep one is just the same. If I don't have a

01:29:08.360 great night's sleep, it shows. And again, that's, I think that's, what's kind of cool about a simulator

01:29:11.880 is you're simulating driving around at 200 miles an hour. And at that speed, things are happening

01:29:17.720 three times faster than when you're in a car on the road, even on a freeway. And of course you're

01:29:22.960 turning constantly, right? So the stakes are so much higher and the forces and inputs are so much

01:29:28.320 more subtle. Any wrong input will lead to a significant change in output. And it is remarkable

01:29:34.780 how you can feel fine for the mundane task, but not for the drive. Add that to your course you're

01:29:42.300 creating, right? So besides the muscle memory, whatever, it's like, you'd love to challenge,

01:29:46.540 especially those adolescents with even whatever's appropriate there. Sleep loss would be better than

01:29:51.920 alcohol, obviously. But think about that as a challenge just for them to say, I feel fine.

01:29:56.820 Give me the rating on the one to 10 scale and then show the objective measurements in driving. Like

01:30:01.520 you say, even the mundane ones. Yeah. Let's talk about other substances. And I think the most relevant

01:30:06.720 is cannabis. What did the data say about cannabis use and driving? It's not good in that what's

01:30:14.400 happened is in alcohol. There's a really good established protocol with, you know, here's what

01:30:21.680 impairment means at point away. And by the way, we can use a breathalyzer. We can use blood. There are drug

01:30:27.420 recognition experts, DREs that can use behavioral measures on scene. And we know they hold up in court

01:30:33.660 with cannabis. There is still ongoing debate of what defines impairment. And so if you can't even get that

01:30:42.320 defined and then you have no roadside kinds of measures and it's harder, as you know, whether

01:30:49.200 you're using breath, hair, all these other things that people are trying to do, it ends up, well,

01:30:54.460 you know, it sticks around in your system longer or can be in the follicles for this long. It's like

01:30:58.740 there is no good measurement. So we don't have the impairment definition. We don't have good measures

01:31:04.640 to get us there to sort of figure out how we do that. And the other thing, which is really

01:31:09.580 dangerous, is the poly usage. It's rarely one of these on their own. With cannabis, you see alcohol

01:31:18.020 or other kinds of things that people are using at the same time. So that makes it even more difficult

01:31:22.780 to figure out, especially in the roadside kind of environment, of what people are actually using.

01:31:28.720 And so, I mean, this is now, I was administrator six, seven years now ago, and all this research was

01:31:35.380 still going on to try and define what I was just describing. That as all these states now start

01:31:40.160 making it more legal, I always tell people when you're voting for that in your state, don't think

01:31:44.880 about the sport of cannabis, but instead think about what that's going to mean if somebody next to you

01:31:49.500 on the roadway has been smoking and in some way is like the level of impairment of alcohol. People just

01:31:56.100 don't think of that part. And unfortunately, while the alcohol is pretty well defined, the cannabis part

01:32:00.480 is very poorly defined. In that time, we still don't have it figured out yet.

01:32:05.360 Is there a manner in which we can at least extrapolate from the standpoint of understanding,

01:32:08.520 right? So alcohol is a CNS depressant. I guess at the risk of just talking through something that

01:32:14.620 seems self-evident, what is the mechanism by which you think ethanol is contributing to mistakes?

01:32:23.800 We're talking now about the subset of mistakes that lead to bad enough crashes and potential

01:32:28.160 fatalities. Like, is it that people end up driving too quickly when they're drunk? Is it that their

01:32:34.200 reaction times are so slowed when they're drunk? And not even drunk, let's not even use that word.

01:32:39.700 When they are under the influence of alcohol, what are this suite of mistakes that they're making that

01:32:45.200 are leading to these catastrophic errors? So I think the two big categories are any performance

01:32:52.120 related ones, which are degraded or impaired. So that would be sustained attention, reaction time,

01:32:59.080 decision making. And the other is the one we talked about, and that is the subjective perception

01:33:04.360 part being disconnected, where they have no sort of, you know, their decision making, like it's really

01:33:09.220 okay. So we know basically with alcohol, all those CNS things that you were talking about are going to

01:33:16.480 be, again, be degraded or impaired in some way. And again, I talk about this the same way with fatigue.

01:33:22.580 And I'll just use that as the example for the moment, but there's this like continuum. Everyone

01:33:26.800 is like, you're asleep here, you're awake over here, but the alertness in between is a continuum.

01:33:32.340 And so I think most people would absolutely agree being asleep at the wheel, not good. But that

01:33:37.640 continuum beforehand, we know that with sleep loss, circadian disruption, alcohol, that basically

01:33:43.200 your performance will degrade way before you actually say fall asleep or drunk enough to go

01:33:48.520 unconscious. And that's why I call that degraded as opposed to the impairment where literally you

01:33:52.600 can't perform at all. But at least in the sleep realm, I mean, we can see 50% reduction in your

01:33:58.200 decision making, 30% reduction in your memory. We can see 50 to 75% increase in your reaction time.

01:34:04.840 These very specific things when you're in the driving mode, trying to deal with a situation are

01:34:10.260 going to affect you and your ability. And that's why I always tell people it's like way before you have

01:34:13.980 you have that fall asleep, go unconscious, you've got these performance degradations that put you at

01:34:19.160 risk. But in that sense, I mean, to me, sleep and cannabis seem very similar in that they're sedating

01:34:27.180 and there's clearly a spectrum from completely debilitated to loosely less functional. And there's

01:34:36.880 no real way to measure it directly the way you can use a breathalyzer for alcohol. Are there

01:34:43.320 differences between how we view drowsiness due to sleep deprivation or sleep interruption and

01:34:51.720 cannabis use, which again, I'm just bringing up cannabis, not because it's the only other substance

01:34:56.420 out there besides alcohol, but because as you said, it's becoming more and more ubiquitous. I think

01:35:01.220 there's a belief on some level, maybe somewhat warranted that it could be less toxic than alcohol. I think a

01:35:13.300 unintended consequences of that with regard to this domain.

01:35:18.260 Well, and I'm in complete agreement with you there, which is that I think it remains hypothesis,

01:35:22.980 but with everything we know about how physiology and the brain and things work, it's got to be.

01:35:27.480 We may not be able to quantify it quite yet. Here's a comment I often make about sleep loss

01:35:32.220 and circadian disruption. When you lose sleep, you disrupt the clock. All aspects of human capability

01:35:37.780 are degraded or impaired in some way. So I mentioned that because a new study comes out and says, oh,

01:35:43.280 we just link this now to cancer in a new way or immune function in a new way. And I'm kind of like,

01:35:48.860 okay, so now we have more data and we can be more exact in certain areas. Good thing.

01:35:53.660 But the overall comment holds, which is kind of what you're saying here too, which is everything we

01:35:58.920 know about how cannabis works in the brain, how it's going to affect us physiologically performance-wise,

01:36:03.560 we know that's not going to improve things. It's going to degrade them. And while we may get better

01:36:08.140 at quantifying how bad that is, let's be clear. It's not a good thing for driving under the influence

01:36:16.020 of that. I think that's what's going to happen here is that we will get there. There's some great

01:36:20.440 people at NIH and other places that are working on how to get that impairment level defined, to think

01:36:25.800 about what the test could be, et cetera. We'll get there. Took a decade or two, frankly, to get the

01:36:30.500 alcohol ones in the way they are now. We'll get there. But for the moment, let's be clear.

01:36:36.000 Everything we know is that's a negative. It's not a positive. And we may get better at it, but it's

01:36:41.260 still not going to improve things in some way. I'll just make another general comment, which is,

01:36:46.100 I think, the sleep part that's interesting. Everyone sleeps. They think they're an expert.

01:36:51.140 And it's kind of interesting that everyone is like, yeah, I've lost some sleep and, you know,

01:36:54.700 I'm still here. I'm fine, et cetera. Whereas, again, people will think about alcohol and cannabis.

01:36:59.320 Just, you know, that's more of a choice. And by the way, I don't drink or I don't do this,

01:37:03.100 et cetera. No one's ever going to say, yeah, I think that sleep thing, I just stopped it.

01:37:06.640 That's a requirement for just our existence. And so it's slightly different that way.

01:37:12.560 Are there any other prescription drugs that play a significant role in this? I mean,

01:37:17.860 the obvious ones that come to my mind would be benzodiazepines. But is there any other class

01:37:22.220 of drugs that frequently enough show up on the impairment side of the equation?

01:37:27.020 Thank you for bringing this up because drunk, drugged, people always think of cannabis,

01:37:33.120 et cetera. But that includes prescription medication and over-the-counters. So any

01:37:37.580 antihistamine that makes you drowsy, any sedating antidepressant, trazodone gets used,

01:37:42.360 any one of those that you're taking has the potential to affect your sleepiness alertness

01:37:48.180 level when you're actually driving. So I'm glad you brought that up because everyone thinks of the

01:37:51.800 big ones, alcohol for sure, cannabis, we got to start talking about. But these others,

01:37:56.600 you think about it, prescription, something you don't want to just read the label. You want to

01:37:59.820 talk to your healthcare provider about that and see how much could that affect me? And the over-the-counters

01:38:03.860 as well. Those are all big issues. And we, again, it's the polypharmacy. You see that a lot,

01:38:08.740 that there's multiple things, but trying to pull out exactly what they were,

01:38:11.900 that's often after the fact. Post hoc, you go interview people and say, oh yeah, it's allergy season.

01:38:16.820 And are you taking the one that's sedating or are you taking the one that's the non-drowsy version

01:38:20.520 of it? Let's talk a little bit about speed, which still makes up a pretty sizable fraction of the

01:38:26.720 contributing forces. Is that changing? Is that less today than it was 30 years ago? It strikes me that

01:38:34.440 clearly cars are faster today than they used to be, but are people necessarily driving them faster

01:38:39.060 today? And is that not offset at all by the far greater impact of technology, which we haven't come to

01:38:45.540 talk about yet. So we should certainly spend some time talking about seatbelts and airbags, but

01:38:49.020 yeah, what's the general throughput of speed in this equation? It's a big factor. Force equals mass

01:38:55.280 times acceleration. It's all about the energy. So when you're going faster, you're talking about when

01:38:59.680 a crash happens, there's just more literally impact of what's going to come out of that particular

01:39:05.220 collision. And so it's interesting. Speed has actually gone up and that has more to do with the

01:39:09.280 capability of the vehicles, roadways, and the technologists. Think about, I mean, you're a racer.

01:39:14.100 There's always like, how do we get better, faster, et cetera.

01:39:17.140 Although it seems that cars outside of SUVs, probably cars are getting a lot lighter.

01:39:22.160 So when you think about mass times deceleration, which I assume is a big part of what that force

01:39:28.140 is that people are experiencing, does that not offset it on the car side of the equation?

01:39:33.260 Not very much because generally cars are getting bigger. Now that we've added SUVs and others,

01:39:39.600 when we look at what's going on out there, there's people describe it as a weight bloat that's

01:39:43.740 occurred. And unfortunately, they do protect you. If you're in one of those energy related

01:39:49.740 kinds of crashes, the bigger, the protective cage you have around you, the better off.

01:39:54.260 People have a tendency to say, I want to be in that vehicle, not in the little small

01:39:58.040 version where I could get killed doing that sort of thing.

01:40:01.700 I think what's interesting is we do see, and this is why it's so complicated, but we do see

01:40:06.420 during the last recession, during the pandemic, the deaths go down, but sometimes other particular

01:40:13.140 things like speeding goes up. And we can have all kinds of hypotheses about, are there fewer

01:40:18.580 people on the road? Is just the density different, et cetera? They're all hypotheses. We think about

01:40:24.020 those. We don't really know. But I bring it up because it's interesting that even the economic

01:40:29.380 environment of our society can end up having effects beyond the usual causal or contributing

01:40:34.640 factors, for example, that we're mostly focused on in this conversation. So even those things can

01:40:40.100 have that societal ripple. And I would just say for the speed side, two other things, which is,

01:40:45.940 I'm sure we'll talk about the safe system approach, and that has to do with safer roads. And one of the

01:40:51.100 things they look at is controlling speed through road design. So that's where roundabouts or what they

01:40:57.300 call speed diets, where you basically two become one lane, and then instead we're going to have a more

01:41:02.180 dedicated pedestrian or cycling lane where you can use, et cetera. And then the last thing I'll just

01:41:06.760 say on the technology is there's a huge debate. It's not where the alcohol is, where the technologies

01:41:11.960 pass legislation, but there's a lot of discussion about speed limiters now and whether or not, as we've

01:41:18.300 been discussing, and I said this earlier, society very often when we have human choices or things that

01:41:23.780 we want to deal with, we go to technology. So all of a sudden now, speed cameras and speed limiters

01:41:29.500 and vehicles are in more discussion than they have been in a decade.

01:41:34.340 Yep. Makes sense. All of these things are, you'd pull them out of our dead lifeless hands if you

01:41:39.820 don't let me drive quicker and don't let me talk on the phone and whatnot. I want to go back to

01:41:44.980 another kind of deconstruct the accident question. Do you have a sense or are these data knowable

01:41:50.320 as to what fraction of those fatalities in the vehicle are a result of the integrity of the

01:41:57.600 vehicle being lost due to the collision itself versus a flip? In other words, how often is avoiding

01:42:04.060 the flip of the vehicle with a lower center of gravity a relevant part of this equation?

01:42:10.540 Yeah, that's actually a big deal. And when you look at the crash tests are done, it's interesting,

01:42:15.560 they don't actually crash vehicle to vehicle. The crash worthiness of your vehicle is tested against

01:42:20.080 a big block of cement, basically. But now there are different versions of that. So there's a side

01:42:24.960 impact. And now over the last years, there's also an angled one as well, and then rear end,

01:42:31.240 and then you can add all kinds of other variations of that as well. And I would say that's the kind of

01:42:36.020 thing I don't keep in my head anymore. But there are some databases that gets at least into things

01:42:41.820 like rollover. And that's where the side impact is critical, right? Because that's mostly going to

01:42:46.120 happen with the center of gravity when you're actually have, you're up on a curb or some other thing

01:42:50.760 that's going to tip you potentially, and again, the direction that you're hit, etc. But there is

01:42:54.840 some data on that to show sort of which are the versions. And that's why they added things like

01:42:59.320 the angled collision, basically, recently, just to figure out that, you know what, there's enough

01:43:03.820 of that going on, we should understand the crash worthiness of vehicles when they're hit at an angle

01:43:07.680 like that. Are electric vehicles indeed significantly safer than non-electric vehicles on the basis of a

01:43:13.780 lower center of mass? And on the basis of not having an engine, typically in the front during a

01:43:19.780 front collision, therefore it can have much more force absorptive capacity? Or is that more of a

01:43:24.940 marketing strategy? We don't have sufficient data yet to sort of make that factual comment per se.

01:43:31.700 But I think you've identified all of a range of the factors that actually differentiate electric

01:43:36.360 vehicles from the standard. And as you know, right now, one of the biggest things is just getting the

01:43:41.280 myths separated from the fiction that is told. There are still more fires in ICE vehicles than there

01:43:46.100 are in EVs. Some of that's about battery. Some of that's about just the population that's out there of

01:43:50.820 these different vehicles. I mean, it's like, you got to get into that level of data analysis. That's why I'm

01:43:55.800 talking about the segmentation and stuff. So some of these things, again, NHTSA.gov is one place, the

01:44:00.820 Insurance Institute for Highway Safety, they do a lot of great crash testing. And I'm sure we'll get into some

01:44:05.540 of that when we talk about some of the vehicle specific things. But you can get a little bit more of

01:44:09.780 that data. My caution is always that some of that is still in development, as far as

01:44:14.300 understanding the kind of differences. Hypothesis, like on what you were just identifying, but the

01:44:18.940 level of segmentation to actually make factual statements, I think we're still developing a lot

01:44:22.760 of that. Let's talk about where autonomous driving comes in. If you believe that north of 90%

01:44:30.980 of these crashes have at their root cause errors by humans, then you would think that an autonomous

01:44:41.960 vehicle provided the entire network of vehicles is autonomous. That's always been the big if,

01:44:47.340 right? It's not enough to just have some autonomous vehicles. That could actually make

01:44:50.680 things worse. You have to kind of have everything being autonomous, would certainly have the potential

01:44:55.500 to do better. I don't think it goes from 94 to zero, but it seems that there's a glad path.

01:45:00.440 What is your view of autonomous vehicles and the hope that they bring to this problem? In many ways,

01:45:07.420 I wonder, is that the solution as opposed to built-in speed limiters, breathalyzers? I was also

01:45:13.840 going to half-jokingly suggest that if we can put all of those things in cars, we can also probably put

01:45:19.520 in eye flicker sensors and we can track micro-sleep because I think back to when I was in residency,

01:45:26.440 how tired I would be driving home and I could barely hold my eyes open. Clearly, we could sense that.

01:45:32.520 So there are clear things you can build in to make humans less likely to hurt themselves.

01:45:38.420 But at some point, you might just say, let's just put all of our energy into autonomous vehicles.

01:45:42.860 So this is a fun part of a conversation, which is the general comment we've already been discussing,

01:45:49.840 which is when society has issues that are around human behavior and choices and errors we make,

01:45:55.460 then do we often look to technology as a way to help us do better, save lives, improve situations?

01:46:02.520 So there's another study that NHTSA did where they actually looked at 14 different technologies as

01:46:09.280 straightforward as seatbelts, airbags, electronic stability control, 14. And over 50, 52 years,

01:46:16.220 how many lives did those technologies saved? 613,501. And those are just the 14 they looked at,

01:46:23.620 which were related to the Federal Motor Vehicle Safety Standards and other things that are

01:46:26.960 in vehicles. So you can't do this. But I point out at 40,000 lives a year, we could have gone 15 years

01:46:35.920 with no lives lost on our roadways. Kind of our conversation, not necessarily injuries or crashes,

01:46:41.620 but technology works. And I'll say this as the caveat to start with is the potential is unbelievable,

01:46:48.540 but we still got to prove it's going to work, fully acknowledging that these new systems will also

01:46:54.920 introduce new risks. Okay. I think again, when I told you people like to point their blame,

01:47:00.100 then it does like still got to prove it. And we also got to acknowledge that now we're going to

01:47:04.320 have software risks and other kinds of machine learning and technological risks that are introduced.

01:47:10.320 We got to make sure we address those as well. Your point that we may get to zero in some areas,

01:47:14.920 should we even expect that in others? Having said that, what you just described actually is the way I

01:47:20.340 think we need to be using technology. So those are really in two big areas. The advanced driver

01:47:28.920 assistance systems, ADAS, are the things that are already in cars. I'm sure you've got them in yours.

01:47:35.540 Automatic emergency braking, lane keeping assist, cruise control, rear camera, all these different

01:47:41.820 systems that are independent. And so when you think about it, those are a part of, again,

01:47:46.680 advanced driver assistance is how do we support the driving task by giving you these things that

01:47:52.040 we know when you look this way, you can't be looking that way, but we could give you visibility

01:47:56.240 to that situational awareness again that could actually help with that. The higher level is a

01:48:01.580 fully autonomous vehicle. And let's just say, you know what? We're going to take out the steering wheel,

01:48:06.700 the pedals and everything else. It's going to do the entire task. And the way sometimes this is

01:48:11.420 separated is at the lower levels. SAE actually has these five levels, and I'm not going to get into all of

01:48:15.980 that. But you can think about it basically as at the lower levels, the human is responsible for

01:48:21.620 monitoring the environment and the vehicle. There's an interim level, level three, where the vehicle

01:48:29.120 will do some of those things. But if it needs you back in the loop, it's going to let you know,

01:48:33.980 take the wheel. And level four and higher is vehicle is going to do it for you. It's going to

01:48:39.060 monitor the environment and the vehicle and handle the driving task.

01:48:42.740 Just so I understand, Mark, is Tesla the best example of, is that the furthest along technology

01:48:49.200 in the driver assist commercially? I'm not sure it's further along, but I would say it has the

01:48:55.060 most integrated systems of pulling those separate systems in. But what you just said that's so

01:49:00.320 important is the label that often gets used, autopilot, gives a sense that it is autonomous

01:49:06.300 driving, when in fact, it's what you just said. It's actually a collection of driver assistance

01:49:12.180 systems, which actually even they publicly, because of regulatory requirements, that's level two.

01:49:17.640 I was just about to say, is that level two or three? But it's level two.

01:49:20.780 Yeah, you got it. You got it. And probably that's one of the more integrated systems that uses

01:49:25.340 all these different things. But you're seeing even the traditional automakers basically offering all

01:49:30.680 these different systems that are in there. So one of the ways, take a step back, is to realize if

01:49:37.340 we're going to take the 280, 300 million cars that are on the road today, it can take anywhere from 10

01:49:44.300 to 12 years to actually penetrate the fleet with new technology.

01:49:50.680 That seems optimistic. I was going to say probably longer. I mean, it would seem to me at least 20

01:49:56.120 years to overturn that fleet. And that would also include, like, this is going to be one of those

01:50:00.680 moments where I imagine a real regulatory challenge. So I love internal combustion engines. I am an

01:50:08.120 internal combustion engine junkie. And not only that, I love naturally aspirated internal combustion

01:50:12.720 engines. And I'm not going to embarrass myself by saying how many of those I have. But I have a belief

01:50:19.180 that those things aren't going to be made for very long. And that's why I love to have them. I like the

01:50:23.940 engineering. I love the sound, et cetera, et cetera. So for example, like a 1967 Corvette, you can still

01:50:29.560 drive it today. Even though it doesn't meet a single safety or emissions standard that exists

01:50:35.360 today, it gets grandfathered in as you go along. But when we start to think about autonomous vehicles

01:50:41.540 and you get to this point where you say, well, look, for the system to really work, every vehicle

01:50:46.580 must be a level four or level five. That is a totally different regulation. That's no longer just

01:50:52.460 saying to the car producer, you can't make a car that doesn't meet these requirements. It's saying to

01:50:57.920 the consumer, oh, by the way, you can't have that car anymore. Is that what's necessary here?

01:51:03.020 My one thing for you is horses. When people say, I'm never going to be able to buy my ICE vehicle

01:51:09.020 again or one in the future, it's like, you know what? People still breed them, raise them, race them, 0.96

01:51:16.380 own them. They're always going to be ICE vehicles. You're going to still have it around and do what you

01:51:21.540 need to do. But I think it also brings up the enormity of the problem.

01:51:25.600 Yeah, this is the biggest infrastructure problem I can imagine.

01:51:30.820 Yes. And that's why when I was administrator, we actually put out the first federal automated

01:51:36.360 vehicles policy and people were screaming, well, that's a nice policy, but like, what are you going

01:51:41.240 to regulate? And I used to say, what are we going to regulate? It's just a huge issue. The autonomous

01:51:46.580 vehicles aren't there yet. There's some great programs going on demonstrating their potential,

01:51:50.740 but they're not there yet for full deployment throughout our entire society. They're not

01:51:56.060 there yet. They're great demonstration, pilot programs, larger ones. And there are a lot of

01:52:01.680 problems that are going on as well with some of the companies. So that is still, again, to be proven

01:52:06.900 and understanding where the risks are with that model, we still need to get there. That's why I think

01:52:12.280 one of the ways to conceptualize this is we're probably on a path. There used to be a debate,

01:52:18.020 and I think we know what the answer is now. Do we just go along the levels, one, two, three, four,

01:52:22.880 you know, until we get up to four and five fully autonomous, or do we jump right to four and five

01:52:26.740 and let's go for the full thing? I think we're seeing it now just by what we're talking about.

01:52:31.320 ADAS is that stepwise, get people familiar with the technology, see where the advantages are,

01:52:37.120 see where it helps you, see where it might actually save your life, et cetera. That'll move us toward

01:52:41.740 it to where eventually it's like, I'd like to just get my hands off the wheel, period,

01:52:45.740 at least for my commute or something else, let the vehicle do it. Again, if you want to race or do

01:52:50.540 something else, that's a different environment, different task you want to do. But I think now

01:52:55.800 that question's been answered that we're probably going to be going through these steps

01:52:58.980 where technology is introduced. We need to get enough data to show it's really going to make

01:53:04.100 a difference. And then when we have that, I think people are like, okay, it's built in,

01:53:07.980 it's working, et cetera. And then there'll be more comfort at some point. The two examples I'll give you,

01:53:12.720 one is when I was administrator, one of the things we did was actually challenge the auto industry

01:53:18.820 to make automatic emergency braking standard on every new vehicle by 2022, or at least 95,

01:53:25.780 99%, like of all the new vehicles. We originally had 10 manufacturers. And a couple of months later,

01:53:32.040 when we launched this program to try and get this going, we had 20 manufacturers basically covering

01:53:37.840 about 95 to 99% of all new vehicles. And the whole idea was they would get in the room and spend some

01:53:43.360 time saying, what are the criteria that we want to see everybody? It's called democratizing safety.

01:53:48.420 So it's not just if you buy a higher end car, or it's an option that I can, it's like, no,

01:53:53.060 democratize means every new vehicle, AEB. And the Insurance Institute of Highway Safety says,

01:53:57.860 if every car had AEB, we could probably reduce 50% of the rear end crashes with that.

01:54:03.260 Wow. Yeah. And just to be clear, Mark, this is not the warning that comes on when you're about to

01:54:09.700 hit somebody, but it actually will break if the distance and rate of speed between you and the

01:54:16.320 car in front of you triggers an algorithm that says, you're going to hit this thing.

01:54:20.180 Peter, this is why it's so great to have some time to discuss this stuff. AEB actually has multiple

01:54:25.300 elements to it. The warning is one. Another is if you don't break, it'll break for you.

01:54:31.100 Another one is if you break, but you don't break hard enough because of the distance,

01:54:35.060 it will actually add breaking power to what's going on. And so that was what this working group

01:54:40.840 was, is let's figure out what the performance criteria will be for this AEB mandate that we

01:54:47.440 want to see. Now, why I'm bringing this up is because it was not a regulation. And pretty much

01:54:51.840 that's because every regulation that had come through recently, like when I was there, was 10 or 12

01:54:57.380 years in the regulatory pipeline. And so we said, you know what, let's just challenge them

01:55:02.420 and see who would agree. And like I say, we ended up having 20 manufacturers come together.

01:55:07.320 And I mentioned that because somebody just put a report out that basically they met that requirement.

01:55:12.640 Now, why is that relevant? Besides the fact that everybody's got AEB and all those new vehicles,

01:55:16.900 again, to our point, that's not retrofitting everything else. So there's still plenty of cars that

01:55:21.600 don't have this yet. It's going to take a while for all of that to change. The other thing that most

01:55:26.820 people, it's not a secret, but most people don't understand this, but to get a regulation through,

01:55:31.640 you need to have an office management budget basically do a cost benefit ratio, right? They

01:55:36.540 got to do a calculation that says it's worth it. It usually takes a penetration of at least 10% of

01:55:41.500 the vehicle population to get enough data to do that kind of analysis. OMB needs to do an analysis

01:55:52.200 of something because it's the government that's mandating it. That's why they have to be able to

01:55:59.320 say, yeah, if we mandate this, the cost to the industry to do this is X, but then how do they

01:56:05.280 assign a cost to life? Oh, the numbers are there. Always debated, but do a search on it. Yes, but the

01:56:11.060 number of the cost of life doesn't include loss or pain and suffering to the people who are left

01:56:16.220 behind. It's economic loss of life, meaning you lost 10 years of your life working where you would

01:56:21.940 have made this amount of money. So it's probably an underestimate of the true value of life by far.

01:56:26.780 Absolutely. And part of what I'm bringing up is you typically need at least 10% penetration in

01:56:31.940 the vehicle population, even to collect the most basic data to send it to them. It's like if it doesn't

01:56:38.580 even have the data for that cost benefit, so they can do their straightforward lives versus we have

01:56:43.560 this technology, how many lives you think you're going to save? And what data do you have to

01:56:46.480 demonstrate it and substantiate it enough for the cost to the car, the manufacturer, the society?

01:56:53.580 So again, I think that's one of the challenges you see going on to what we were saying before is that

01:56:58.440 seeing these stepwise additions to these driver assist systems are getting added. Things like AEB that

01:57:05.480 are now standard allowed that. So the agency just came out recently with, there's something called an

01:57:11.100 advanced notice for proposed rulemaking and a notice for proposed rulemaking. I don't know

01:57:15.180 which one it is, I forget, but basically they're now going to regulate AEB. And I like to point out

01:57:20.240 as part of the reason that they'll be able to do that at an even better level, like maybe extending

01:57:24.300 it to trucks, heavier vehicles, et cetera, is because of this other work that was done that made

01:57:29.300 democratize it as something that needs to be out there that everyone should be able to have in

01:57:33.460 their vehicles. So again, I think we will see this stepwise going, but I also point out that when we want

01:57:39.900 to get to the full autonomous vehicle, you said this earlier, I want to bring us back to it.

01:57:45.260 The mixed fleet is going to be a huge problem for everybody. And if you think about it, it's going 1.00

01:57:51.080 to be at least three groups, fully autonomous vehicles, vehicles that have a lot of the ADAS

01:57:56.880 support systems, and people that are just driving cars that don't have any of that stuff in it.

01:58:01.780 And they'll all be on those streets. You won't know which is which necessarily.

01:58:05.560 Mercedes just came up with a different color light for when it's in autonomy for some of the

01:58:10.400 things that they're doing. But earlier, our conversation about defensive driving is like,

01:58:15.080 just think about how more challenging that kind of environment is going to be.

01:58:18.960 Well, let's touch briefly on drowsiness. I realize this is kind of your life's work.

01:58:24.120 It's also the one that I think people who listen to my podcast regularly are very familiar with the

01:58:29.120 implications of sleep deprivation. What are the most important things you think you want people to

01:58:33.340 understand with respect to how compromised sleep impacts driving?

01:58:39.600 I'm going to repeat what I said earlier, which is lose sleep, disrupt your clock, the circadian

01:58:44.580 clock in our brains. Lose sleep, disrupt the clock, and you will pay for it by having human 0.89

01:58:51.280 capabilities degraded or impaired. That is across the board. So I don't care what your job is or what

01:58:59.100 task you're on. But if you're not getting the right sleep at the right time, you're going to

01:59:04.040 pay for it. This is one of those you can't fool human nature, just nature, period. You can't fool

01:59:08.880 it. There will be a price to pay. And is there any regulation about when a truck driver can drive?

01:59:14.380 I understand there's clear regulations about how many hours he or she can drive. But are there

01:59:18.980 regulations that say, we'd like you to stay within your circadian rhythm and do your driving?

01:59:23.780 We'd like you to see a sunrise and sunset and drive accordingly, even though I know that that's not

01:59:28.820 necessarily convenient and it might be far more convenient to drive through the night.

01:59:32.240 So what's really interesting is if you understand sleep in the clock, some of what we would do as a

01:59:38.360 society is exactly opposite the way the world actually runs. And so there are regulations in

01:59:45.540 aviation and trucking and rail and a lot of other places, as you already mentioned, the resident

01:59:52.100 training hours, et cetera. To your point, I always point out there's the two physiological elements

01:59:58.580 you've got to deal with, the sleep part and the clock. Usually it's easier to deal with the sleep

02:00:02.580 part. And people still don't do that well. But as you know, just to be clear, if you don't have 10

02:00:08.160 hours off your job, commuting, getting eight hours of actual supine sleep opportunity doesn't mean

02:00:13.980 you're going to get the sleep, but at least eight hours opportunity. Maybe nine hours if what residents

02:00:18.460 used to do is live in the hospital, maybe. But so yeah, you can have a little bit better on the

02:00:23.080 sleep part. But to your question, the circadian piece is a lot harder. And keeping that stability

02:00:29.380 in a 24-7 society, military operations, air traffic control, moving goods, operating airports,

02:00:36.140 it's like these are 24-7.

02:00:38.120 This speaks to, again, I think the general theme here that I'm hearing, Mark, is we're going to

02:00:42.520 quickly approach the limits of what humans are willing to do in terms of the sacrifices they're

02:00:48.160 willing to make. And therefore, we have to come up with technology to work around that. We're not

02:00:54.180 going to be willing to live in a society where nothing happens at night. So we have to automate

02:01:00.000 those processes so that we don't have to rely on individuals functioning at fractional reserve and

02:01:06.600 fractional capacity.

02:01:08.380 You got it. But we are so far away from that in our society now.

02:01:11.460 Yeah, it might not be in our lifetime. You and I probably won't appreciate the full extent of

02:01:17.080 pick your favorite sci-fi movie where the trucks are just autonomous vehicles that are transporting

02:01:22.100 things on freeways at night.

02:01:24.060 Right. I say that because start at the beginning, which is the information, even naivete and ignorance

02:01:31.980 of society around the dangers, not just the performance, but that is still an emerging.

02:01:37.860 I mean, it's been going on for a while. I feel like Bill DeMent, who-

02:01:41.460 wasn't just a professor of mine, but became a friend and colleague and teacher and mentor,

02:01:45.840 all those things. He was probably one of the leading voices who said, we got to pay attention

02:01:50.180 to this because the price is too high. And he always pointed out that if we don't start with

02:01:54.580 education, so that people appreciate not just what the cost is, but for me and Bill, there's a

02:02:00.080 lot of like, if you appreciate what the benefits are of getting the sleep you need and keeping the

02:02:05.240 schedule, et cetera, that should outweigh putting yourself through this other stuff. That's the

02:02:10.300 individual piece. But then it gets to what you just described is if we really accept that as a society,

02:02:15.320 then we have to move to what are the solutions going to be, right? And again, it's not always

02:02:20.140 intervening on the individual, give them drugs. It's like, no, it doesn't have to be that. It's

02:02:24.340 going to have to be these other things like technology that's helps us offset those choices

02:02:29.060 we make as a society that could still put us at risk. Yeah. Let's close our discussion, Mark,

02:02:34.760 with a couple of things that people can take away as far as actions that they can take to

02:02:40.920 avoid hurting others and being hurt. And let's start with that of being a pedestrian.

02:02:46.700 pedestrian. So what can an individual do as a foot pedestrian to minimize their risk? Let me start

02:02:54.820 with a basic question. How many times is a pedestrian killed when they have the right of way in a

02:03:01.420 situation? So they're not jaywalking across a red that they shouldn't be. They're doing the right

02:03:06.560 thing. They're either up on the sidewalk and a car comes up on them or they're crossing when they

02:03:11.500 should be crossing and someone runs in there. Do you have a sense of that? I don't. It's a great

02:03:16.080 question. And I'm not sure we have that data just because of the categories that these would be

02:03:21.160 placed in. Because again, if they weren't doing something illegal, I'm not sure that that kind

02:03:25.440 of information would necessarily be collected. So in other words, we could probably back into it

02:03:30.060 if we, I assume the data are tabulated for every time the pedestrian is quote unquote at fault. And I

02:03:35.960 know we're trying to get out of the at fault mentality here and just solve the problem. But you could

02:03:41.040 say, look, of the 42,000 people who died, 6,000 were pedestrians. And we have a really clear example

02:03:48.500 of a thousand of those pedestrians were the ones at fault. And the driver tried to avoid them at the

02:03:54.200 last second, but couldn't. And then we could maybe say, well, look, in 5,000 of those cases,

02:03:57.920 the pedestrian was not. But do you have a sense of what a pedestrian can do specifically around

02:04:03.320 their situational awareness? What do they need to be on the lookout for to reduce those odds?

02:04:08.580 You bet. And I think it is worth noting, again, in the last decade, that number has gone up 50%.

02:04:13.920 This last year, the estimates are 7,500 people have died as pedestrians of those numbers. So

02:04:20.960 that's significant. So to the specific actions, the choices that you have,

02:04:25.840 one is if there's a sidewalk, be on it. And again, that's just separation right from the driving

02:04:31.620 that's going on. If there is no sidewalk, you actually want to walk against the traffic.

02:04:36.620 So that's for visibility stuff of what's going on. You already mentioned it. It ends up,

02:04:42.320 I mean, we do know 64, 67% of those fatalities occur outside of intersections. So they're happening

02:04:50.800 elsewhere. But it does mean that if you're crossing, you want to find an intersection that's got a

02:04:55.200 crosswalk if you can, and has some system of stop signs or counting for when you should go.

02:05:00.720 And you want to follow those rules as best you can. Okay. The other one, I think, is visibility.

02:05:06.740 I can't tell you how, especially at night, which is when most of those occur, over 50% are simple,

02:05:13.020 straightforward visual things, making sure people can see you. And the other one, I think,

02:05:17.320 nowadays in particular, we mentioned it for driving, but situational awareness is critical.

02:05:21.620 And so it's amazing how many people could be either drinking or on their phone with head down.

02:05:26.660 So distraction, the kind of things you think about as a driver, you would apply to pedestrians and

02:05:31.640 cyclists as well. Are cyclists included in that 7,500 deaths per year?

02:05:36.780 No, I think those are just pedestrians. And I don't remember the cyclist number, but it's there.

02:05:41.740 Yeah. As someone who rides his bike still outside, though, not nearly as much as I used to. And

02:05:47.000 frankly, a lot of that is due to just my lack of faith in drivers. I certainly adhere to the

02:05:52.860 principle of always assuming the car doesn't see me and always assuming that the driver is a moron. 0.99

02:06:00.080 And those assumptions are surprisingly accurate more often than I would like to acknowledge. I mean, 0.99

02:06:07.440 the amount of people that will pass you, so they've presumably seen you, only to then want to turn

02:06:13.520 right directly in front of you is remarkable. But yeah, I think all of those make sense. The walking on

02:06:20.540 the other side of the road, the being completely aware. And I just think as a general rule, as a

02:06:25.420 pedestrian, also assuming drivers don't see you, they're calibrated to look for bigger, faster

02:06:31.580 things than you. So you just have to kind of assume they're not seeing you.

02:06:37.140 You know, in California, pedestrians have the right of way wherever they are. But I always point out,

02:06:42.180 having a lawn on your side.

02:06:43.160 Having a lawn on your side means nothing if you're hit and dead or injured. I always find that to be

02:06:47.700 the dumbest argument ever, which is assume you never have the right of way, even when you do have 0.95

02:06:54.000 the right of way, when you have that big green light and the big walking sign, still assume that a 0.99

02:07:00.300 driver is going to make a mistake because the consequences are much higher for you than them,

02:07:05.120 even though legally the law's on your side. Anything else, Mark? We've sort of throughout this

02:07:10.520 discussion infused as much insight we have as far as behaviors that you can take into your own hands.

02:07:17.640 And aside from the obvious, I think the obvious ones are so clear. But I do, again, want to

02:07:22.140 reiterate this thing of looking both ways through the intersection. I always tell people left, then

02:07:28.020 right, then left, because it's that left one that's going to kill you when you're driving. So left,

02:07:33.480 right, left scan before the intersection when you have right of way. The other thing for me,

02:07:39.240 Mark, is when I'm on a four-lane road, two in each way, I'm never in the left lane unless passing,

02:07:46.960 just because, especially with these non-median roads out here, it's that much harder for somebody

02:07:52.620 to cross and weave into my lane. So it's staying in the right-hand lane. It's really funny. We've got

02:07:58.640 these 75-mile-an-hour roads out here that have no median, and there are typically two or three lanes

02:08:05.220 only in total, two in one direction, one in the other. And sometimes they're one in one.

02:08:10.240 And those are nerve-wracking. You know, a one-on-one 75 with no median. Those are roads

02:08:17.060 where I will not talk on the phone. I will not listen to a podcast. All I'm doing is imagining

02:08:23.140 how quickly could I veer off to the right if a truck got into my lane.

02:08:27.960 If I can just a slight nuance to what you said is part of the defensive piece of that is don't

02:08:34.560 expect or assume that people are going to actually follow the rules. And so you may be, you know,

02:08:40.740 at that intersection and looking, but it's like, don't think because someone's got a stop sign

02:08:45.340 there, don't assume they're going to follow the rules, whatever they are. And that includes yellow

02:08:50.080 lights, red lights. No, if there's someone there and their potential risk, your choice is whether you

02:08:55.820 decide to proceed through, slow down, pause. Those are the things that are under your control.

02:09:01.420 Yeah. It's really funny. Two days ago, I was driving somewhere and I was in the right-hand lane

02:09:06.200 and I was going to make a right-hand turn into something. And I didn't actually realize that

02:09:12.500 there was a driveway before where I make my right. I put my right blinker on, and this is a pretty

02:09:18.120 quick road. It's about a 50 mile an hour road. And I'm coming up to, I'm now probably about,

02:09:24.800 I don't know, 300 feet before where I'm about to make my right. And a woman came to the edge of the 1.00

02:09:32.220 road where, again, I didn't even realize there was a road. It's easy to miss road. And she looked,

02:09:37.680 saw me, but saw that I had my right blinker on and just pulled out, assuming I was going to turn

02:09:43.680 right there. Now you could argue, well, technically, I guess she's right. I had my blinker on,

02:09:48.020 but I had that blinker on long before she showed up thinking I'm turning, you know, whatever,

02:09:52.140 three, 400 feet up the road. And again, I had to really slam on the brakes not to hit her.

02:09:59.060 And she would have taken the brunt of that, not me. Regardless, again, it's not about fault. It's

02:10:03.580 what is the consequence of this awful collision? And I was amazed that she pulled out. She pulled 0.98

02:10:10.300 out when my rate of speed hadn't even begun to slow. And I wondered to myself, how does she not see

02:10:16.560 how fast I'm going? Like, what did she think I was going to do? Somehow, like I'm on a rail that's

02:10:22.380 going to allow me to turn that quickly. But see, that's a perfect, perfect example of what we just

02:10:27.340 talked about. Don't assume people are going to follow the rules or that what you're seeing is

02:10:31.200 what's going to happen. And to what you just said is she wasn't thinking, you know, she just assumed

02:10:37.380 that that right signal meant you were turning right into where she is.

02:10:40.680 And my point is, even though maybe I was at fault, when it's all said and done, maybe it's my fault

02:10:45.400 that I had the blinker on when I wasn't planning to turn right there. But I think taking fault out

02:10:49.520 of this is the right way to go. And instead thinking root cause effect, root cause effect,

02:10:56.040 because the only accident I've ever been in my life, Mark, was a very bad accident where the

02:11:01.840 other person was 100% at fault. They ran a red light while I was going correct speed limit,

02:11:08.940 which was 50 miles an hour through an intersection. And luckily for this woman, 1.00

02:11:14.500 I hit the passenger side, not the driver's side. She was the only person in the car.

02:11:20.640 I'm in a pickup truck. She's in a sedan. I hit her so hard that both vehicles were a write-off.

02:11:29.100 It's the only vehicle I've been in where the airbags deployed. And what struck me is

02:11:35.160 how amazed I was at how quickly it happened, because I was driving this way and she was

02:11:42.660 stationary to make a left. So it wasn't like I was looking at her. I saw her the whole time,

02:11:49.520 but never thought she would just jet out in front because it wasn't in my usual going through the

02:11:55.740 intersection, somebody over there, somebody over there. It was the woman right in front 1.00

02:11:59.320 who erroneously thought she had a green light to go, even though she had a red arrow.

02:12:05.300 And she just pulled out slowly. And I did probably get the brakes on before I hit her. 1.00

02:12:10.880 But I think about that. And I think had she had a passenger in that car, could have been a very

02:12:16.920 different situation. And it happens so quickly that you don't have to make the mistake to suffer

02:12:22.820 the consequence. I really appreciate you telling these stories. Some people are good at it. Some

02:12:28.120 people don't like to talk about them, but this touches everybody. I mean, you've lost someone

02:12:33.560 in a roadway and it's like the emotions you were experiencing are so justified and understood and

02:12:39.400 your approach to how you think about it now that are based in all of that. Same thing here. Those two

02:12:44.860 drivers you just talked about with different things. It's like neither one of them thought,

02:12:48.920 again, I'll go out today and see if I can get killed. It's not just the other way, right? It's that

02:12:53.000 one too. It's like, let me see if I can run into somebody to do it. It's like, they're not doing

02:12:56.960 that. But it's a complex dynamic environment with humans that are imperfect and make bad choices and

02:13:03.620 also errors in what they do. And how do we try and do that more safely? And I think, again, those

02:13:08.680 encapsulate everything we've been talking about. Big societal changes. How do we make this better

02:13:12.980 for everyone? But a lot of it comes down to your behavior, what you can control, the stuff that's

02:13:18.700 out of control, you can still do something about in some situations. I think it's also, we didn't talk

02:13:24.560 about this much, and you made just one quick comment in there. There's a whole culture around

02:13:29.160 cars that's part of what's going on here too. I mean, it's our independence, it's economics,

02:13:35.900 it's family, it's just sort of the American way, right? Don't tell me how. So there's a lot going

02:13:41.760 on here that also are at play, which I think when we talk about, are we ever going to have fully

02:13:46.640 autonomous so we can save lives? I'm not sure we know that yet. Maybe we work that way when the data

02:13:51.320 can justify it. But for the moment, there's a huge car culture that's in operation here as well

02:13:55.840 that clearly is affecting our choices and willingness to give certain things up for these

02:14:00.920 other societal benefits that we know we could attain. Yeah. Are there any resources, Mark,

02:14:06.440 that you would point people to who are kind of interested in the, what can I do? There are,

02:14:11.780 again, lots of things that hopefully will happen technologically and we can debate the policies all

02:14:16.840 day long. But at the end of the day, everybody listening to this is going to get in their car

02:14:20.340 today. And I hope we've given people a lot of strategies. Is there anything you would point

02:14:25.820 people to, especially parents maybe who have teenage kids who are starting to drive where we can,

02:14:33.740 as you said, I think get them started in the best way possible because you start to create

02:14:38.820 patterns of behavior early. So let's hold the kids for a moment. And I think the place to start a

02:14:45.560 few websites, NHTSA.gov, NHTSA.gov is a great website that you can go on. It has a VIN, vehicle

02:14:54.820 identification number lookup. So if you're wondering, do you have a defect and recall that you should

02:14:58.640 take care of, you can do that. Car seats, you can literally go on there and put your kid's birth

02:15:03.160 date, height, weight, and it'll actually give you recommendations of whether it's backwards,

02:15:07.600 forwards, booster seat, et cetera. Really good with that. And all the data we're talking about that I

02:15:13.540 can't keep in my head is on that website. And so it's a great source for things. We haven't talked

02:15:18.780 about it, but NHTSA runs a new car assessment program, NCAP. It's the five stars, stars for cars.

02:15:26.300 It's what's on the Monroly label on your new car. That's the stars evaluating its crash worthiness and

02:15:31.940 some other factors. What's it called? It's NCAP new car assessment program. Okay. So again,

02:15:38.160 when you're thinking about where should I start? Well, to start with the ratings. So NHTSA has got

02:15:42.580 all this different stuff. That's a great resource there. Next would be the Insurance Institute for

02:15:47.820 Highway Safety, IIHS. They do a lot of the crash testing, but they also have a lot of research that

02:15:53.420 they do. So they're the ones who do the top safety pick and other kinds of things. So that AEB challenge

02:16:00.000 I talked about, that was actually collaboration between NHTSA and IIHS doing that challenge for the

02:16:04.980 industry to democratize AEB. And besides the crash information, they also have other studies and

02:16:10.280 things they do. So really good there. National Safety Council, NSC. They have great data because

02:16:17.660 they all have car stuff. They have pedestrian stuff. They have cyclist information. They're really good

02:16:23.280 for general resources for safety, including roadway pieces. Sorry, that one again was?

02:16:28.940 That's the National Safety Council, NSC. NSC. Okay.

02:16:31.900 And the fourth one I would mention is Safe Kids Worldwide. And they actually certify car seat

02:16:39.780 technicians because it ends up, and I'm going to get this number wrong, but 60, 70% of car seats are

02:16:46.740 installed incorrectly. I know it's over 50%. So it's great to do all that work to get the right one,

02:16:51.940 but then if it's not incorrectly. So Safe Kids Worldwide actually does that too. I think those four

02:16:57.120 are great resources for different things. When you get to the teenagers, a couple things. AAA

02:17:03.580 actually has different kind of driver contracts that you can use. So that's the insurance company

02:17:09.580 also has a foundation. So they have a lot of good auto safety information there as well. But they

02:17:14.640 actually have contracts that you can do with your kids. And I would just say you should also look up,

02:17:19.380 because depending on the state, they may have graduated licensing. And that licensing for new

02:17:24.420 drivers actually includes, can you have other kids in the car? No. Can you drive after dark? No.

02:17:31.440 This for six months, this for a year, et cetera. And so I was just talking to a friend. It's like,

02:17:36.600 oh, we did the AAA. I'm like, you should look at the graduated licensing along with the AAA and just

02:17:42.020 see what the elements are and then come up with your own. And I hate to say this because as a scientist,

02:17:46.980 you'll appreciate this too, but feel free to experiment with her. Put that training program you were

02:17:53.400 talking about. There's some simulator work. There's also some actual driving work. There's

02:17:58.040 the intellectual part that would supplement any formal stuff she gets from another source.

02:18:02.760 But do that. I mean, that's the intervention. I would do your own assessment then to sort of,

02:18:07.060 could you measure some of that in the simulator, for example, to see if now you have an end of one,

02:18:11.160 totally get it, but it's her life you're trying to save. So whatever you could do there to try and

02:18:16.680 bolster her with more education, experience, et cetera, that would make her again,

02:18:21.940 more situationally aware and defensive when she's driving along with the skillset, I think would 0.73

02:18:25.980 be great benefit. The other thing I'll just tell you that's interesting is we've seen in the last

02:18:30.440 five to 10 years, young drivers are less and less interested in getting their driver's licenses right

02:18:35.880 away. So we're seeing kids in college that don't know how to drive, you know, or they're waiting till

02:18:41.160 after college when they have a job or they're picking their location by public transit.

02:18:45.560 And this might be a great trend given what you said about the maturation of kids anyway. So yeah,

02:18:52.020 you know, one thing I'll just share with you with my daughter, she's not psyched about this, but

02:18:55.600 she's going to be driving a manual transmission truck. That's actually going to be her car. And 1.00

02:19:00.780 every time she fusses about it and says, Oh, you know, I just say, I mean, you could take the bus,

02:19:06.480 like that's cool too. But if you want to drive, you're going to be driving stick. And there's a big part

02:19:11.440 of that is the connection to the drive. And it reduces at least one of those variables in my

02:19:16.540 mind, which is the distractibility. And she's nervous about rolling backwards on hills. And I'm

02:19:20.640 like, good, concentrate harder. Yep. And learn the skill. Yeah. So that becomes a non-issue for you.

02:19:26.520 Yeah. Well, Mark, this has been really illuminating. It's quite disturbing as we've talked about. And I

02:19:32.640 almost wish there was a way that we could convey every one of the 42,929 stories of the people who

02:19:41.860 died two years ago, the most recent data we have here. I think it, on some level, it sometimes takes

02:19:47.340 those things. Unlike when you think of all the chronic diseases that I spend most of my time

02:19:51.120 thinking about and talking about, they disproportionately affect older people. Your risk of death from

02:19:57.600 cardiovascular disease, stroke, cancer, and Alzheimer's disease. This is disproportionately a disease of the

02:20:04.760 elderly. Whereas when I last looked at this, which was 2020, and I can't imagine the data have changed

02:20:11.040 that much. The thing that stuck out to me the most was that automotive deaths were the most uniform cause

02:20:19.720 of death by decade of anything in the top 10 causes of death. There was nothing more uniform

02:20:27.580 than dying in a car. Equal opportunity killer. Which is why let me, I know as we start wrapping

02:20:35.240 just two things that I think would be good, especially for you to be thinking about. One is

02:20:40.620 that, and these are where the parallels are there. One of the differences, however, is between health

02:20:46.120 and say the safety element, et cetera, is, and one of the reasons I've really looked at safety and

02:20:51.660 transportation from my sleep interests is death can be instantaneous. I mean, it's like in milliseconds and

02:20:59.080 you don't go out that day thinking it's going to happen. And as you were saying. Yeah, you don't get

02:21:02.160 to say goodbye to your grandmother who's got cancer for two years. Yeah. Someone comes to your door.

02:21:08.360 They're not coming home. It's like, just like that, right? And so I think that's one thing that

02:21:13.880 separates it in some ways from those chronic illnesses and many other kinds of things. But I would also say

02:21:20.180 there's some real significant parallels when I think about this, because one of the big pushes

02:21:25.020 that I've tried to make happen in the automotive realm is trying to move from a more reactive

02:21:30.700 safety culture to a more proactive one. Which again, I think the parallel there of what you do for

02:21:36.880 prevention and thinking about how do we eliminate or mitigate the things we know that can cause bad

02:21:42.540 stuff, promote the ones we know that will make good stuff happen. How do we do that? And I think

02:21:48.700 that's been a challenge having worked at NASA as a NASA scientist on the aviation side. They're very

02:21:53.180 proactive. I mean, in aviation, they went 12 years. In the US, no person died in a commercial aircraft

02:22:00.640 crash in like 12 years. Okay. And then it was one. And I always tell people, it's like, it's so hard to

02:22:06.560 get there. It's even harder to stay there, saving all those lives. But that takes a proactive culture

02:22:13.120 that says we're going to do what we can to eliminate or mitigate, diminish what those risks are and

02:22:18.840 promote the good stuff we know that's going to make a difference. You do that all the time. This

02:22:22.840 is why I said thank you at the beginning, because bringing this into that realm, and not just public

02:22:28.240 health, this is like societal safety and what we get here. There's a real parallel there that there's

02:22:33.080 a chance to be more proactive. Now you think about it, investigation is reactive. Bad thing happens,

02:22:38.620 we investigate, but it's to make it proactive. That what we learn from there has to be translated

02:22:43.880 into some action that's going to prevent it from happening again. And that's where, frankly, and

02:22:49.200 always, again, I've already said thank you, I'll do it again. Just having people discuss this and think

02:22:53.460 about it, tell their story to someone else, those can save lives. I don't know your daughter at all.

02:22:59.560 I've known you for this period of time, but it's like, you're way better off, both the fact you've

02:23:03.960 talked about it, you've thought about it, and will basically provide that context and skill set to

02:23:09.420 your daughter. You may not even be around when it saves her life, or her kid's life, or some other

02:23:15.800 person in her sphere, because that's the societal change in a proactive way that's going to make the 0.77

02:23:21.560 big difference. And yeah, I think we need to save them one at a time. Your friend, my father,

02:23:26.140 the consequences are so significant, but we can't bring those back. How do we use those

02:23:30.880 as opportunities, though, to make the future for you, your daughter, her family, my kids, etc.?

02:23:37.100 What do we do to make that in a proactive way safer for the future? You're doing that straight

02:23:41.600 on with the healthcare. The parallel is there in this realm as well, because the costs are so high.

02:23:47.600 If there's one thing I could do based on this discussion, Mark, because I now realize there's so

02:23:51.680 many pieces that are changing with the technology that are going to move us in the right direction.

02:23:55.500 But if I could make a change today, based on what we discussed, it would be that I wish that for

02:24:01.520 every significant accident or fatality that occurred in a given city, the story was told

02:24:08.000 in two fronts. The story was told in the human sense of the story, so that we understood the life

02:24:15.620 or lives that were lost, and the consequences, and how that's going to ripple through forever.

02:24:19.880 How Nick's wife lost her husband and Nick's kids lost their dad, but also in a very clinical

02:24:26.340 autopsy-like manner of the accident. I really think that every time a horrible accident or fatality

02:24:33.760 occurred, if each of us could see a 60-second video that would say, this is what happened on

02:24:40.600 this date and this time, and these were the contributing factors, that's it. And everybody

02:24:45.500 has to invest. It's not a big price to pay for those of us who are alive, that for every time

02:24:50.300 somebody dies, you got to invest a minute in hearing how it happened. No other technology at

02:24:55.680 this point in time, just the explanation of what happened, is going to make us better drivers.

02:25:02.260 No question. This is why I was not kidding about write those down, the four-minute video or like what

02:25:07.220 you're talking about now. And I would just slightly extend to say, not just identifying the causal

02:25:13.120 contributing factors, but if there's any action that people could actually take to do differently,

02:25:18.000 and think about this with that local intersection you keep talking about, without what you just

02:25:21.940 described, it's going to keep happening. People aren't going to know what happened there, or if

02:25:26.000 there's anything different they could do to spare their lives, it's not going to change without what

02:25:31.200 you were just describing. Yeah. Yeah. Figure out how to do that. Mark, thank you very much. Really

02:25:37.320 appreciate your time and your insight today. Thank you. Enjoy the conversation.

02:25:41.020 Thank you for listening to this week's episode of The Drive. It's extremely important to me to

02:25:46.480 provide all of this content without relying on paid ads. To do this, our work is made entirely

02:25:51.460 possible by our members. And in return, we offer exclusive member-only content and benefits above

02:25:57.880 and beyond what is available for free. So if you want to take your knowledge of this space to the next

02:26:02.420 level, it's our goal to ensure members get back much more than the price of the subscription.

02:26:06.920 Premium membership includes several benefits. First, comprehensive podcast show notes that detail

02:26:13.700 every topic, paper, person, and thing that we discuss in each episode. And the word on the street

02:26:19.320 is nobody's show notes rival ours. Second, monthly ask me anything or AMA episodes. These episodes are

02:26:27.420 comprised of detailed responses to subscriber questions, typically focused on a single topic

02:26:32.280 and are designed to offer a great deal of clarity and detail on topics of special interest to our

02:26:37.640 members. You'll also get access to the show notes for these episodes, of course. Third, delivery of

02:26:43.380 our premium newsletter, which is put together by our dedicated team of research analysts. This newsletter

02:26:48.920 covers a wide range of topics related to longevity and provides much more detail than our free weekly

02:26:55.160 newsletter. Fourth, access to our private podcast feed that provides you with access to every episode,

02:27:02.040 including AMA's sans the spiel you're listening to now and in your regular podcast feed. Fifth,

02:27:08.960 the qualies, an additional member only podcast we put together that serves as a highlight reel featuring

02:27:15.160 the best excerpts from previous episodes of the drive. This is a great way to catch up on previous

02:27:20.340 episodes without having to go back and listen to each one of them. And finally, other benefits that

02:27:25.340 are added along the way. If you want to learn more and access these member only benefits, you can head

02:27:30.740 over to peteratiamd.com forward slash subscribe. You can also find me on YouTube, Instagram, and Twitter,

02:27:38.200 all with the handle peteratiamd. You can also leave us review on Apple podcasts or whatever podcast

02:27:44.780 player you use. This podcast is for general informational purposes only and does not

02:27:49.980 constitute the practice of medicine, nursing, or other professional healthcare services, including

02:27:54.360 the giving of medical advice. No doctor patient relationship is formed. The use of this information

02:28:00.460 and the materials linked to this podcast is at the user's own risk. The content on this podcast is not

02:28:06.860 intended to be a substitute for professional medical advice, diagnosis, or treatment. Users should not

02:28:12.420 disregard or delay in obtaining medical advice from any medical condition they have, and they should

02:28:17.160 seek the assistance of their healthcare professionals for any such conditions. Finally, I take all conflicts

02:28:23.500 of interest very seriously. For all of my disclosures and the companies I invest in or advise, please visit

02:28:29.840 peteratiamd.com forward slash about where I keep an up-to-date and active list of all disclosures.

02:28:42.420 Thank you.

The Peter Attia Drive - March 25, 2024

#295 ‒ Roadway death and injury： why everyone should care and what you can do to reduce risk ｜ Mark Rosekind, Ph.D.

Episode Stats

Summary

Transcript