“Safety isn’t expensive, it’s priceless.”
In the Autonomous Vehicles field, safety is at the front, back and center. In fact, all the other features that are included only make sense if a vehicle is proven to be safe.
That’s why AV companies are investing heavily in proving that their vehicles are safe. The fact that they have to meet higher safety standards has made them rethink their approach to it.
In this episode, the second part out of two, Ed Bernardon interviews Mark R. Rosekind, Chief Safety Innovation Officer at Zoox, Inc. He’ll help us understand the safety features that are included in their vehicle. He’ll also share with us how his vast experience in transport has helped him in his current role.
Some Questions I Ask:
- What should you do to regain alertness if you feel sleepy while driving? (06:34)
- What are the key areas that you consider when you arrive at a crash scene? (10:41)
- What exactly are you doing to make the stopping distance shorter? (22:51)
- How do you make the vehicle safe in case a system failure occurs? 24:29)
- What needs to be done to our current infrastructure to make AVs safer? (29:17)
What You’ll Learn in this Episode:
- What inspired Mark to pursue the field of safety (00:41)
- The difference between NHTSA and NTSB (08:56)
- The differences between how regulators and AV companies approach safety (12:14)
- The most significant safety innovations by Zoox (17:05)
- How Zoox uses simulation to test their AV (26:21)
Connect with Mark R. Rosekind:
Connect with Ed Bernardon:
- Future Car: Driving a Lifestyle Revolution
- Motorsports is speeding the way to safer urban mobility
- Siemens Digital Industries Software
Ed Bernardon: Have you ever been a passenger in a car or bus and almost feared for your life? What if those uncomfortable rides were a thing of the past and we could count on safely and efficiently getting from point A to point B? When it comes to ensuring safety in a vehicle —either human or autonomous driven or really anything else — it’s always better to be proactive rather than reactive. This is the Zoox approach for autonomous vehicle safety: take action to prevent situations where accidents may occur well before the accident is even about to happen!
Today, I continue my conversation with Mark Rosekind, Chief Safety Innovation Officer at Zoox and a former member of the National Transportation Safety Board. We discuss the proactive safety measures behind the Zoox AVs, including Mark’s deeply personal inspiration behind promoting vehicle safety. I also talk to Mark about what to do if you ever feel yourself starting to doze off behind the wheel, the one thing he’d like to “uninvent.” Tune in and hear part two of our conversation, on this episode of The Future Car podcast.
Mark, welcome back to the Future Car podcast.
Ed Bernardon: I’d like to take a step back and talk a little bit about your background and your inspiration. I think your father was a motorcycle policeman and was in traffic enforcement. I believe that was a big inspiration for you early on.
Mark Rosekind: I don’t talk about this a lot. Actually, it was at my Senate confirmation hearing for being NHTSA Administrator that I actually talked about this publicly, but my father was a motorcycle policeman in San Francisco and he was chasing a traffic violator when somebody ran a red and hit him. So, he was killed in the line of duty – I was three and a half, my brother was two. And so we were one of those headlines: “Officer killed in the line of duty, leaves a widow with two young kids.” And so, it’s one of the reasons I know 38,680 lives lost in 2020. For some, that’s just a number or statistics. For me, every one of those numbers is a father or a mother, sister, brother, a community member, one of your colleagues — those are people and we can’t bring them back. We can talk about all this stuff for safety, but we cannot bring those lives back to us. And so, yeah, this is personal for me, I don’t lead with that. But when people want to talk about these issues about safety, and choices we make, this is personal – very, very personal for me. And so, whether it’s a law enforcement side, people lose significant members of their family and friends, it affects your entire life.
Ed Bernardon: And I think, too, that, as we’re saying, safety is a key thing, especially with autonomous vehicles. And it’s one thing to say, “Hey, we gotta make sure we also make it safe, besides all the AI and everything.” But it’s really another level when the need for safety is a passion that’s always been with you, I would imagine, throughout your life, and certainly, your career shows that.
Mark Rosekind: You’re on target there, Ed, which is why, for me, safety is always foundational to everything. Safety first – not as a priority because priorities change, but safety is foundational. It is what you build everything on top of. And you’re right. For me, when I look across my career – and I’m an old guy now, been around for some decades doing this stuff – safety is always what drives me; what’s it gonna take to save more lives?
Ed Bernardon: And I think when it comes to autonomous vehicles, one might say, “Oh, we want autonomous vehicles because they’re going to be more cost-effective, or they’re going to be more sustainable.” I mean, you can come up with a long list of things. But no matter what, none of those things really mean anything unless it’s safe. And being safe, I think, is a foregone conclusion – you’ve got to prove that it’s safe before you even start talking about the rest. I want to ask you about the earlier parts of your career. You were in an area that had a lot to do with sleep research and alertness. How did you get into that? How did you go from that to transportation? Obviously, they’re related because there’s probably one of the biggest things that hurt safety is when people fall asleep or aren’t alert while they’re driving. But how did that interest come about?
Mark Rosekind: I’ll keep it short, but I’m an undergraduate at Stanford, pre-med, and I take a course called “sleep and dreams.” And the one thing everybody knows is REM sleep, Rapid Eye Movement sleep. Well, the course is being taught by the guy who helped discover that. And it was just fascinating because everybody sleeps and it affects everyone. So, that’s how I got into Sleep and actually decided to pursue a Ph.D. so I could do research. I ended up working with Dr. Dement – the guy who helped discover – for a bunch of years. And ended up getting my Ph.D., doing a postdoc with somebody from Stanford who was at Brown, doing a bunch of sleep research. When I came back, I worked for Dr. Dement again for some years. And then I started working with NASA. And to your point, the NASA work was focused on pilots and astronauts and how do you keep them awake on the job. So, that was my transition to transportation. So, in the ‘70s, for about seven years, I ran a program called the Fatigue Countermeasures Program at NASA; keeping pilots and astronauts awake. So, we were highly involved in aviation: pilots, military, commercial, business, air traffic control – you name it, everyone’s got to be alert on the job to keep things safe. And eventually, at NASA, it became clear it wasn’t just aviation; all modes of transportation had fatigue issues. And so that started in the ‘70s, but by the mid-‘90s, I was pretty much involved in human fatigue issues across all modes of transportation. And so, for me, it’s actually been about 25 years I’ve been in the transportation industry. And the thing I’ll just mention, what’s interesting is so many people are familiar with my work in Washington, at the NTSB or NHTSA. I had a long, multi-decade career doing sleep work for a lot of years before that, and really was my first intellectual passion, which is fun, but you already made the connection. Clearly, that’s a key to safety in any mode of transportation.
Ed Bernardon: Well, along those lines then. So, you’re an alertness expert of some kind, certainly the greatest alertness expert we’ve ever had on The Future Car podcast. So, I want some advice for our listeners. So, certainly, you want to be proactive, but let’s say you’re in a reactive mode, you’re driving, you catch yourself starting to do that head nod, little bit dozing off – is it better to roll the window down and have cold air on your face, or maybe slap your face a little bit maybe, or turn the radio, or pullover is probably the best thing to do? What do you recommend there to get yourself alert again?
Mark Rosekind: So, Ed, we are about to save some lives. Are you ready?
Ed Bernardon: I’m ready.
Mark Rosekind: Great question. We’re going to save some lives with this. And you’re right, this is my area of expertise. I’m the Lifetime Achievement winner from the National Sleep Foundation, I have over 150 publications. This is stuff I do; keeping those pilots and astronauts awake. So, what do you do as a driver? First of all, when you get the head nod, pull over. One of the studies I did — I have one right back here — was giving pilots naps in the cockpit. Why did we do that? Because they were falling asleep, two or three at a time, including when landing a 747. So, if your brain is sleepy enough, it will fall asleep even in a life-threatening situation. So, if you’re in your car and you’re getting a head nod, it’s too late, you’ve got to pull over. And don’t just pull over, you want to pull over, and if you can, take a short nap, because only the sleepiness is going to help reverse that. So, to your question, brilliant, because that’s what people say, “Can I roll the window down? Can I slap myself? Can I turn the radio up? Can I turn the lights on?” Those things work for 10 minutes. So, they’ll work just long enough to get you to a safe place to pull over. But if you do those thinking you’re going to make the next hour, half-hour, even 15 minutes of your ride, that could be when the crash occurs that could take your life.
Ed Bernardon: That’s great advice. And I think, too, nobody wants their face slapped every 10 minutes.
Mark Rosekind: Yeah. And it doesn’t work after a while. That’s my point, is your brain will shut you down. And we’re lucky, often, because there’s nothing that’s life-threatening. But if in that moment, a car comes our way or there’s a pole in the way, that’s where the crash and your life gets threatened.
Ed Bernardon: So, you took a move into government, you were in the Obama administration, head of NHTSA and then there’s the National Transportation Safety Board, which I think was before that. What is the National Transportation Safety Board? What is the National Highway Traffic Safety Administration? What’s the difference? What are their roles? You’re in both, maybe you could tell us a little bit about what they do for us?
Mark Rosekind: You bet. And you know what, Ed? It’s so good that you’re asking about that because especially in the auto industry, people are very familiar with NHTSA, National Highway Traffic Safety Administration because it’s the modal agency within the Department of Transportation that has responsibility for auto safety regulation enforcement. So, that’s what NHTSA does for the whole country. And in the past, it was barely 650 people and not even a billion-dollar budget responsible for all safety regulation enforcement. The NTSB, the National Transportation Safety Board. So, at NHTSA, I was Administrator; at the NTSB, I was one of the five board members. And the NTSB, basically, investigates major crashes in all modes to make recommendations so those crashes don’t happen again. And so, what’s critical there is the NTSB does the investigation part making recommendations, but it doesn’t really have any power to enact any of those recommendations. It’s up to companies, governments, state, local, federal to actually make the changes to bring those recommendations to fruition. So, NTSB’s investigations with recommendations; NHTSA, actually, is the one that implements the safety programs, does the regulation, does the enforcement. They’re very complementary in that way.
Ed Bernardon: So, one of the roles of the NTSB, when you were there, is to try and understand why your crash occurred. And you’ve said that crashes are complex. So, what are the key areas that you consider when you arrive at a crash scene? I mean, you’re sitting there, you’ve got these cars, it could be pretty mangled. Is there a systematic way that you try and figure out what’s going on?
Mark Rosekind: Absolutely. And the NTSB is superb at it. And I don’t mind saying, even as a board member, did I learn from those professional investigators who are superb. So, when you show up on scene, you’re looking at three things: the human, the machine, and the environment. And you go through each one of those areas. So, the human is everything — Let’s say, it’s an, again, automobile crash. You’re looking at the driver – did they have their license? Were they trained to do what they were supposed to do? Were they drinking? Were they sleep-deprived? You look at the human. You look at the environment – so, that’s gonna be everything from weather to the road, etc – and you look at the machine. In this case, it’s the car, for example. Was there a defect? Were the brakes working? So, the human, the machine, the environment, those three things. An in-depth examination and analysis of each one of those areas, which the NTSB usually comes up with a finding, a probable cause, but also contributing factors. The complex point is that is never one thing, there’s always multiple things in a chain of events that create a crash.
Ed Bernardon: You’re somewhat in a unique position because of your experience in government, and now in one of the leading companies that’s creating autonomous vehicles and a service-based on autonomous vehicles. How do you compare the perspective of government regulators versus people on the commercial side of things when it comes to safety? How do they look at it differently? How do they look at it the same?
Mark Rosekind: So, I think, to your point, the ones who keep safety literally in the bull’s eye as the primary target, look at it the same. It’s about saving lives, preventing injuries, preventing crashes, etc. But to your point is, what path do you take to get there? And so government investigators or regulators are going to look at the levers or tools they have available to make changes that will save those lives and prevent crashes. Whereas in a company, it can’t control the whole system, it’s got to be thinking, “What product do we create that’s going to help us save lives or prevent those injuries?” So, they may have the same central objective, but they will come at it from different kinds of angles. People say, right now with autonomous vehicles, that the regulators are always going to be behind because the technology’s changing so quickly. I always point out, the government doesn’t actually build stuff, that’s up to car manufacturers, plane manufacturers, iPhone manufacturers, technology, and other companies that build things. The regulators are always going to be late, if you will, because they come after when the new technology has been innovated. And then the job is to make sure the government regulators can come – hopefully early enough – to make sure the guardrails are around so that new technology is safe. And I always say this, which is, even now, with all the talk about AVs, great potential, we still have to prove it, we still have to prove that it will give us the benefits that we talk about.
Ed Bernardon: From the commercial side, having been on the government side, if you could give some advice to the people in government right now, on something — let’s just call it the top thing that they could do sooner rather than later, to improve the environment to get autonomous vehicles out there and operating? What would you recommend to the federal government to do?
Mark Rosekind: Can I do two?
Ed Bernardon: I’ll let you go for as many as you want.
Mark Rosekind: Well, I think, actually, one of the first things that government can do, which is a wonderful role they can play, is to bring the different parties together so the industry can get together and say, “We want to do these safety things.” But sometimes, as we discuss their competitive concerns about intellectual property or whatever, they don’t always talk to each other. If the government says, “I’m getting you all in the room because I have this safety objective, I want to know how we’re gonna get there”, they can be a great motivator to bring those people, not just in the room, but to problem-solve. So, that’s the first thing. I think NHTSA and DOT can be a great forum to bring groups together. And that’s why the second thing, just to give an example of that, is, there’s a huge problem – I’ve been bringing this up a few places but there’s a huge problem right now with all the different language that people use to describe all this new technology: ADAS, supportive systems, ADS (Automated Driving System); what do all these mean? AEB versus ACC, and I’m using these acronyms for a reason, which is, people don’t necessarily know what they mean, and what does it mean, actually, in my vehicle. So, again, a great example of what government could do is get people together and say, “Okay, let’s nail the language down. And going forward, this is what we’re going to call X, and Y, and Z.” So, when we were talking about communication earlier, the same issue is how do we make sure consumers or the road users have a consistent language to understand about this new technology? That’s just one example.
Ed Bernardon: If the government brought you all together, you’re all in the same room, you’re the leader of the discussion, what would be the first topic you would want to discuss? What’s the top topic that you’d love to discuss and would hope that you could come to some resolution on or to some benefit?
Mark Rosekind: The question you asked: How safe is safe enough? So, it couldn’t get it figured out for the last 100 years, how do we figure out how safe is safe enough for moving forward? We can say zero, that’s our objective; how do we get there?
Ed Bernardon: And what better time than now when everyone is innovating and coming up with new ways to do transportation unlike ever before it? This is your chance, our chance to do it.
Mark Rosekind: Exactly. So, just to go back to the beginning, my title is actually Chief Safety Innovation Officer. And I said that because I have pointed out since being an administrator, we need as much innovation in the safety arena as we do in all the technology arenas. That’s how we move to get to zero.
Ed Bernardon: Well, let’s talk about some of those safety innovations because the claim is that you have over 100 safety innovations. So, by innovations, it’s things that don’t exist already. So, tell us about your 100 safety innovations – the most significant, the ones you’re most proud of.
Mark Rosekind: So, we just put out a second safety report. And it’s just zoox.com/safety. Everybody submitted them to NHTSA. We did our first one on our philosophy. The second one is where you got that statistic that we’re talking about, where we identified – you’re right – 100 safety innovations that don’t exist on cars that are on the road today. In that report, we highlight nine of them. They fall into three areas, one had to do with driving controls. So, it gets to our four-wheel steering, braking, and why those matter. It actually talks about the accuracy of our steering, even at speed around corners; braking, which should be obvious, because it means you can stop shorter, react more quickly, all that kind of stuff. The second category mostly had to do with what we call our safety objective of no single point of failure. That means designing systems at an aviation level of safety – so you have redundancies to make sure that if one system goes, there’s a backup and another backup. And so we have three examples of that, including monitors that monitor all these different systems in the vehicle. So, again, we’re bi-directional. If something happens that cuts off some capability in the front, won’t be bi-directional. Let’s just go in the other direction, bypass those until those things can get fixed. And the third area is innovations in traditional arenas. So, for example, seatbelts and airbags. So, in a Zoox vehicle, when you get in, everyone has to be buckled up before it will move. 91% of people wear their seatbelts, but in crashes where people die, 50% of them were not wearing their seatbelts. A Zoox vehicle won’t start moving until you’re buckled up – so, that’s built-in. And the airbag system, I hope people go to the report, but it’s this beautiful curtain airbag that comes down. Again, we have a clean sheet, what can we do? Instead of in a steering wheel, it’s got to come out at you. Well, this is a curtain that comes down, there’s a secondary one for each passenger that can come out, and then they’re everywhere else – side, back, etc. And you’ve got to read about it because it’s so cool. But we actually have an airbag module that can determine where that impact is coming from and potential speed, so it can actually determine what airbags need to release when based on what crash is about to occur. So, that gets to our prevention that we were talking about, as opposed to reacting after things actually happen.
Ed Bernardon: We hear front airbags, side airbags, rear airbags; you also have a horseshoe airbag. And I think you’re right, that image you have on your site, you’re completely enveloped in airbags, so you’re covered in every direction. One thing also is a little bit tricky when it comes to safety, and I’m sure you’ve thought about this, is, in conventional vehicles we’re facing forward. Now, suddenly, you have two people facing back. And in fact, it says it can go in either direction, they’re both front and back. And keeping someone safe when the seat is behind you, or I guess in front of you, however you want to look at it, it’s a little trickier. How do you handle that?
Mark Rosekind: Yeah, good for you. It does make it more complex. We’re already doing crash testing. And I always tell people, we want to meet or exceed what the requirements are. And so, to your point, most of that is built for front rows front-facing, we will meet those. But again, you’re completely correct as we actually have to think about that from our carriage seating perspective. And so our testing, our crash testing, and all the other things we do; we are trying, as we’ve discussed, to provide the same level of safety. In NCAP terms, that would be five-star crash-worthy test, crash-level protection in every seat in our vehicle. We alluded to this briefly. But the way the current crash systems are set, they have higher criteria for the front seat, sometimes no criteria, or lower criteria for the back seat. For us, because of what you just described, we have the same high level intending to get five-star crashworthiness for every seat in the vehicle.
Ed Bernardon: And that’s self-imposed, or is there a government regulation that says, “Hey, no matter how many seats you have and what direction they’re facing, you must be safe to some level.”? I mean, I would never have thought of that.
Mark Rosekind: And it doesn’t exist.
Ed Bernardon: Yeah, you go from having all the seats facing forward. And like you just said, sometimes there isn’t even a regulation for what happens in the backseat. And now, these configurations, who knows what they are. I mean, that must make the job for the safety regulators very complex. Fortunately, you are doing some self-regulation.
Mark Rosekind: So, one of my favorite mantras at Zoox is: “Set the bar.” These brilliant young engineers come in with a problem they’re trying to solve. And in some cases, for traditional automaking, you can literally go to the book on the shelf, pull it out and say, “Oh, this is how you solve that.” With a lot of what we’re doing, it’s never been done before. And so that’s why my mantra is: “Set the bar for safety where you think it needs to be.” So, that’s the example you just gave. There is no regulation that says, “If you have this kind of seating, here’s the level of protection you should provide.” So, we’ll set the bar for what we think is, “You know what? Every seat should have the same five-star level of safety for everyone.” Hopefully, people will emulate or follow us for where that bar gets set.
Ed Bernardon: Sometimes with autonomous vehicles, we don’t think about the more traditional aspects of that vehicle. And you actually touched on them, at least mentioned them a few minutes ago: four-wheel steering and shorter stopping distance. I mean, those are things that you could put in a traditional car. So, the four-wheel steering, maybe you could explain that a little bit for our listeners, but it makes the car more agile, the ability to steer more quickly or to tighter spots. And also, what exactly are you doing to make the stopping distance shorter?
Mark Rosekind: I hope people go to the report because it gives really nice diagrams of all that stuff. And I think the example I would give is you’re driving in a city, where there are cars parked on the right, and you see a bunch of kids over there playing, and now you’re worried about that ball coming into the street. So, that’s the model. The four-wheel steering gives you the accuracy to know, “You know what? We can just move six inches, a foot further left into a lane that gives us more time to react, if that ball should come.” The braking system, same kind of thing, is that being able to stop in a shorter distance means if that comes, not just a quick one, but seeing it with our perception systems, there are actually kids over in that area, gives us a chance to slow down – so we’re actually breaking from a slower speed than having to break at, say, a higher speed. Again, you can read the specifics of the technology that’s in there. But it’s thinking of the scenario where you would use both the steering and the braking to make a difference on what kind of safety you could be providing for that kid that might dart out, that all of this fear, could dart out in front of our vehicle.
Ed Bernardon: You’re giving the vehicle better tools to be able to deal with situations that you normally might not have to if you didn’t have those two things. You also mentioned no single point of failure. Your battery system dies, your computer dies, who knows what; you want to make sure that the vehicle, just like a real driver — the driver that had fatigue, we were talking about before — you’d better pull over before you get too sleepy. How do you make it so that if there are system failures of some kind that the autonomous car can pull over safely?
Mark Rosekind: So, I’ll just give you a couple of quick examples. One is, since you mentioned the battery, we actually have what will be one of the largest batteries on the road. But it’s actually in two forms. There are two batteries; one on the one side of the vehicle, one on the other. And so if one actually goes out, the other one can get you to your destination and then back for maintenance. That should be invisible to you, even if it goes out; that’s the redundancy part. And then I just alluded to the monitoring system that monitor all the different sensors, the same kind of thing. Let’s say, your lights going in one direction fail for some reason; you would not only be able to sense that, but you would be able to literally decide which direction should the vehicle go in so it can deliver you to your destination, invisible to you that there was even a system issue, and then go back for maintenance so it gets fixed before it goes out again. And so whether it’s battery or light systems or braking systems, there are redundancies – anywhere from one to three or more deep – that allow, basically, compensation without even letting the rider know about it. So, you get to your destination without even knowing it, and then it’ll get fixed before the next riders get in.
Ed Bernardon: Humans have to take a driving test to make sure that they’re capable. Your code, the AI engine, the brain, you’re using a lot of simulation to test things before they go on the road. Can you tell us a little bit about that? What types of simulations do you use? How do you use them? How do you know when you’ve simulated enough and it’s time to put it on the road?
Mark Rosekind: Well, why that’s so important is because people talk about public testing. And I always point out, there’s a long path before you ever get to public testing. We have laboratory setups that do simulation, and mechanical, we’ve written about it in some things recently. But we have something called DynoBot – it’s one of our videos we put out, etc – where we have all the mechanical systems we get to test, but in a lab. So, we’re creating all these new things, you don’t want to test those on a vehicle that’s on even private roads. So, we have all kinds of laboratory testing and simulation, etc. And for the simulation part, the beauty of that is: Give me an interesting scenario. In simulation, I can do hundreds, thousands of variations on that, that could take me billions of miles to actually experience. And so that’s the beauty of simulation – you can do all that. And by the way, you then go from the laboratory, to private tracks, to private roads, eventually to public roads, and eventually to public servants. The point that’s really critical there, is it’s not like people just say, “Oh, that’s interesting. Let’s go test it on public roads where there are people. Here’s all this other stuff you do.” And I think what you’re highlighting there is there’s a lot of stuff that never makes it out of simulation or the laboratory that never gets onto the vehicle, because it’s like, “Well, that doesn’t work the way we intended. Let’s figure out how to do it right.”
Ed Bernardon: When it comes to simulation, are the simulations homegrown, proprietary, built in-house? Are they tools that are publicly available, and really you just add data? What actually do you use when you do simulation?
Mark Rosekind: We create our own. We actually have a lot of folks from the gaming industry. So, the 3D richness of these environments is just amazing. And so we create our own. This is the part of having it tightly integrated, so it reflects our experience, as well as the other kinds of things. As you know, some people do offer simulation capabilities to different companies, etc. We grow our own. We also do our own maps – that’s, again, our integrated system that we think gives us some advantages for different things. And so, again, personally, I think in the future, simulation will become the predominant testing environment. You have to make sure that the simulation can be validated by your on-road testing. But at some point, simulation gives you so much more capability than, again, the hundreds of millions or billions of miles you could drive, simulation is the place to be if you can do it. So, I think some balance, when we figure out, “Well, this many millions of simulations are equal to X number of miles,” we’ll be able to get that equation that allows us to use simulation even more.
Ed Bernardon: Another part of safety is the infrastructure within which these vehicles are going to operate. So, for human drivers, there are traffic signs. And now, of course, there are connected vehicles, could have human drivers in it. And connected vehicle technology certainly applies to autonomous vehicles as well. You’re in three different cities, they probably have a different infrastructure. What can you tell us about the level of infrastructure today, and what you really think needs to be done to improve that infrastructure to make autonomous vehicles safer?
Mark Rosekind: What’s interesting about this issue is that Zoox has the AI built into the vehicle, and we have mapped the area we’re going to be driving in. And so any city that maintains its infrastructure, which means street lines are painted, it’s got stop signs that are visible, or that your flashing lights really flash the right color. If it just maintains that, then we’re good. And we’ve always pointed out that the better your infrastructure is, then the more accurate our map is, and the more accurate our perception in the moment is going to be, the safer and better it’s going to be able to drive. When you get to the connected vehicle, whether it’s vehicle to vehicle or infrastructure or pedestrians, etc, again, our vehicle is actually created to be able to operate on its own with its perception systems, etc. But when those become more commonplace, then we can incorporate those, and they’ll just help expand our envelope for operating for safety, etc. But we don’t want to have to rely on a big city investment to allow us to come in and operate in their environment.
Ed Bernardon: So, the vehicle itself has to be safe on its own with zero help from the infrastructure, that’s a given. If you have infrastructure available, then, potentially, you could navigate intersections more efficiently, or who knows what. If you could add infrastructure, what aspects of infrastructure do you think you would want to add that would help you, say, navigate more quickly or get through intersections faster? What would you ask for from a city?
Mark Rosekind: So, let’s just say the basic infrastructure is good, painted lines, etc. Two things. I think any communication system between the infrastructure, so we’ll know lights are green, yellow, red, all that kind of stuff would be great. And then the other thing would be any “pop-ups” — construction, emergency response, any of those kinds of things beyond lights and siren, that could be put out, or construction that just got created today, but it’s going to be there for three weeks — if there are ways to send information out about that, then we would be able to totally avoid that situation. Same thing with emergency vehicles.
Ed Bernardon: I want to ask you, it’s the year 2050, we’ve had all this great communication between all the suppliers of vehicles and services and all that kind of thing. What’s it going to be like? I’ll pick Boston here. I’m going to step out of a building, I’m going to look around at the sidewalk, I’m going to look around at what’s on the road – what do I see?
Mark Rosekind: Hopefully, you’re going to see a whole new mobility system that gives you even more access to whatever you need. So, if you want to take public transit, there’s some system, like a Zoox vehicle, that could literally pick you up at your door and take you to a bus station somewhere, or a rail system somewhere. Or you can come out the door and you can get your app, and basically, “You know what? I need to get out of the state and go skiing in Vermont.” And you can pick someone else who could do that for you. Or you’re coming to your door because there’s a delivery that was supposed to show up, and it’s not somebody on a bike or in a car, it’s literally a robot that’s showing up at your door for you to be able to take what you need and put in your return. Just so you know, what I’m talking about here is, everyone who’s crystal-balling, I don’t think we know. And the example I give is the iPhone. Over a dozen years ago, if someone said, “What would you pay to have a phone in your pocket? Maybe two, one for work, one for personal. And by the way, you’ll use it for texting.” “What’s that?” “Well, that’s an app.” “What’s an app?” I think the same thing is going to happen here, but with our whole mobility system. Things we can’t even imagine yet are going to be out there on the street helping you and so many others move around that we haven’t even thought of. And the number one thing I hope in 2050, if we’re not at zero fatalities, we’re getting pretty close.
Ed Bernardon: What do you think the mix of vehicles will be? Like, you’re starting to see scooters, and your little shuttle, there’s public transportation, there’s drones now and Urban Air Vehicles. What’s the mix of what I’m going to see when I step out there on that 2050 sidewalk? And it doesn’t even end at the sidewalk, probably; you were talking about little robots that interact with us, they probably walk into the building and out of the building with me, delivering packages. What do you see there, the actual devices that are going to move us and packages around?
Mark Rosekind: This is another one we could take a few hours on. This, I think, is some of the most exciting things. And we’ve been talking roadway, we didn’t even talk about drones during delivery or moving around in urban mobility, when you think about the whole system. And the reason I say that is because I think we haven’t even come up with all of those solutions. And one of the things that is a gap, though, is we’re not talking about your question enough. So, let’s just stay with roadways. Even by 2050, there’s still going to be drivers that have their hands on the steering wheel of their cars. There’ll be some people with ADAS, they’ll have different kinds of cruise control and blind spot. And then there’ll be truly self-driving vehicles. All on the same roadways. We’re not talking enough about what the next 20 or 30 years are going to look like. And to your point, what does that mean for all the scooters, and delivery, and air drones, etc? There are not enough discussions of those things going on. There have been a couple of questions – so, good for you for asking an insightful one – but there haven’t really been meetings, saying, “Okay, how are we going to get an integrated system within Boston that’s going to actually deliver on all the benefits that we keep saying are going to be available from those systems?”
Ed Bernardon: What do you think Zoox has to do? What are the next things that Zoox has to do to take steps towards that 2050 world you were just describing?
Mark Rosekind: The really exciting part is when I talk to, again, all these brilliant young engineers, the technical challenges are known and we’re going to get them, we’re going to take care of them. I think the biggest challenge, for Zoox and everyone else, is going to be actual deployment in cities; how do we figure that out? And you’re asking some of the questions. We need to make sure the grid is there, places to maintain these vehicles. Where’s the pickup and drop-off? Do you have specific lanes for certain kinds of, whether it’s bicycles or automated vehicles? These are the, literally, rubber-meets-the-road operational issues that need to be addressed to have successful operations in cities. So, I think those are the things that Zoox is starting, not just to acknowledge, but is looking for, basically, how do we answer these with the same innovation that we do with our technology to make sure we’re going to get the most benefit.
Ed Bernardon: Mark, thank you so much for joining us today. I think you’ve done a tremendous job of letting people know when you say, “Oh, my vehicle is safe,” then all the aspects and what it really takes to make that happen. Thank you so much for joining us here on The Future Car podcast.
Mark Rosekind: My pleasure. Hope we do this some more.
Ed Bernardon: What do you wish you understood better?
Mark Rosekind: Detailed engineering.
Ed Bernardon: If you could uninvent one thing, what would you uninvent?
Mark Rosekind: War.
Ed Bernardon: And if you could magically invent one thing, what would you invent?
Mark Rosekind: Anything that saved lives.
Ed Bernardon: I believe it was at Stanford, you spent time in the dream laboratory. And you were fascinated by that course you took. Tell us what is the best dream you’ve ever had. Remember our audience here now, the best dream you’ve ever had.
Mark Rosekind: Well, I will just admit that I’m a pretty sound sleeper, which means that I don’t remember my dreams very much. And that’s relevant. My wife wakes up all the time telling me her dreams – and I’m not going to give you any of hers.
Ed Bernardon: You beat me to the punch. You worked in the Obama administration. Did you ever meet President Obama?
Mark Rosekind: I went to multiple White House things and activities where he was announcing, etc. So, I never had a sit-down with him where we chatted about stuff.
Ed Bernardon: But you’ve heard him speak?
Mark Rosekind: Oh, went to the holiday parties, yes.
Ed Bernardon: I think one of the most interesting things about President Obama is he’s actually a pretty funny guy. And he could almost be a stand-up comedian sometimes when you see him out there. What do you think the funniest thing is you ever saw President Obama do?
Mark Rosekind: What I can say is that at some of the more informal holiday gatherings, there were chances to do exactly what you’re talking about, which is less scripted and more just engaging with people. And he was just superb at that.
Ed Bernardon: All right. And here’s the very last question: Tell us something about yourself that would surprise your friends and family, besides the fact that you’re a magician. Besides that, tell us something about yourself that would surprise your friends and family.
Mark Rosekind: I still get nervous whenever I do presentations or any of these kinds of interactions. In spite of decades of doing this stuff, I still get nervous every time.
Ed Bernardon: Mark, you certainly didn’t show it today, though. Not on this podcast. Thank you so much. Really appreciate your time and answering all those questions. Thank you so much.
Mark Rosekind: Thanks, Ed. Really enjoyed it.
Mark R. Rosekind, Chief Safety Innovation Officer at Zoox, Inc.
Ben is the founder of Fering Technologies. He has devoted his career to whole-vehicle design, predominantly in motorsports and supercar design and is the brains behind the Fering Pioneer. He previously worked for Ferrari and McLaren, and was involved in the development of the Caparo T1 project. He has a Bachelor of Science from City, University of London.
Ed Bernardon, Vice President Strategic Automotive Intiatives – Host
Ed is currently VP Strategic Automotive Initiatives at Siemens Digital Industries Software. Responsibilities include strategic planning in areas of design of autonomous/connected vehicles, lightweight automotive structures and interiors. He is also responsible for Future Car thought leadership including hosting the Future Car Podcast and development of cross divisional projects. Previously a founding member of VISTAGY that developed light-weight structure and automotive interior design software acquired by Siemens in 2011. Ed holds an M.S.M.E. from MIT, B.S.M.E. from Purdue, and MBA from Butler.
If you like this Podcast, you might also like:
- Zoox’s Journey to Safety with Mark R. Rosekind – Part 1
- Carlo Mondavi’s Autonomous Electric Tractors for Sustainable, Affordable Farming – Part 1
- The Next Leap for Electric Vehicles with Will Graylin, Indigo Technologies – Part 1
The Future Car Podcast
The tech-driven disruption of the auto industry cuts across domains, from silicon and software to sensors and AI to smart traffic management and mobility services. Get the chip- to city-scale story in regular interviews with technologists at Siemens and beyond.