Your mileage may vary

California is proposing a bill to alert drivers to their speeding. Anthony suggests automatic tickets. Elon doesn’t understand data or is flat out lying about the safety of Autopilot. Remote controlled trains remind us of GM Cruise. Miles per Gallon is not as helpful a measure as gallons per 100 miles and we discuss why AV companies don’t follow the Department of Transportation’s Technology Readiness Levels… which they helped write.

Mentioned Links:

Subscribe using your favorite podcast service:


note: this is a machine generated transcript and may not be completely accurate. This is provided for convience and should not be used for attribution.

Anthony: You’re listening to there auto be a law, the center for auto safety podcast with executive director, Michael Brooks, chief engineer, Fred Perkins, and hosted by me, Anthony Cimino for over 50 years. The center for auto safety has worked to make cars safer.

Good morning, gentlemen. Good morning world. Good morning to you too. Hi, good morning to you out in listener land. Listeners, how are we doing? You can’t really respond to me, so I’m just gonna assume everybody’s doing great. Unless you’re a pedestrian and an electric car is coming near you, cause According to a report in the Guardian, electric cars are more likely to hit pedestrians than petrol vehicles.

Petrol. Ha! For those who won that war, it’s called gasoline, dammit. Hybrid and electric cars are more likely to strike pedestrians than petrol or diesel vehicles, particularly in towns and cities, according to an analysis of British road traffic accidents. I think it’s because they’re driving on the wrong side of the road.

Fred: That’s got to be a factor.

Anthony: Without a doubt. And does this come down to just the fact that electric cars are silent and they don’t make sounds? Do they not have those requirements, like in the U. S. you see an electric car go past you and you hear the whistler sound?

Michael: Oh, it, it could be. I want to point out.

Just initially that this study used data from 2013 to 2017, which was prior to both the United States federal motor vehicle safety standard number 141 and the EU requirement for noises. Both of those requirements came out and I think the U S was 20, 20. Hey, we beat the EU to something. US was 2020 and the EU was 2021 and that required the little sounds that you hear when a hybrid is passing, I believe it’s under 18, 19 miles per hour when they’re required to do that.

This is only happening in urban environments, that there’s no evidence that there are any more dangerous in rural environments. Of course, the urban environments, the ones with the most pedestrians, right? Who’s walking in the rural world, right? It’s maybe it’s the noise. I, I.

I think there may be something to the fact that electric cars accelerate so much more quickly than ice vehicles and, consumer and driver and experience with that faster acceleration might contribute to more pedestrian crashes. But the noise is certainly a factor.

Anthony: From this article, it says people relied on sound to judge the present speed and locations of vehicles.

When these cues are missing, this could be very problematic for people in busy urban areas.

Michael: 100 percent true. Yeah. And it’s why those standards were put in place. I’d really be interested to see a study that was based on more recent data that could account for the change after the U. S. and the EU implemented the sound requirements.

Anthony: Yeah, this study, it just seems it’d be like a study coming out now saying Betamax users are really disappointed in their rental options. What are we doing? It doesn’t make any sense. But with noise related things, let’s talk speeding cars. Speeding cars in California, they passed a bill, oh, no, it’s not, it’s a bill up for a vote, that would force new vehicles to send out visual and auditory alerts every time a driver goes ten miles past a speed limit, which sounds and horrible at the same time because I’ve yet to be on a road anywhere in this country where anyone is not going at least 20 miles per hour over the speed limit.

So there’s just gonna be nonstop. I don’t know what the sounds are, but I’d like to, uh, put forward my resume.

Fred: Sadly, these are only internal sounds. So there’s no warning people outside of the car that the car is speeding. And that would be a nice thing, actually, if you were able to warn all the people around you that you’re driving recklessly.

Anthony: I like that. So what’s, Michael, what’s happening with here? Because the impetus behind this is that there’s a lot more crashes related to people speeding. And so the idea is, Hey, we’ll warn you, remind you that you’re speeding, but everybody knows they’re speeding.

Michael: Yeah. This is part of my, I don’t have a lot of ream of data to back up my views on this, people who really want to speed are going to speed regardless of whether they’re getting beeped at these systems beyond the beeps, there’s, I, They can be turned off.

There are ways to deactivate the system, which is probably shouldn’t be permitted if you really want to permit the people who want, prevent the people who want the speed from speeding. And ultimately, although I think that, intelligent speed or adaptive speed, intelligent speed assistance and these types of systems might have some ultimately you’ve got to tell the speeders. No, and force their hand. You can’t put in a system that could slow them down or could stop them and a system that they could just turn off whenever they don’t want to be bothered or whenever they want to speed.

Because, people that. Speed do so pretty consistently across their driving, and that’s the habit we need to break and not just break it, but just stop it cold in its tracks by putting speed governors on these vehicles. If someone is being alerted that they’re speeding and they don’t comply with the speed limit, then the governor would.

Slow the car down. It’s essentially I think it’s the only way we’re going to stop the large toll of deaths and injuries created by speeding is to actually require the cars to slow themselves down. Take that choice away from the drivers. Which is something that, other vehicle groups are not excited about doing.

In some ways, people see their ability to choose to speed as a fundamental right. Again, I don’t want to have to give the driving is a privilege speech again, but it, when you’re in a car I ha I personally have no problem and I would love to have my speed limited. Then I wouldn’t have to worry about looking at my speedometer every week.

Two seconds while I’m driving to make sure I’m not going too fast. And also I’d be, assured when driving in the city and other spots that, I’m not going to be put in a situation where I might be, legally responsible for a crash. When I’m driving over the speed limit. So I don’t think the majority of America sees it the same way as I do.

And I don’t think the majority of America wants to be restricted in their speed. And automakers are hesitant to put this technology into vehicles because they won’t sell the cars that it’s in if it’s, mandated a speed governor. There’s a long way to go on this issue, just not only to convince Americans, but to convince regulators and auto manufacturers that this is something that is going to work.

I don’t think there’s any question that it’s going to work. Ultimately the technology is available and could be working now. It’s simply a matter of whether we’re willing to accept it.

Fred: So what is the role of language here? Because they’ve got all these signs up all around the country that say speed limit is X.

But everybody completely ignores that because it’s really not a speed limit. It’s a speed advisory. And, I once back in my, Foolish youth, when I was stopped for speeding by a state cop in Massachusetts, I asked, what are your guidelines for giving out tickets? And he said if you’re more than seven miles an hour above the speed limit, then you’re likely to get a ticket.

Now, I just thought that was curious because If here’s a cop saying we don’t really enforce the speed limit. Why should anybody take the speed limit as a limit? Why, it’s really an kind of an advisory or a suggestion. And if you’re 20 miles an hour over the speed limit, people will say you’re reckless driving, but that’s the real speed constraint that law enforcement uses.

So would we be better off if. The sign said, speed advisory, you’re stupid if you drive faster than X miles per hour, discuss among yourselves.

Michael: I think we’d be better off if, the actual speed limits were enforced. First of all, there’s a lot of leeway given and depending on, what jurisdiction you’re in, the cops might have a 7, a 9, a 5, a 11, 12 mile per hour limit on what they’re willing to, Pull over a driver for doing so it changes all over the place.

There’s no real set rule. And so it’s, it really allows people to creep up and test the limits because there aren’t true limits. There’s not a hard right line. That drivers can go by where they know if they exceed that speed, they could be in trouble. They could lose money. They could ultimately lose their license.

They violate it too much. There’s certainly an enforcement issue here. Speed cameras and other things could help there. But. But ultimately, I don’t know why we would waste our time putting up speed cameras and all this stuff into infrastructure and the cost of, millions, maybe billions of dollars when we can be sticking this technology into cars from the start and stopping speeding where it starts.

Anthony: I think automatic speeding tickets. So every, your car is already connected. You’ve got a phone in it, or you’ve got an easy pass transponder or something like that. And as soon as you’re going over the speed limit for a certain amount of time, you get an alert. Hey, bing! 20 fine. 50 fine.

Points on your license.

Michael: So it’s detecting you via GPS.

Anthony: Yeah! Why not?

Michael: No car required.

Anthony: No! Boom! And you start getting points. Yeah. Yeah. Because I’m thinking, for most commutes and everything, it’s not gonna, you’re you speeding versus not speeding. It’s a one, two minute difference. For longer road trips, I think it’ll be quicker because most road trips where I’m taking these five hour trips there’s traffic jams happen when everybody sees a cop car parked in the media and they all slow down to 20 miles below the speed limit and this backs up and this huge thing, whereas you get rid of that entirely, everyone’s doing 55, 65, whatever it is, Smooth sailing.

Michael: I’d have to push back on that a little bit. If you’re on a, oh, if you’re Anthony, when you’re on a bus going to Atlantic City, wait to the casinos and you, the bus driver goes well over the speed limit and your phone is in the bus with you. Do you think you deserve a speeding

Anthony: ticket? No.


It’s tied to the vehicle. Tied to the, so it’s tied to the vehicle? Yeah. Tied to the vehicle. Not fun. You tie it to the transponders for the toll systems.

Michael: Ultimately, that’s great. It’s a fine. It doesn’t stop the behavior immediately that’s causing the safety problem.

Anthony: You get points on your license.

Eventually, by the end of a road trip, you don’t have a driver’s license anymore.

Michael: It’s yet another maybe small step on the way to preventing speeding, but ultimately, we have the technology to stop it now. Why don’t we just go all the way, stop it, save all those lives, prevent all those injuries, dust our hands off, and move on to the next problem in traffic.

Anthony: Because mine’s inspired more by a little bit of anger. So you hate freedom, Anthony, is that the deal here? I hate other

Fred: people’s freedom.

Anthony: Okay. I want them to slow down, not me.

Fred: And a corollary question. How long is your commute, Anthony?

Anthony: Let’s see. From my bedroom to my office. It’s, uh, it depends if all three cats walk in front of my path.

That’s a traffic jam. I might have to stop and pet them. 38 seconds if I crawl.

Fred: Some of our intelligent listeners might recognize that you have a smaller commute than many of the people who might be listening. Just as dangerous, though. But anyway, I still wonder about the language.

If, if it’s not a limit, okay, let’s say it’s not a limit. If it is a limit, treat it like a limit. The the automatic penalty. I don’t think it’s going to be wildly popular, but hey, why not?

Anthony: Take it to the limit. Is that see that you propose my ridiculous idea, then people will fine, I’ll just accept this other thing.

You go to the extreme where you anger people and then you get what you want. All right. We’ll shoot for marriage advice. Visit my other podcast called sleeping on the couch.

Hey, let’s go on. You guys were we’ve talked about this small little company and this very humble man Tesla and Elon Musk.

The world’s smartest man, because he’s told us he is Elon Musk says that his his autopilot is basically his adaptive cruise control. When people use it, they’re 11 times safer than other cars not equipped with the software. Ooh wee, guaranteed. And if you act now, I’ll throw on a big cowboy hat and some pretty juicy steaks.

In Gizmodo autopilot is primarily used on U. S. highways, while the national highway transport nor NHTSA data considers accidents on all roads. Accidents are more common on city roads and undivided roads. NHTSA data also accounts for incidents including all vehicles, including trucks, buses, and older car models.

It is not a clean comparison. So what they’re saying is that Elon is saying, Hey, here’s my apple. I’m comparing it to your bowl of garbage. Wait, no. I have a bo No, Tesla, I have a bowl of garbage. Look at that apple.

Michael: One way to think about this is when you look at accident statistics, that interstates and divided highways have very low fatality rates.

Per million vehicle miles, 100 million vehicle miles traveled than other roads, arterial roads and then urban streets. There’s a whole list of other roads that are far more dangerous than interstates, which is primarily where people are using autopilot those and what they’re doing is they’re taking the autopilot data that is primarily being used on interstates and divided highways.

We hope at this point, there’s still that camera and issue and geo fencing issue that hasn’t been resolved by Tesla. And they’re comparing that data from this, very peachy circumstance of being on the interstate to everything else that that all other drivers encounter in other cars.

On the roads, and they’re comparing it to, semi trucks and every other type of driver. And so there’s, they’re not comparing it directly, for instance, to us drivers on interstates or drivers in the same circumstances as people. The circumstances under which the autopilot is operating.

So just from the very beginning, the Tesla safety report is a lie. It’s incredibly misleading to use the data that way. And not only that. All Tesla’s providing is basically a chart on its website as its safety report. They don’t give us any of the underlying data. No one is allowed to look at that data.

Which is a sure sign that, Tesla’s just making this stuff up.

Fred: Michael, so is this your candidate for gaslight illumination this week?

Michael: It is. And it will be every time they put this report out because it’s a lie from the ground up. And it couldn’t. It’s not a good comparison.

It’s not the way anybody with any sense would, the only person who’s going to make this kind of comparison is someone who’s trying to lie to you and sell you something.

Fred: I’ve got another candidate for gaslight illumination in it. And since Anthony does not, he’s going to be the judge this week.

Anthony: I’ve got one, but yeah, you go on. I’ll judge you two anyway.

Fred: All right. This is related because of cars on roads, but Union Pacific Railroad is my candidate because there was an incident in Buffalo where a young man was playing near the near the train tracks. And a remote controlled train ran over him and his sister heard him screaming, had to run from the house to the railroad tracks where she saw her younger brother writhing in pain.

And then while they were there together, the locomotive started moving again. The train started moving again because there was no human being supervising its its activity. In the rail yard, and as she pulled her brother away from the train, which suddenly started moving, he lost his leg. So terrible situation, but Union Pacific responded by saying a locomotive is a massive piece of equipment, said Kristen South, a spokeswoman for Union Pacific.

But that pedestrian is going to be hurt, whether it’s remote or unconventional. Whether it’s remote or conventional locomotive, they wouldn’t be able to stop close quote. So I love this as the candidate for gaslight, because not only is it completely neglect the human situation has no sense of empathy for the injury, whatever, but it also says that, really.

The locomotive has no responsibility. It’s transfers the risk over to the victim completely by saying, they wouldn’t be able to stop because it’s a big, heavy thing. I think it’s with a rhetorical flourish of both avoiding any sense of empathy and also transferring the blame from the locomotive that, and train that ripped off this kid’s leg onto the kid himself.

I think that Union Pacific deserves this week’s Gaslight Award. It’s my candidate.

Anthony: That’s a pretty good one.

Fred: Anthony, over to you.

Anthony: That’s a really good one, Fred, and that’s from an article we’re linking to in the New York Times. It reminded me, here’s a quote from the article. No one from the railroad had heard the boy’s screams or noticed him trapped on the tracks.

The train had no conductor or engineer on board. Instead, its movements were being controlled by a remote operator who was not aboard the train. And under railroad protocols and could have been more than a half mile away. It reminded me of GM crews driving over a pedestrian and be like let’s just move the car to the side.

It reminds me of every AV that has no human content and no sense of shame, empathy, or fear of the future,

Fred: fear of consequences.

Anthony: Alright, that’s a pretty good gas lighter, but, and Michael, you had a very good one, come on, Elon is always a contender, but, you both lose, this week the winner is Mary Barra, CEO of General Motors.

Why? Because they keep making 9, 000 pound electric vehicles saying, hey, isn’t this good for the environment? Nope. They just came out with a GM Silverado, which is 9, 000, 90, 000. It’s not going to a job site. It’s not doing anything. It’s just gonna haul soccer balls. As the voter, I win.

I think that’s good, but I gotta support my gaslight because it’s two dimensional.

Both transfers the blame onto the victim, and also says that there’s nothing we can do.

The judges have listened to your thing and what, who, what do you win for winning the Gaslight Award?

Michael: Ego. Okay. I think we should get another one of Fred’s wonderful stories, but I’m not sure if he’s out from yet.

Anthony: I

Fred: could talk about the time I I hopped the train and That would be relevant.

Anthony: Okay. We’ll save that for some other time if we have time, but Fred, I enjoy listening to your dulcet tones. So you asked for some extra tau time today. So let’s do some tau time. And this episode is complimented use of gay boy, URL 4, 600 D O T R L for AV safety, who comes up with these ridiculous acronyms?

Fred: I don’t know, but who do, how do we make AVs safe? Let’s assume that’s an objective we want to accomplish. It’s a technical achievement if we can do that. So the question is there a precedent for how to develop technology? And is this precedent accessible to the Department of Transportation, for example?

So how would you do it? In a normal system engineering approach, you start with a set of requirements. And then you go down based on those requirements to the design. And then after the design is complete, and you’ve developed the technology, then you validate the requirements and that, the vehicle actually has the requirements and that it satisfies the requirements.

Very reasonable approach. The first step in that setting up the requirements is. What do you really want to do? Why do this at all? And what is our objectives? And that’s where it’s all

Anthony: question.

Fred: Yeah. And that’s where I don’t get it. So how would this work? You have technology readiness levels.

That’s that’s a concept that’s been well developed in the military and it turns out that the department of transportation does this as well. So one of the requirements. That you would impose on the development is the technology readiness level that you’re trying to achieve. For example, you have a different program if you’re trying to achieve a low technology level, and it’s just a laboratory experiment.

If you’re trying to achieve deployment, you have a much higher bar. There’s a lot more that you have to do. When you set up your development program, the first thing you need to establish is what are you really trying to do? It turns out that the AV industry has skipped over a lot of that. Because what what a reasonable person would do is they would take their set of requirements, and we submit that we have a pretty good one called the Consumer’s A.

V. Bill of Rights. And you would attach to that the technology readiness level you’re trying to accomplish so that you’ve got a good balance between the requirements you’re imposing and the objective you’re trying to achieve. Then you would use something like the UL 4600, Underwriters Laboratory 4600, which is a standard for safety case analysis.

To determine whether or not you’ve actually achieved those objectives. So that kind of closes the loop. You get to the end, you develop something, you got the requirements, and you come into UL 4600 or equivalent to say, okay, here’s the guideline. That says how we need to do this and, who’s going to sit down and what is the determination you’re going to make.

It’s a very large and well established cookbook, but it’s a cookbook nonetheless. You can just open it up. It’ll tell you exactly what to do. Given that the technology readiness level tells you a fundamental requirement of what you’re trying to achieve. It’s time for a spot quiz for our participants here.

Anthony: Oh goody.

Fred: Yeah, so there’s a document called the da. Let’s see. How do you spell it? Da? Da. It’s the Technology Readiness Level Assessment Guide, and it was developed by the Government Accountability Office. Participants in developing that guide include Waymo, which made a significant contribution and the Department of Transportation, which had 5 of their staff people on the committee developing this.

So it’s interesting that NHTSA seems to be unaware of these because it provides a progressive program, which leads to the deployment of new technology. So let’s see. Now we’re so here’s where the quiz comes in, because to achieve, as I say, to achieve a given technology readiness level, you should answer yes to every question.

So gentlemen, I’m going to ask you these questions and you tell me whether or not you can answer them with a yes. Did we do this before? We didn’t do it in this much detail. Okay. Okay. But for some reason, my computer isn’t helping me with this because I can’t see what the hell they are.

Anthony: You’re pulling that up.

I have some questions. So with autonomous vehicles, the goal, no one really clearly states the goal cause it’s Silicon Valley and they go, we’re changing the world, man. And it’s all bullshit. Yeah. So is the goal is to remove human error from driving. Is that fair to say? That’s fair to say,

Fred: okay, but okay, getting back to the, the technology.

So there’s nine technology readiness levels. And number nine is for technology refined and adopted, and that’s the implementation level. So if we look at the questions, is the technology deployed in its intended operational environment? You can say hot road testing would do that. Is information about the technology disseminated to the user community?

Maybe. Is the technology adopted by the user community, which is the public at large? The answer to that is clearly no. So they’re not ready for implementation. Going down number eight, does the technology meet a stated purpose and functionality as design?

No, we have a lot of reports of vehicles going up the wrong way of roads and driving drunk and doing all those kinds of things.

So I would say that’s a no. You agree? Agreed. All right. So that means it’s not level eight. So we’re two away from deployment. Number seven. Are all interfaces tested individually under stressed and anomalous conditions? We have seen no evidence of that, have we?

Anthony: What is, what do you mean by stressed and anomalous conditions?

Fred: Let’s say you’re going down the road and you’re designed the vehicle to travel at a certain speed and to turn right at some other speed. Okay. So what you want to do is you want to establish how close you are to critical boundaries. So you would do a test where you’d say, okay, I’m going to make a right hand turn at 10 miles an hour because I’ve designed it to go 30 miles an hour.

Okay, that’s fine. You test it again at 20 miles an hour, and it does okay. You test it again at 30 miles an hour, and let’s say for the sake of argument, it does okay. Now, to stress it, you would say, okay, I’m going to test it at 35 miles an hour. And you test it at 35 miles an hour, and it does okay. That means you have a 5 mile per hour safety margin associated with that maneuver under those conditions.

Okay. If, on the other hand, you tested at 35 miles an hour and it goes wildly out of control, you know that the limit you’ve placed on the vehicle is very brittle and you’re very close to a precipice, if you will. So you’re right on the edge. And if you exceed that limit, you’re going to go over the edge.

So that’s what stressing means.

Anthony: Okay.

Fred: So then, get back to that. So for AVs, any brand, you gentlemen can pick the one that you like, are all interfaces tested individually under stressed and anomalous conditions. I would submit the answer is no, because what they’re trying to do is just drive these things on the road and see where the problems are and then backtrack and say we’re trying to fix this

Michael: later on.

That’s a tough way to address it. The question here, because you’re the way they’re testing them is, in a sense, they’re discovering anomalous conditions while they’re testing. They’re not. So much testing to those conditions as they are finding them out as they test, they’re not quite up to snuff here yet.

Fred: And another example, the crash that occurred in Mountain View, California, where the Tesla ran off the road, and I’m including Tesla in this because Tesla’s Despite their protestations that they’re a level two, they’re really a level three car. De facto, that’s how people use them. And, so they’re included in this.

Anthony: People use them as a level four car, but technically they’d be more of a level three.

Fred: So the case that happened there was that the vehicle failed to recognize that there was an exit ramp. Called a gore and it did not navigate the gore properly, ran into a barrier, killed the driver injured driver in another vehicle.

If you say that’s a stressing case, then it would make sense for companies to set that up on a test track somewhere and make sure that the vehicles they’re designing can handle that perfectly well. Is there any company doing that? No, we haven’t seen that.

Michael: They’re clearly using humans as those test cases.


Anthony: Alright, there’s a video I’ll put a link to, this guy using full self driving, quote unquote, around New York City and he has to interview you a couple times because full self driving doesn’t understand taxis or pedestrians in Manhattan. Oh.

Fred: So clearly the answer to this is no, and so we’re not at that TRL 7, so let’s go down to TRL 6, which is prototype demonstrated in a relevant environment.

And it says does the prototype satisfy all operational requirements when confronted with realistic problems? I’d say a realistic problem is an exit ramp that the vehicle cannot execute and instead kills the driver. You guys think that you guys think that there’s any on the road? A realistic problem.

Michael: A realistic problem here would be what a Navy should do after it hits a pedestrian, which we saw in the GM cruise case. They have not. Established any way to, to fix that situation.

Fred: We’ve also seen Waymo going up the the wrong way on a multi lane highway. And by the way, Waymo was one of the people who one of the companies that contributed significantly to this development.

So be hard for them to say that they don’t know what this is all about.

Anthony: So this sounds slightly unfair, but what these questions you’re putting forth, have any of these companies. Publicly defined. Hey, these are our operational goals.

Fred: If they’ve defined them, they haven’t shared them with the public.

There we go. All right, so clearly we’re not at level six, which is prototype system demonstrated in a relevant environment. So level five is integrated components demonstrated in a laboratory environment. A question here is component integration demonstrated in a laboratory environment, i. e. fully controlled setting.

I would argue that this is a yes. I would say that this is the technology readiness level that all currently operating on the roads. Have achieved so the question really behind all this is, why is the Department of transportation, which had 5 people in the development on the committee developing these guidelines and has their own section on developed Department of transportation.

Technology, readiness, local guidelines. Why is this not part of the process for approval of on public roads? Why. Why hasn’t the department?

Anthony: I know the answer to that. The answer to

Michael: that is because there is no process.

Anthony: Because it’s done at the state level, not the federal level.

Michael: Yeah, and there is no federal process there.

And even the state level would allow for the qualification of vehicles that don’t get anywhere near technology readiness level number nine, which is where we should be when we’re deploying these things on public roads. Because

Anthony: these companies give a lot of money to politicians. For more information, visit www.

fema. gov Really,

Fred: That’s all true. So I would argue and helpfully. I will let the Department of transportation listeners know that they should really be looking at page 124 of the of the document that they subscribe to, because that’s where the are and please don’t approve cars. Or use on the public roads until you’ve achieved tier level 9.

And you’re able to demonstrate that you’ve gone through this progressive development process that you endorsed and that you participated in before you put these damn things on the road. People are dying and, you’re doing people no favors by skipping over the steps of the development process, just because the companies think it’s convenient to go ahead and put this stuff out there.

And Thank you. End of rant.

Michael: Fred, one question about these. In the context of AVs, companies would be seeking to reach level 9 within a particular specified operational design domain. They wouldn’t You know, level nine doesn’t mean that you’ve got an AV that can drive in mountains and snow in the winter.

It means that you are at level nine, say, in Chandler, Arizona, on flat roads and in ideal conditions.

Fred: No, that’s right. The other, the three questions that you need to answer yes to, to consider your vehicle to be at implementation readiness at level nine. Number one, is the technology deployed in its intended operational environment?

That’s the ODD. Number two, is information about the technology disseminated to the user community? Presumably that would happen as part of their marketing. And finally is the technology adopted by the user community. It’s a little bit nebulous, but that’s what I think that the way you would approach that is, are people approaching your showroom with shovels, pitchforks and axes, rather than with a happy attitude about buying this vehicle.

But, this that requirement is a little bit nebulous. At

Anthony: what level is my car free of body thetans? Free of what? Body thetans.

Michael: Are you going into Scientology? Ah,

Anthony: yeah, a little bit. What, so Fred, I get the impression that the, that you like this, the, what the DOT came up with, right?

That, those levels are good? I do, yes. Okay. Yes,

Fred: absolutely. It’s been in use for many years in many different programs, and it works.

Michael: And it’s not just the D. O. T. coming with them. I think there’s a lot of people involved, including NASA use this system to, figure out if their spacecraft are ready.

Fred: Absolutely. NASA, the Department of Defense, a couple of big users, but so is Boeing involved in it. And, the company, virtually every company that makes complex technical, systems. Items for use by the public is involved in the development of these technology readiness levels and they use them because they’re useful and they’re appropriate.

So again, the complex of setting requirements. Using the technology readiness levels to say, this is what we’re trying to achieve at this level of development using the consumer 80 bill of rights or something equivalent to say, this is what you’ve got to accomplish regardless of the technology readiness level.

But, ultimately, before deployment before level 9. You’ve got to have all those requirements satisfied some subset for lower levels of the requirements, okay? And using something like Underwriters Laboratory 4600 as a way of testing whether or not your actual performance conforms to the requirements you’ve set out for yourself.

It’s, three items that close the loop on how to develop AV safely. And to our friends at Waymo, Cruise, Aurora, and a lot of other, Kodiak, a lot of other companies, you’re welcome.

Anthony: What I don’t understand is none of these AV companies have come out and really said, hey, this is the problem we’re going to solve.

They’ll sum up and say, we’re going to make the roads safer. 40, 000 people die a year, and we’re going to make it safer. Okay, how do you do that? And they go, magic. None of them really have laid out how They’re doing just, it’s we do it. It’s going to be amazing.

Michael: Most lay it out in terms of avoiding mistakes that humans make without even mentioning all the mistakes that computers make.

But also we

Anthony: don’t even know all the mistakes that humans make.

Fred: That’s another, it’s another rhetorical flourish. Okay. Cause they’ve misinterpreted a study and said 94 percent of the accidents are caused by human beings. Damn the torpedoes, full speed ahead. We’ve got to get this technology out as quickly as possible to save people’s lives.

Of course, if you read that study correctly, it says something quite different. And there’s nothing in there that says 94 percent of accidents are caused by human beings. But, the first thing you want to do in sales is to create a sense of urgency, you need to get somebody to do something right now.

So that’s a rhetorical flourish that they’re using. And it was endorsed by a lot of people at NHTSA in the previous administration, because frankly, it’s simpler to do nothing than to do something and to say in the future, everything will be better if we just let. Things develop by themselves. That’s not actually going to work.

Anthony: Listener, if you don’t go to autosafety. org right now and click on donate, bad things will happen. You might get a paper cut today. Hey, just think bad things happen. Salt might fall on it and a little lemon juice. So go to autosafety. org today and prevent these lemon juice salted paper cuts. Why not?

That’s how I generated urgency. That may be a bit of a reach, Anthony, but I think donation is a great idea. Yeah, you’re not in our sales department. So around this, I’m watching this guy drive with full self driving, beta, whatever, blah, blah, blah, around Manhattan. And the way he does it actually I think is really good because he is 100 percent like he’s probably using a lot more cognitive load having the system on than not having it on, than if he was just actively driving because he’s, he’s making a video for his YouTube fans but he’s fully engaged, fully seeing, okay, what, it’s not doing exactly what I want, I have to take over, I have to engage and whatnot.

If people actually use these systems in that way, maybe that would make things better. Better. It was also forced to

Michael: understand that it doesn’t work in order to safely drive it, whereas what Tesla is saying to you is this works, right? And that’s the fundamental problem. There’s that drivers if they don’t think if they think the technology works.

Then why are they going to waste all that extra cognitive power monitoring it? Whereas if you tell a driver it works sometimes, but not always, and you really have to keep an eye on it. It’s a totally different approach. Exactly.

Fred: And that’s the whole point, I think, of the lack of safety with these AVs.

The Swedish study that we quoted a couple of weeks ago says that, or actually demonstrated that, it takes about 15 seconds for somebody who is distracted. To actually take full control of the vehicle again, even in fairly benign takeover circumstances, if there’s an emergency, it’s even worse. But think of that as, you’re actually increasing the cognitive load of the driver if they have an AV system running.

And they’re supposed to be the supervisor of that because you’ve got to constantly make the decisions about what you would do if you were driving the car. And also, you got to monitor what the car is doing, because you don’t trust it. So you’re, you’re actually in a worse situation than if you didn’t have any automatic driving at all.

Anthony: Exactly. Okay. So let’s go into some more pop quiz territory. How’s this? Okay. Our friends at Vox they put together a lovely article about gas mileage, fuel economy, and how it’s nonsense. So their pop quiz starts off with which of these saves more gas swapping a car that gets 25 miles per gallon from one that gets 50 miles per gallon.

Or, replacing a car that gets 10 miles per gallon with one that gets 15 miles per gallon. Or C, if a man gets on a train in Chicago and somebody else gets on a train in Philadelphia. No. No, that’s not part of the quiz?

Michael: No. Word problems get me enough when you’re trying to add a C in there. It’s hard.

Anthony: Okay. Fred, did you read this article?

Fred: I did read the article.

Anthony: Dammit, then you already know the answer, probably. You’re, you and your math brain probably already knew the answer ahead of time, huh? Yeah. I’ve got

Fred: some experience with something called long division, and it’s really helpful when you’re trying to analyze something like this.

Anthony: See, in my day, we had calculators, and there’s the kid next to us, and he wrote really big, and just look at him. I know. Basically, this is saying, hey, though, the one that gets from 10 to 15, I don’t understand it. Fred, help me under, help me understand why, if going from a car that gets 25 miles per gallon to 50 miles per gallon, Is not as big an improvement as 10 to 15.

Fred: The underlying assumption is that the the cars are going the same distance. So it depends on how far you’re going. But if you’ve going from 10 fif 10 to 15, you’ve got about a 50% improvement in mileage. And because you’re burning up a lot of gas the denominator is the numerator, which is the gas is quite large.

So 50% improvement in some, in a number that’s quite large is better than 50% improvement in a number that’s relatively small, right? So if you talk about the mileage of a high mileage car. The number of gallons it consumes is much smaller than a low mileage car. So the same percentage improvement in a car that’s not burning a lot of gas is not as consequential as a percentage improvement in a car that’s burning a lot of gas.

So that’s the heart of this puzzle.

Anthony: Okay, and it seems that the way that we measure miles per gallon in the U. S. is different than the way the Europeans do it. And I’m not just talking because they use liters.

Michael: They use liters. They use what is it, liters per 100 kilometers driven. So they’re doing the math that’s required to figure this problem out.

Whereas in America, we’re just using miles per gallon, which doesn’t take into account, the distances traveled and the savings that are relative to each one of those scenarios. And it’s a problem because it’s, it, the miles per gallon. They call it the miles per gallon illusion here because it is somewhat an illusion.

They use that in calculating the corporate average fuel economy standards. And there’s a pretty good argument that this has contributed to some problems there. It’s also, the authors here contend that it, The mile per gallon illusion helps conceal the distortions of the structure that’s been created in CAFE, which has one fuel economy standard for passenger cars and another standard for like trucks, SUVs and pickups.

And that’s what we’ve called in the past, the light truck loophole which doesn’t, instead of putting pressure on manufacturers to make those Poor performers better. It allows them to do things like make their best performing cars, a little bit better, and then use credits from those to offset the bad cars they have, but ultimately we don’t get really any improvements and overall fuel economy or emissions.

And they make a pretty good argument, for changing the entire system to go more towards a. Hundred, how many gallons does it take to go a hundred mile system versus how many miles do you get out of a gallon? But it’s, it’s definitely a word problem and a math problem that is beyond many of us out there.

Anthony: But it sounds like a relatively easy fix. It seems like it would be.

Michael: It sounds easy, but when you look at the docket the federal docket for the corporate average fuel economy each time they change it, you’re talking about hundreds of thousands, if not millions of pages worth of stuff, data and other things submitted by manufacturers, government, environmental groups, all sorts of different Stakeholders, and there’s a lot of politics that goes on there and making common sense solutions happen in that environment is difficult.

I think it will be difficult to change the way that we have it set up in America under the mile per gallon standard.

Fred: And it’s going to happen in my lifetime.

Anthony: Wait a second. You’re always the, things will always be better in the future,

Fred: huh? That’s an aspirational statement, but yes, of course, we’re always headed in the right direction.

Anthony: Okay, good. Hey, last week we talked about Waymo, and about how, hey, we’ve got video of them driving drunk on the road, and this past week NHTSA sent a letter to Waymo, Saying, hey, this is from our Office of Defects Investigation, which I have no idea why this is not a Law Order show. Law Order.

ODI. Kung. No, it’d be like an engine. Brum. That

Michael: would be awesome.

Anthony: That would be awesome. So this is a PDF and we have a link to it. And normally I get it if you’re like, I don’t want to click on a PDF. This PDF is 100 percent worth clicking on it. Yeah. Because they list out all the things.

All of the incidents that NHTSA identifies saying, Hey, this is the issue. But beyond that, they list out a whole bunch of YouTube and Reddit links of people reporting stuff. And these YouTube videos are the best thing. They are so much better than the 11 foot, 8 inch truck bridge overpass. They’re totally worth it just for that.

Check these out.

Michael: I’ve looked at thousands of information requests out of NHTSA over the year, and this is by far the most entertaining one and interesting one, because it has, concrete examples videoed by folks writing or around. Waymo vehicles when they are doing bad things, including not being able to exit a postal service parking lot where the vehicle is basically driving in circles until someone, an actual driver comes out to remedy the issue.

And some of the other ones that we’ve talked about with the unicycles and Waymo is going down the street the wrong way. Waymo is turning into the wrong traffic lane when trying to enter traffic and going the wrong way again, basically you look at it and you say, these vehicles are drunk, there’s something wrong with the driver here.

And these are the type of incidents that, we point to every time someone from the A. V. I. A. or Waymo or Cruz says that humans are the bad drivers here. And these vehicles are the ones that are going to save us from ourselves. They’re not there yet. They’re not even close. And you can watch them and see them.

And it’s. It’s very clear to anyone watching these videos that these vehicles are not there yet and that they are every bit as bad, if not worse, than a lot of human drivers.

Fred: So I’m starting to get the impression that you gentlemen feel that the idea of putting multi ton vehicles on the highway that don’t control themselves very well and that could impact pedestrians and other vehicles is a bad idea.

Anthony: Have I got that right? You have that 100 percent wrong because without those vehicles, what else am I going to watch on the internet? Oh, that’s true. Good point. Alright, let’s go into recalls. This week we’ll start off with Kia. Kia, the 2022 Kia Niro EV. The Kia Niro, which is the Chevy Nova of car names.

Certain 2022 model Niro EVs manufactured from some date to another date. The contact surfaces of the high voltage battery safety plug may develop high electrical resistance. Resistance. All the exact cause of this high resistance and the safety plug is not being confirmed. KIA suspects it is due to supplier problems.

Blah, blah, blah.

Michael: Yeah, this one is basically there is a supplier problem. They’re contending. They don’t appear that they have figured out exactly what happened there, but they’ve got a female to female terminal that was apparently not working. And the recall fix completely removes the female to female terminal.

It looks like. So even though they don’t know the exact cause here, it looks like they’re going to have a fix out and available. In about two months, July 19th is what it looks like. They’re telling owners they’ll have that fix available with a solid bus bar connection. No more female to female terminal.

Anthony: Did they supply this car with a little violin?

Michael: They should.

Anthony: They should, indeed. I, again, Tia, your marketing department, I get maybe Nero means something different in Korea. But, historically, not that great of a name. Wasn’t

Fred: that, wasn’t that part of the Titanic Nero, my god, to thee, as they were playing it as the ship was going down?

I never saw that movie. Now you’re stretching. That’s too much of a stretch already.

Anthony: Alright, next up, Jaguar, 22, 268 vehicles, the 2018 2022 Land Rover Range Rover Sport. Man, how do you get a job in naming vehicles? You just is it a drinking game? Is it really a, the Land Rover Range Rover Sport?

It is a non compliance recall. A concern has been identified on certain Range Rover Sport vehicles equipped with the Surround Camera System. No, it’s the camera. Where the reversing camera will display a poor image or no image at all as a result of water ingress into the camera housing bezel. Keep in mind that Tesla has all of their self driving cars just by cameras.

And this will never happen to them.

Michael: Yeah. This one’s basically water getting into the camera and, that violates it basically prevents the rear camera image from reaching the driver. And so it violates federal motor vehicle safety standard number 111 for backup prevention.


Anthony: All right. Last recall, Volkswagen, 17, 797 vehicles, the 2022 to 2024 Q4 e tron SUV and sportback affected vehicles were equipped with headlight control module software intended for the EU market that may not comply with FMVSS. Come on, what’s the number? It’s not 111. It’s not 120. Oh yeah, someone cheated.

He looks. You didn’t even get bread a chance

Michael: that one is. This one’s interesting. Basically they’re building all these vehicles and then right before they ship them to America or Europe, they install the whatever countries laws are controlling, they install their, Software based on those.

So here they shipped a lot of Audi’s to America that had the headlight software that was intended for vehicles in the European Union. And since our standards don’t match up with their standards, they’re out of compliance. Here, it’s basically if you’re driving one of these out is you have the EU software in your vehicle now, and they want you to put in software that’s compliant with the American version.

I, there’s a lot there as to which one may be, may or may not be safer. There are different glare standards between the countries and, there’s, Essentially, it’s a non compliance, but there’s not a safety defect here.

Anthony: Brad,

Fred: do

Anthony: you disagree?

Fred: No, I was just going to say, you’re probably better off to keep the old one.

Because what was originally installed in there was the adaptive lighting system that will be legal on U. S. cars, I think, what, next year?

Michael: Yeah. It’s coming up soon. And, I even, there’s even a triple A study that compares the European and United States headlamp specifications and finds that, the European standard may be a little better.

They’re pretty close, but that’s why, wanted to bring that up that, this may not be a critical safety issue. You’re functionally just bringing your vehicle into compliance with us regs.

Fred: Yeah, so to be clear we here at the Center for Auto Safety endorse getting your your updates as provided.

In this case, it’s a little questionable, but, be on the safe side and go ahead and get the recall implemented. Yeah,

Michael: if this was an update that, corrected. If your vehicle had been programmed to think it was driving on the left hand side of the road versus the right hand side of the road as a few European islands allow for, then that would be a very significant and scary issue.

But here, yes, go get the recall and get it repaired. But note that, the European standard is not going, probably not going to endanger you.

Anthony: All right. And with that listeners, what have we learned today? We’ve learned that humans probably aren’t that bad. We’ve learned that naming vehicles is a drinking game.

We’ve learned that If you want to have really fun YouTube videos, look at the Office of Defect Investigations PDF we’re linking to. I I think that’s about it, listeners. Do you have anything else to add?

Fred: Not a thing, but I would like to say thank you for listening. I, of course I’m listening.

It’d be rude if I didn’t. I wasn’t talking to you, Anthony. Okay. I was talking to our esteemed

Anthony: listeners. Alright, I call them donors. Hey! Michael, do you have anything to add before we wrap up? No, I’ll wrap watched

Michael: today.

Anthony: Thanks for sharing. Hey, everybody, come back next week. We have a lot more stories.

Each week we have a ton of stories we have to skip or else this would be an 18 hour podcast. And it wouldn’t be a podcast telethon. Thank you. Goodbye. Thank you. Bye bye.

Fred: For more information, visit www. autosafety. org.


Join the discussion