Safety on Autopilot: The Pitfalls and Promises of Partial Automation

[00:00:00] 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.

Hey listeners, welcome to their auto be a law. This week we’re brought to you by. I got nothing. I didn’t plan ahead of time. I was just trying to riff and see what will come up. And nothing did, but Hey, where are your seatbelts while you’re listening to this podcast? If you’re not in the car, go get a seatbelt, make him out of a household appliance or a t shirt or a rag.

[00:00:46] Michael: Don’t do that.

[00:00:47] Anthony: Okay. Don’t fine. Whatever. Don’t judge. And

[00:00:51] Michael: you have a pretensioner for that rag.

[00:00:53] Anthony: I, Oh, I don’t know what that means or why I made that sound. But anyway, look, we’re on the center of road safety [00:01:00] podcast. This week.

[00:01:02] The Challenges of Partial Automation Systems

[00:01:02] Anthony: Let’s start with assistant driving, automatic driving assistance.

There we go. So there’s a couple of good articles I want to link to. One is from the insurance Institute for highway safety. And it’s talking about how, when we start using partial, I can’t speak, partial automation systems like lane keeping assist or adaptive cruise control or Some mislabeled things like Autopilot, that we we trust them a little too much and we learn how to get away from their quirks.

IIHS did a great study, and quoting from their article, These results are a good reminder of the way people learn, said IHS president David Harkey. If you train them to think that paying attention means nudging the steering wheel every few seconds, then that’s exactly what they’ll do. Basically, their study shows that, People do not become safer when you’re using these partial automation systems.

[00:01:57] Fred: Talking is so difficult.

[00:01:58] Michael: It’s really hard. You [00:02:00] know, apparently driving safely when you’ve got one of these partial automation systems on is difficult as well. Because it’s essentially the insurance Institute did two different studies. I believe one was on the Volvo pro pilot or pilot assist.

Yeah. And the other was on Tesla autopilot, two separate studies where they looked at how drivers work with these partial automations. And what they really, I think what the upshot of the whole thing is that drivers, humans learn quickly and drivers adapt to these systems.

Fairly quickly and start using their additional time not to make sure they’re, driving safely and getting from point A to point B without incident, but by distracting themselves with other things. And, for instance, in the Tesla system, I don’t believe Tesla relies on those steering wheel nags anymore.

I think they. Took that out of the system recently, but at the time of the study, they were using the steering wheel mags [00:03:00] and part of the study, they found that the Tesla drivers were had trained themselves essentially to simply touch the steering wheel every half minute or every few seconds, whenever they needed to avoid the warnings, but.

They weren’t necessarily paying more attention to the road or focusing while they were touching the steering wheel. It was simply their adaptive behavior that allowed them to keep focusing on their movie or their lunch or whatever else they had going on. It’s a couple more studies that are in a group of studies that have come out recently that suggest Some of these, adaptive cruise systems and other automations aren’t really contributing to safety.

They’re really a convenience feature for drivers.

[00:03:43] Personal Experiences with Adaptive Cruise Control

[00:03:43] Fred: That mirrors my own experience with my 2020 Subaru. It’s got a similar sensor system and wants you to move the the steering wheel every few seconds. So naturally I became compulsive and determined that it’s a 10 second interval. [00:04:00] By counting down and, then after I discovered that the adaptive cruise control wears out your brakes very quickly, I immediately changed to compulsively watching the mileage meter on my car when I use the adaptive cruise control and switching it off and on, depending on whether or not I’m going uphill or downhill, which is, gives 10 seconds here, because I live in a hilly area.

Yeah, that happens. People become very adept at either defeating or accelerating their adoption of these automation features as they become used to them.

[00:04:39] Anthony: I love Lane Keeping Assist and Adaptive Cruise Control. They’re great. I don’t start watching movies. I don’t start cooking a meal or anything like that.

I just For my steering wheel, I don’t have to touch it, I just have to weigh, weights on it. So I basically just hold my hands on it. So my hands are still active on the steering wheel. And I feel, oh, cool, look at man, it’s turning. Oh, [00:05:00] this is cool. Adapter of cruise control I love because every car behind me hates me because I’m the only car on the highway with a three car gap.

Between the car, a car in front of me, which I, again, I love because people are nuts. But then Fred, you mentioned this before in a previous podcast that your brakes wear out more on adaptive cruise control. And I don’t know if that’s unique to Subaru’s or I don’t, that’s something I’ve always wanted to know with my adaptive cruise control.

Like when it slows down, I don’t, is it doing compression breaking with the engine? Is the brakes activating and the brake light actually appearing on my car? I don’t know.

[00:05:35] Fred: I don’t know about your car, but for my car, it’s solely through the application of brakes. It’s no compression braking.

And the lights do come on if the, if there is enough of a deceleration going on. Okay. So in fact, I’ve got a display on my, Dashboard that shows the lights going on and shows what’s going on when you’re in adaptive cruise control.

[00:05:57] Anthony: Oh, mine doesn’t do that. Huh.

[00:05:59] Fred: [00:06:00] Yeah. But I found so here’s my reference point.

I had a 2010 Subaru Outback, which had cruise control, which then only What’s not adaptive. So it would turn itself off whenever you had to. And it when it was going downhill, it didn’t do anything. It just coasted and my brakes lasted the first 70, 000 miles. In my current Subaru, which is 2010, the brakes were completely worn out after about 30, 000 miles because I was using adaptive cruise control a lot.

So yeah, the brakes are wearing out a lot faster.

[00:06:35] Anthony: Yeah. The main complaint I have with mine is if. Say you have adaptive cruise control set to, the speed limit’s 65 miles per hour. Someone gets in front of you and they drop down to 40. And your car slows down, which is great, but then they get out of the lane, and all of a sudden my car is like, LET’S SLAM ON THE GAS!

And it drives in a way that normal humans wouldn’t. And the problem I have with this is it tends to wake my wife up and she’s What are you doing? It’s [00:07:00] not me, it’s the car, which is the wrong thing to say. Never tell anybody you’re in a relationship with that you’re not driving the car. It’s, the car’s driving itself.

It doesn’t end well.

[00:07:10] Fred: No I’ve got both lane keeping assist and adaptive cruise control. And I found that when I’m passing a truck and I’m following another car, And all of a sudden the other car clears, moves into the lane in front of the truck, and my car lurches forward. It also tends to lurch to the right directly in front of the oncoming Class 8 truck.

So it’s scary as hell. I rarely use them anymore. Oh, wow. The more I use them, the less confident I became that they were keeping me safe.

[00:07:43] Anthony: Thankfully, I don’t have that experience. Michael, do you use these features in your car? Does your car have these features?

[00:07:47] Michael: I don’t, my car doesn’t have the lane keeping assist.

I, I doubt I would use it at all if I had it. I just, because I don’t know if I’d be able to trust it. And there’s, I think the typical experience, is. [00:08:00] At least my parents experience is they turn, they try it out, it annoys them a few times and they turn it off. And I, I think it’s somewhere around 60, 60 percent or higher of consumers that, effectively abandon those systems after trying them out because they just don’t like how they work.

I can see in the future if, these systems become better. And anticipating the type of traffic in front of them. The one thing I noticed mostly about it on the highways is cars who are in this mode will pass me and then slow down when there’s a car in front of them and they won’t even notice they’re slowing down because the computer’s doing it for them.

So I, continually catch up and repass, and then they repass me and it’s, it seems counterproductive to a good flow of traffic, maybe if everybody had them turned on, it wouldn’t be that way.

[00:08:50] Anthony: Yes, everyone use adaptive cruise control. From the Ars Technica article, it has an interesting little line that something we’ve talked about, and it says the Volvo [00:09:00] and Tesla studies were complicated by software updates during the study periods.

Two that affected the Volvo Pilot Assist, and 12 different versions of Tesla software during the study’s duration. And we’ve talked about this a lot, where, you know, someone, the version of software we’re using today, and they say, Oh, there was a problem with it. Then they have a new version tomorrow.

Maybe that fixed that problem, but we have no idea if it introduced anything else, or these other cases.

[00:09:26] Michael: Yeah, it has to be just incredibly frustrating as a researcher at the Insurance Institute to conduct a study on a vehicle that is constantly updating its software. It’s, I guess it makes it hard to draw firm conclusions when the autopilot software has been updated 12 times over the course of your study.

So I think they did a good job.

[00:09:51] Fred: It’s incredibly difficult and forces people to aggregate the data in ways that are completely inappropriate. We’ll talk about this when we get to the Waymo [00:10:00] data as well. So I’ll leave it there.

[00:10:03] Anthony: All right.

[00:10:03] Waymo’s Safety Claims Under Scrutiny

[00:10:03] Anthony: Hey, maybe let’s get to some Waymo then. Look at that. I’m going to I’m going to go into Gaslight a little bit here, because Waymo’s PR team was on overdrive the past week and a bunch of articles were written about how Waymo’s terrific and it’s the best thing and it’s from organizations that things I like, such as Ars Technica. And a lot of different places are like, Waymo’s safer than a human.

Of course it is. And so my Gaslight, jumping right into it, is an article from Ars Technica. And the line that is pure gaslighting is, 20 injuries might sound like a lot talking about Waymo crashes, but Waymo’s driverless vehicles have traveled more than 22 million miles. Which version of software? So driverless Waymo taxis have been involved in fewer than one injury causing crash for every million miles of driving.

A much better rate than a typical human driver. This is pure gaslighting nonsense, because we’ve discussed this. Oh.

[00:10:58] Michael: I, my biggest [00:11:00] problem with the way that Waymo and, reporters are characterizing this, the crash rate and the death rates that we see on the roads from humans don’t really reflect good human drivers.

They reflect the worst human drivers. Making a claim or setting a benchmark as Waymo has done here that corresponds to the absolute worst drivers in America is a non starter for me, as someone who is hoping that these vehicles are going to significantly increase safety on the road.

I want. I want Waymo’s to drive better than me. I want them to drive better than, someone with a clean record at the DMV, someone who doesn’t have a trail of speeding tickets, someone who doesn’t have, doesn’t drive drunk, someone who doesn’t use their phone while they drive. All these, all this data simply proves is that, Waymo in certain areas, primarily cities with lower speed limits [00:12:00] is.

Not making as many mistakes as the worst drivers on the road because they’re using crash data, which essentially is going to pull in the worst of the worst. They’re not comparing this driver to a driver that has a clean record and seeing which vehicle. The data is hard to work with here, and I acknowledge that.

And even Waymo acknowledges that it’s really difficult to find good data to create benchmarks for how safe. Human drivers are versus the way most system, but ultimately what they’re comparing to is the worst drivers. The drivers responsible for all the crashes, the drivers who are drunk, the drivers who, you know, and it’s just not compelling to me if we’re going, if these, they continue to advertise these vehicles as going, they’re going to save us all from our own inherent Who are driving and ultimately to do that, I think they’re going to have to be better than the reasonable human driver.

They’re gonna have to be better than what we would expect from a [00:13:00] decent drive on our roads. And right now they’re only comparing themselves to the lowest common denominator, which is the people who are speeding and crashing and drunk. So it’s not compelling to me, from a data perspective.

[00:13:13] Fred: So I’m going to agree with you guys and dive in a little bit deeper. I think it’s brownwashing rather than gaslighting and brownwashing for our attentive listeners playing the home game is a changing bullshit into purity of the driven snow. So let me quote from 1 part of their website. It says, quote, focus should be put on injury, causing crashes, low speed crashes that result in minor damage.

Can cause property damage that will be quickly repaired. Close quote. The answer is no, because way more statistics would then be based upon accidents or collisions that rarely cause damages because there’s speed limited in the [00:14:00] city to 25 miles per hour. Below 25 miles per hour, people are only rarely injured in crashes.

So if that’s your basis, then you’ve automatically skewed the data to a very safe perspective compared with the human drivers who are driving at a wide range of speeds. So let me just suggest how to make it. Apples to apples comparison apples to apples comparison would include comparison with other expensive new cars rather than with the aggregate population that includes a lot of bad cars, a lot of old cars not the overall crash rates, it would compare the routes traveled by the waymo versus the routes that everybody else is traveling.

It would include vehicle speed restrictions that waymo invokes rather than. The unrestricted human drivers, it would include comparison of weather and time of day and environmental conditions missing from the data that they present. It would [00:15:00] also include the software version, which ideally would show statistically significant decreasing incidents with successive software versions, which nobody has ever shown.

Waymo has shown

[00:15:10] Anthony: they haven’t done that as they’ve driven into a telephone pole with a recent software update.

[00:15:15] Fred: In fact, showing that cars are getting safer, their vehicles are getting safer with successive versions of software seems like the easiest thing in the world to show, and yet nobody that I’m aware of has illustrated that in any meaningful way.

So isn’t that

[00:15:33] Michael: because isn’t that because, if you’re updating your software, even say every six months, you’re never going to build up enough miles driven in those six months to even make a statistically significant assertion about anything. So it

[00:15:48] Fred: could be, nevertheless, nobody has shown that nobody has shown that what is being put on the roads now is any safer than what was being put on the roads five years ago.

Where is [00:16:00] that? That should be, there should be some indication of that. So that’s why I’m supporting exactly what you said about gaslighting, except expanding it to brownwashing. Anthony.

[00:16:09] Anthony: I approve your upgrade. So I just did some back of the envelope math and I assume, roughly I can guess, based on how my age and how long I’ve been driving, I’ve driven probably about a half a million miles in my life.

And that includes on highways, in snow, on ice on dirt roads, on off roading, on roads that don’t seem to exist. And I’ve had zero crashes. None. Not a single one. You don’t see me going around Hey, aren’t I great? Waymo, compare yourself to the average person. Cause the average person doesn’t get into crashes.

[00:16:47] Michael: Again. Yeah, you should knock on wood right now, first of all. Aw, damn it. Also, one thing that the data clearly shows is that Waymo vehicles are getting hit by humans a lot more than they are hitting [00:17:00] cars driven by humans. I think there were 20 or so severe crashes.

I think Waymo is track 16 or 18, but the author in Ars Technica found 20 crashes, and I believe the Waymos were only responsible for the collision in two of those. crashes. Yes, you could take that and say, yeah, we’re safer than humans, but really, you’re only safer than the humans that collided with the Wayman Waymo, which probably aren’t very safe drivers at all.

[00:17:29] Anthony: So Waymo through 2020, June 2024, has reported nearly 200 crashes. 23 were very severe, where they caused an injury or had an airbag to the beloy, or both. But, really, it’s just, unfortunately, this, I hate to say it, but it’s a little kind of we’ve done 20 million miles, aren’t we the best thing ever?

All the, most of these crashes happen at, one mile per hour or less, because they’re happening in a city at very slow rates of speed, and typically they’re happening at [00:18:00] intersections. Transcribed it’s a little listener, what I’m saying is if you’re listening to this, have you donated? No, that’s not what I’m saying right now.

What I’m saying is take all of this with a giant heaping of salt, unless you have high blood pressure, then use a sodium alternative.

[00:18:16] Michael: And Waymo is getting onto the freeways now, right? The next data update may not be quite as kind. I hope it goes well. I don’t want to see anyone get injured.

And I hope that they can work on highways at some point and at higher speeds and in more locations, that would be great, but I’m still somewhat skeptical, even though, I think it’s been pretty clear that I think Waymo is doing a better job of this than any of the other. Robo taxi and companies and fake robo taxi companies like Tesla.

[00:18:47] Anthony: But by doing a better job, you mean that they’re being more open?

[00:18:51] Michael: They’re being more open. They’re being more transparent about the process. They’re actually I don’t know exactly what they’re doing behind the [00:19:00] scenes as far as, some of the lobbying that we’ve seen from state to state passing some really bad policies.

Laws around autonomous vehicles. I’m sure they’re parts of part of that effort. But when it comes to, I think being careful and not overstating the the success of their technology, they, they aren’t throwing humans under the bus quite like cruise GM cruise did, humans are terrible drivers and that kind of nonsense.

That their approach has been more responsible and they’ve played it out over almost a decade now and slow played it in a manner that I think is a lot more responsible than some of the other companies that have come and gone in, in the meantime certainly more responsible than Tesla, which is basically lying its way through the process of pretending they’re building a robo taxi when we don’t see anything of the sort.

I’m gonna agree that it’s better, but it is still a long

[00:19:56] Fred: way from good.

[00:19:57] Anthony: There you go. That and I’m gonna [00:20:00] disagree and say they are saying humans are bad drivers, they’re just not saying it the way GM Cruise did. They’re saying, look how much better we are than Then human drivers, ignoring the fact that a lot of crashes are caused by drunk drivers and people who are bad drivers because they’re playing with their cell phone.

[00:20:16] Fred: And again, there’s a lot of serious limitations on this and yeah, you can look at many things Waymo is doing because they put it on their website, but it’s very carefully crafted. and not in a way that increases the insights into the underlying statistical data that they’re basing their conclusions upon.

So are we into gaslight illumination? Sure,

[00:20:39] Anthony: we can go. You can you wanna gaslight it up, Mr. Perkins?

[00:20:43] Fred: Sure.

[00:20:43] Kodiak Robotics and Autonomous Trucking

[00:20:43] Fred: I’ve got a new contender this week. Kodiak Robotics. Kodiak Robotics is primarily in the business of heavy trucks and automating heavy trucks for the highways, but happily for our purposes, they’re now getting into light [00:21:00] trucks as well.

Uh, let me just refer to them and their website, which at the very first line in their website says safely driving an autonomous future. So you think that safety would be something that they’re really concerned about. And so diving into it on their website, and all of this is based upon their website.

They talk about their revolutionary sensor pods and they show some pictures and they say that our revolutionary sensor pods use powerful cameras, LIDARs and radars to perceive any environment, keeping you moving forward safely and efficiently. Now. That’s interesting, but it’s important to note the sensors don’t perceive.

Perception is a, uh, what did I say? It’s a cognitive function. It’s not a sensory function. So there’s already starting off with a little bit of gaslighting and conflating things that really should be separated. So it’s not clear if that function is embedded in our sensor pods. [00:22:00] which is implied by what they say, or if it’s in a separate module, which would make sense because the sensor pods don’t look at the whole road.

They’ve got one on the right and one on the left, so there’s got to be some kind of function that integrates. Those two different points of view. And by the way, neither, neither of those have a field of view that includes it right in front of the truck where a pedestrian or child or any small object will be located.

So this is a lot of dead spots. No pun intended, no fun, no pun intended is correct. That’s what they do have a bunch of articles on there. One is from 2023 setting the stage for an autonomous future. And they say they introduced their first autonomous vehicle. A Ford F 150 updated with Kodak driver.

Maybe this will be coming soon to a light automotive venue near you.

[00:22:52] Michael: It looks like they’re primarily targeting like defense or government [00:23:00] contracts or security type for now. Yeah,

[00:23:03] Fred: because that’s where they got the money. But I think if they’re successful with this, they’ll certainly putting out in the public.

They talk about how easy it is, and then they say it’s taken in less than 6 months. They upfitted a new off road ready vehicle with a Kodiak driver, something 6 months. This is not exactly a walk in the park. So they’ve got no further word on the project success. Now, a year later, they talk about JB Hunt, Bridgestone and Kodiak surpassing 50, 000 autonomous long haul trucking miles.

And they then go on to say a two person team of Kodiak safety drivers. Overseas, the seamless and continuous operation. So that’s not exactly autonomous, right? That’s human supervised with 2 humans to take the place of 1. I’m not sure where the reconomics come from, and they don’t talk about the number of interventions or how the humans respond or if they’re sleeping or.

What’s [00:24:00] going on, but it’s, in fact, very few miles, 50, 000 miles of supervised computer driven operation, rather than strength of their autonomous systems. They talk about collaborating with Textron with defense pods and swappable and military, but they’re using the same core technology that’s embedded within Kodiak’s autonomous long haul trucks.

Which have been validated across nearly 3 million miles of real world operations with the human or not, I think it’s with human supervisors. So it’s really not autonomous, right? Michael, there’s nobody who’s running full autonomous trucks on the roads. Now, certainly not 3 million miles ago. No,

[00:24:40] Michael: certainly there.

Aurora is. Blaming, they’re going to start doing it in three months in December. So it’s very interesting. And

[00:24:47] Fred: then they go on to talk about, for example the market and how they’re, the market’s going to be increasing rapidly compared to what it does now. There’s no data yet because there’s no autonomous trucks yet.

So I’m [00:25:00] not sure where that’s coming from. They talk about well, then they go on saying in early 2024, we became the 1st autonomous truck company to build an autonomous or truck platform with all the redundancies required for safe driverless deployment. That sounds pretty good, but they don’t talk about where those requirements came from.

They’re not listed. They’re not referenced. So overall, safety is the 1st word on their home page. But their aspiration is not supported by the information they provide. There’s no data supporting autonomous operation safety claims. No evidence of independent validation of design or implementation or results of their safety systems.

And no evidence that there are less than overwhelming mileage accumulated on various implementations supports an aggregate safety claim. So I’m therefore calling this my gaslight candidate for the week.

[00:25:56] Michael: And it’s often fun to note that making safety claims on the [00:26:00] same web page as showing your technology being used on a tank, a machine of war is always somewhat of a contradiction.

[00:26:08] Anthony: They had me at using an adverb for the first word on their website. Sorry, you’re not passing seventh grade English. Michael, what’s your gaslight?

[00:26:20] Issues with Traffic Law Enforcement

[00:26:20] Michael: I’m diving into a problem we’ve discussed in the past, but there was a story in the Washington Post a couple of days back that, a 12 year old girl was hit by a car in an intersection by a vehicle that was running a red light. She was ultimately okay. She was injured but she survived and appears to be recovering just fine.

But they found out that the vehicle that, that hit her had 19, 770 worth of unpaid tickets. And almost all these tickets were speeding and red light camera, all camera tickets. I believe the person who owned the vehicle was a resident of [00:27:00] Maryland. They, so they didn’t have to pay the tickets.

There’s not a reciprocity agreement between Maryland and DC or DC and Virginia. So a significant portion of the cars commuting in DC that are receiving these tickets while they might have to pay them and collections could come after them in some way or another at some point. They can’t they don’t affect the driving status of that individual in their state.

So a Maryland driver is not going to get in trouble or lose points off their license for having racking up, a hundred speeding tickets in DC. They simply don’t count. And, reading the story I, I first wanted to give the gas light to the guy from Maryland who was driving the vehicle and kind of Pretended that he had nothing to do with it.

And basically saying, you can’t prove that I was driving when I got all those other tickets and my traffic record of the vehicles irrelevant to this, even though, he had four tickets for running red [00:28:00] lights since July ran this red light and almost hit this kid. He’s. Trying clearly, I wanted to give him my gas light for that.

Anthony pointed out that you should give the gas light to the D. C. Department of Motor Vehicles because they’re failing here. And, thinking about that they’re not the ones that are responsible for this problem because they don’t really have the authority to force Virginia and Maryland to take into account.

the camera enforcement tickets from DC. So ultimately I’d have to settle on there was a court that ultimately started part of this problem. They ruled that they ruled that essentially those camera tickets are not going to be able to be used against Virginia and Maryland drivers or even DC drivers in some circumstances.

So it’s a problem here, of, Enforcement and courts. And it’s something that needs to be solved immediately. I believe there’s a, an act called the steer act that’s moving or has just been passed in DC, that’s going to address some of these [00:29:00] things, but ultimately, when you don’t enforce the speeding and red light tickets, you’re going to end up in this situation where those.

It goes back to our discussion about Vision Zero. If you’re not actually enforcing and implementing these measures and you’re just paying lip service to a Vision Zero campaign, it’s not going to result in reduced fatalities and injuries on the road. And you’re seeing that play out in D. C.

where they’re. They’ve been essentially restricted by the courts from enforcing their own traffic laws. And then that’s a problem. So I’m going to give the gas light of the week to the courts who are causing, who are essentially causing this problem. And also to the states of Virginia and Maryland, who.

should acknowledge the DC tags. We need reciprocity in order for, safety to work in the DC area given, the proximity of Maryland, Virginia and the district.

[00:29:54] Anthony: It seems that even if you’re, have DC tags, they’re not going to do anything to you either. From the [00:30:00] Washington Post article, a DC police spokesman said, Officers are not responsible for enforcing camera tickets, the Department of Public Works is, and generally don’t have the equipment to do it.

[00:30:10] Speed Cameras: Revenue Mechanism or Safety Tool?

[00:30:10] Anthony: So everyone’s it’s not my job, it’s not my job, why spend all this money on speeding, the speed cameras, if they’re just there for show. I don’t get it.

[00:30:19] Michael: It plays right into the argument of people that are completely against speed cameras, that they’re simply a revenue mechanism.

Because essentially here you’re not enforcing, you’re not using them for safety enforcement. You’re really just collecting money and going after people in collections. And it’s there to make the district money versus being there as it should be to try to. Change behaviors and, provide disincentives for people who are not following the rules of the road.

And it’s just not doing that properly.

[00:30:52] Fred: I’m going to, I’m going to argue with you a little bit, because when I was driving into Washington, I did get hit with one of those automatic [00:31:00] speeding tickets one time, and it was a hundred bucks. And then I was a lot more conscious of driving slowly, at least on those areas that have the speed camera.

So I, I do think they’re effective in reducing the overall speed where they’re located. I think studies have supported that as well, but yeah. I’m going to certainly agree with you overall that there’s something wrong here and there’s someone who really needs to get their gas light extinguished.

And I think it’s interesting though, it’s oddly similar to the 1st story we talked about of people learning rapidly where the limits are and. Rapidly learning how to work around those limits for these people who are scofflaws once they find out that there’s no consequences for being bad they just get worse.

[00:31:51] Michael: And that would be Mr Curtis in the article who wrapped up 94 unpaid tickets and they don’t seem to be changing his driving behavior. So [00:32:00] he’s the opposite of what you’re describing as how you responded to the situation.

[00:32:04] Anthony: Yeah. It seems there’s no enforcement at all.

[00:32:06] DC’s Struggle with Unpaid Traffic Tickets

[00:32:06] Anthony: Cause I found an article from last year from May of 2023 titled DC struggles to rein in risky drivers.

One car has 186, 000 in tickets. From the article, more than 2, 100 vehicles have at least 40 outstanding tickets according to data from the DC DMV and about 1, 200 cars are linked to fines exceeding 20, 000 for the past five years. Topping the list of offenders is a car with Maryland tags that has 339 outstanding tickets worth more than 186, 000 in fines and penalties.

In all, more than 6. 2 million traffic tickets. Totaling nearly 1. 3 billion dollars in fines and penalties have not been paid to D. C. since January 1st, 2000. This is I think a speed camera is a good idea, but it seems there’s no penalty for [00:33:00] not paying them. And it upsets me that I moved away from the District of Columbia because somebody who uses my car sometimes gets speed camera tickets and I pay them like a chump!

But I think New York actually enforces them.

[00:33:10] Michael: Yes.

[00:33:10] Anthony: Why I’m literally at a loss for words. D. C., why can’t you enforce? Oh, they don’t have home rule and it’s not really a state. It’s just a little colony. And that’s a different podcast.

[00:33:22] Michael: It’s complicated.

[00:33:24] Anthony: Oh, D. C. We feel bad for you.

[00:33:27] Fred: Wouldn’t D. C. have to go through the national legislature to get an agreement signed that binds the drivers in Virginia and Maryland to D. C. law? It’s incredibly complicated. Doesn’t it literally require an act of Congress to get something like that done by the D. C. government? There’s,

[00:33:47] Michael: without diving too deeply into the history of D.

C., they can pass laws, through the city council, and but there is some type of oversight that can be exercised [00:34:00] by Congress in certain circumstances, but it’s, it’s I think that the STEER Act, which they passed, is going to take effect without being without Congress messing with it.

So hopefully they can get this done without going to Congress because I don’t see Congress moving on this issue because they don’t move on a lot of safety issues.

[00:34:20] Anthony: I think they shouldn’t call it the STEER Act. I think somebody on the D. C. Housing or Homeland Committee in Congress will be like, oh, why are they going after my STEERS?

I vote against this.

[00:34:32] The STEER Act and Roadside Safety

[00:34:32] Anthony: Anyway, not all government is bad, folks. There’s, in fact, Congressman Rudy Yakum? I’m gonna say Is that how I say it? We’ll say Congressman Yakum? Yakum. Y A K Y M. I apologize for mangling your name, sir. Or ma’am? Yes, sir. He introduced a bill called Preventing Roadside Death Acts.

This legislation would require the DOT to collect, analyze, compile, and publish accurate, detailed data on disabled roadside [00:35:00] vehicle accidents and develop a strategic plan to stop these accidents. Sounds like a great idea. I’m all for it.

[00:35:08] Michael: Yeah, it is a good idea. I think mainly because it’s focused on enhancing the quality of the data that’s available on roadside crashes, in the early 2000s, we were tracking a lot of fire related collisions with Ford Crown Victoria vehicles, which As our listeners probably know are used in a lot of police, the Crown Victoria police interceptor is used in a lot of for police cars.

And they are inevitably parked along the side of the road after making a traffic stop and we, we see. It’s just thousands and thousands of collisions in America every year where there’s a car on the side of the road and an antenna driver or a drunk driver or distracted driver or a bad driver veers off the [00:36:00] road and onto the shoulder and strikes the rear or the side or another, area of the vehicle that’s parked on the side of the road.

It’s a very powerful collisions. Typically. And in the Crown Victorias at the time, they were, they had a fuel tank that was located behind the frame rail in the vehicle, which is always a problem. And they were, exploding, catching on fire, killing police officers and other folks in the process.

And. When we went to the, where we’d usually go, the fatality analysis reporting system or FARS that, that the DOT has, it’s basically a collection of all police reports on fatal crashes every year, we found that there weren’t. A lot of these crashes were missing because the Crown Victorias were parked on the side of the road, they were not moving and in many situations, they weren’t considered a vehicle in transit, which is a definition that is strictly enforced around what type of accidents or excuse me, what type of crashes are [00:37:00] collected in the federal database.

There’s a data problem there. I, I know that there have been steps taken since. The early 2000s to fix that problem and to include vehicles that are on the roadway, but not actively moving, they’re involved in crashes to be included in the national data. But I think this bill takes that a step further and also, looks at pedestrians who are on the side of the road or who may be fixing a vehicle.

And that’s something that’s really necessary. The data on these types of collisions is pretty And what this bill essentially does is tries to make better data happen in this area. And so we fully support this bill.

[00:37:40] Anthony: Great. And maybe Waymo can get it and they can massage it and say how they made it, made everything good now.

[00:37:47] Understanding Four-Wheel Drive Mechanics

[00:37:47] Anthony: Let’s talk I’ll save the, my, my question around recalls to our recall roundup section for now, how about let’s let’s Fred, can you explain to us, in your Tao of Fred, or Tao, we still haven’t decided, why [00:38:00] do four wheel drivers, why does Anthony lack the ability to speak today?

[00:38:05] Michael: You’ve now entered the Tao of Fred.

[00:38:08] Fred: I can’t address that, but I can talk about four wheel drive. Somebody memorably, that four wheel drive is God’s way of making sure people get stuck in deeper snow. Without getting all theological about that how can that be? So why do tires slip? Tires slip, a little physics here, because the torque applied causes too much force and the weight of the Vehicle on that particular tire times the coefficient of friction determines how much force the tire intersection with the road can sustain before it starts to slip.

Okay, so that makes obvious sense when you think about it. If the road is really slippery. The tire is going to spin more easily, [00:39:00] right? Calculating exactly how that works is very difficult, but in relative terms the tires will slip about a factor of 5 between dry pavement and wet pavement or icy pavement, I should say.

But the tires will slip when the coefficient of friction drops and the torque increases. So with four wheel drive, what you basically are doing is you’re taking the torque of one wheel and you’re distributing it over four wheels, right? Or maybe you have two wheels if you have an anti slip differential on your two wheel drive car, but basically you’re dividing that torque among different tires.

So the individual amount of weight on each tire versus the amount of torque needed to make that slip moves in a favorable direction, right? So the torque goes down. If the force remains the same for each tire to slip, then you [00:40:00] start to move. And people know this intuitively, they put sandbags over the rear axle, for example, in a pickup truck to make it more stable in snow.

But what happens then is that people start, the car starts to move. All of a sudden, and it’s no longer slipping. So they get overconfident and they start to go down the road and they think that because they’ve got the 4 wheel drive and the road is slippery. They’re good to go. But what happens as the speed increases is that.

The inertial forces start to dominate and so if you think of it as the car is going fast and you turn the wheel a little bit to the left or right, all of a sudden, there’s a big sideways force because your vehicle is heavy and that sideways force wasn’t there when you were stuck in the snow before.

And so you’ve, you’re assuming mentally. The things are still copacetic because you [00:41:00] got out of the sticky situation once and now you’re going too fast down the road and what happens is the inertial forces take over because the car is going faster and you turn the wheel and all of a sudden you break traction on one wheel because the coefficient of friction is different going sideways to going forward and in reverse, right?

And so when that happens, when you lose traction on one wheel. All of a sudden, the other ones follow right along behind it. So it’s very difficult to regain control of your vehicle at high speed, because all four of your tires are going sideways now and what you had mentally. Accommodated at lower speed with the tires moving, predictably in the snow, and then the automotive engineers would call that control of tracking of the vehicle.

You lose when the vehicle goes faster because instead of the work from the engine [00:42:00] dominating it, all of a sudden, the initial forces associated with the motion of the car start to dominate and you end up going sideways. Losing traction and ending up in a deeper snowbank than if you were just traveling along at a low speed.

This actually goes back to what we were talking about before about the way most traveling at slow speed versus high speed and just being a lot safer at slow speeds. That’s true for humans as well. Particularly it’s true when the roads are icy. And another factor at work is of course on a slushy road, the weight of the tire can cause the slush to compress and turn it to glare ice under your wheels.

So there’s a lot of dynamic effects going on that are all very difficult to calculate, but the calculations that have been done in relative terms are pretty straightforward. Slippery pavement, icy pavement is about only one fifth as much traction available as a dry pavement. If you think of that, you [00:43:00] really need to go slower to have the same amount of safety, even when using four wheel drive, because of the slippery road conditions where you’re engaging the four wheel drive than if you were operating on a dry pavement.

So what does that resolve to? In order to avoid getting stuck in deeper snow, just be aware of the conditions deteriorating and the fact that the car can go sideways unexpectedly, because the inertial forces associated with the car moving faster are very different than the forces that you encountered when you were simply trying to get moving from a, a slippery situation when you invoked four wheel drive in the first place.

Does that make sense? Is that too convoluted? Were you there on that?

[00:43:50] Anthony: What was the question? No, it makes sense. I think it’s the myth that, I, I remember saying this a lot in D. C. that people couldn’t drive in snow because they would do the classic mistake of, there’d [00:44:00] be like the smallest amount of slush, something like that, and the response is if I gun it, and their wheels would just spin, and people think I got four wheel drive, I’m good, I’ll just gun it, and it’s the same effect is what you’re saying.

[00:44:12] Fred: Yeah actually, gunning it does work, because what happens when you gun it is, you generate a lot of friction just due to the motion, and you melt the snow. And ice underneath your wheel. But then of course, that just gets you into the next patch of ice and you have to keep on spinning in order to do that.

So you can make very slow progression if you spin the wheels, you need to spin them long enough to wear things out before you make any real progress, but the four wheel drive. Yeah it’s better. And the all wheel drive is better. I enjoyed being in my Subaru driving around. Buses that were stuck in the snow in Washington made me feel quite superior, but very slow speed, so it worked.

[00:44:55] Anthony: But I hear that Elon Musk has invented a laser system that’s gonna, it’s already installed on all [00:45:00] the Teslas that are ready to go. They just have to pay a fee. And the laser beams shoot out in front of the tires and melt everything on the road and everything’s perfect all the time. And jelly beans come out of the car too.

It’s weird.

[00:45:11] Fred: There was, of course, the Jewish space lasers that were invoked by legislators. And they may be useful for melting the snow in front of the, in front of the Teslas. I’m not sure exactly how they’re being deployed.

[00:45:22] Anthony: I don’t know either, but I, hey, listeners, if you know how they’re being deployed please listen to a different podcast.

Please go talk to a family member, get some help. And while you’re talking to that family member, Tell them to go to autosafety. org and click on the donate button and donate again. Donate like you believe in conspiracy theories. That’s how much you should donate. We haven’t committed to any, we haven’t taken part in any conspiracy theories in at least 17 days.

[00:45:48] Recall Roundup: Who Pays for Recalls?

[00:45:48] Anthony: So now let’s go into some recall roundups. Strap

[00:45:51] Michael: in. Time for the recall roundup. Michael,

[00:45:53] Anthony: I got a question for you. Let’s say I buy a car. Encino. And it gets a recall on it. [00:46:00] Say like my car, my Toyota, it has a recall. Um, who pays for that recall?

[00:46:07] Michael: So there are a couple of ways recalls remedies work.

You can, a manufacturer can, Bring the vehicles in for a parts replacement, inspection, that type of thing. The manufacturers have an option to, simply buy your vehicle back. In some cases, we see some recalls where the manufacturer says, yeah, these are so screwed up. It’s not even worth us paying to repair them.

We’re just going to trash them and give customers their money back. So that, that’s typically how it works, but in, Any case, whether it’s a repair or replacement, inspection and repair, software update, whatever it is, the manufacturer is 100 percent responsible for taking care of the costs associated not only with the part, but also with the repair and the labor involved.

[00:46:56] Anthony: And this is something legislated at the federal level? [00:47:00]

[00:47:00] Michael: Yes. And this is something that has been required. I, I don’t believe it was the free repair and the remedy sections were required right when the safety act was passed in the late sixties, but they were added relatively soon after because manufacturers and dealers were, as they do sometimes jerking consumers around on the repair costs for recalls.

And the government saw a urgent need to make sure that they were, that none of that took place and that these were just. Repairs or any type of remedy were carried out immediately by the manufacturer at the manufacturer’s cost because recalls are ultimately the manufacturer’s fault.

[00:47:38] Anthony: Okay. What happens if I buy a car from my buddy named Heinrich?

And he has a little car called the Fisker Ocean. And then Heinrich’s starting an auto company is really expensive. And goes, I’ll run a business but I sold you junk, really expensive junk that has recalls on it. [00:48:00] Who pays for it then?

[00:48:02] Michael: Fisker is about to discover that they’re going to be on the hook for those, even though they are now telling all of their owners that their owners are going to be responsible for paying for the cost, I believe it’s the labor involved.

What Fisker does is, essentially telling its owners is we’ll give you the parts for your recall repair, but you’re going to have to find the service place to do it. And you’re going to have to find the money to pay for the labor at that place. Unfortunately for Fisker, they need better lawyers or something over there.

That’s not going to work. That’s a violation of federal law. And, I can’t even believe it’s gotten this far and they haven’t been called out on an enforced, at least warned by NHTSA. It’s a, that this that their interpretation of how the situation is going to work is not going to work. And even, Congress has been fairly aware of this issue as well.

We had the case of Takata starting in 2014 and when [00:49:00] Takata went bankrupt and essentially, that really messes up the. The chance for a recall repair for a lot of customers, if a company is no longer solvent and doesn’t have, repair facilities, replacement airbags coming in and all of that.

So Congress said, look, even if you are in bankruptcy and your company is failing. You still have a hundred percent obligation to, to perform recall repairs. There’s really no distinguishing between whether your company is bankrupt or not. You built the car, you put it on the road and sold it.

You have a responsibility to pay for all the recall repairs. So I’m not sure what Fisker thinks it’s doing here. But they are very incorrect and they’re, if they proceed on this path, they’re going to be in violation of federal law and they’re going to get in trouble. So So hopefully he’s at Henrik Heinrich listens to podcasts and can save his company from even more embarrassment in this area.[00:50:00]

[00:50:02] Anthony: Fred’s very animated right now, but he pressed the mute button. So I’m going to assume this is what he said. Anthony is very good looking man. And I would like to buy him a new car. Is

[00:50:12] Fred: that correct? I think you need a new hat. No, I, what I was actually going to say is that How much trouble can they be in?

Because they’re in chapter 11 liquidation. Presumably the executives have already walked away with as much money as they possibly can, leaving the consumers holding the bag. So how much trouble could they actually be in? You can’t get blood from a turnip.

[00:50:35] Michael: They’re required in that, that law that Congress passed requires The recall obligations of a manufacturer to be placed very high up in the chain of Their creditors in a bankruptcy So that’s one Thing that they’re going.

I mean they essentially are going to have to take care of it either set aside like I think in the Takata case, there was a set aside [00:51:00] of a large amount of money that was intended to cover those recall obligations at a later date, and in this case, I think they’ll have to do the same thing in bankruptcy and set that aside.

But additionally, there could be civil penalties levied against the company that would also appear very high up in their bankruptcy. Simply for the violation of federal law there because it’s a violation of the notification and remedy provisions of the safety act they could be in trouble to an extent I don’t think anyone’s going to jail and the bankruptcy certainly complicates things but if you bought a fisker were probably you probably knew going in that it you know It was a somewhat risky venture and if you You weren’t doing your homework.

[00:51:41] Anthony: Yeah, so I apologize if you’re one of those people who own one of the 2023 to 2024 Fisker Oceans. That have five open recalls on them I know you have the money to pay for anything you want, but you don’t have to pay for this recall. Just, leave the car outside, make it into a piece of art, [00:52:00] do something else with it, just don’t drive it.

It’s, they’re incredibly unsafe. Stay away from them, or you can go and visit, you can send them off to Heinrich Fisker’s address, and I have it right in front of me, it’s 3212 Hideaway Lane, Jerk Face, California. Who knew? No, let’s go into some real recalls now. Mazda, 672 vehicles, that’s an adorable amount.

The 2024 Mazda it says it’s the Mazda 3 the CX 30, CX 50, The CX 30, CX 50. And it has a problem with their high beam control. Their headlights will not automatically adjust for approaching vehicles.

[00:52:38] Michael: Yeah. So there’s a problem, essentially the problem starts with the forward sensing camera on the vehicle.

And the far sensing camera is having, when they ship them out of the factory, they never changed the mode or the setting in the camera to normal mode from. [00:53:00] Production mode, which apparently they leave it in during production. And so these cameras are on vehicles operating on the road, but they are still in production mode.

They’re not in the mode that they’re supposed to be in when the vehicle’s out on the road being driven. And so what the camera does is causes some kind of malfunction happens. And it messes with your high beam control system that automatically adjusts headlights from high to low beam. When you’re approaching vehicles, that’s going to make other drivers mad.

It messes with your automatic emergency braking system, which may, which will not activate if you’re in an emergency situation. That’s really important. And it also prevents lane keeping from working properly. So that’s why you need to make sure that all the T’s are crossed and I’s are dotted before you let a vehicle leave the factory.

And, you make sure it’s not still in production mode when you put it out there. Mazda is going to have. Owners being notified, I think in early November. So if you’re a [00:54:00] Mazda owner, be on the lookout for this one. And in the meantime, drive with the understanding that you may not have automatic emergency braking.

So keep your eyes on the road.

[00:54:12] Anthony: And actually that’s one of my favorite automated features, not the automatic emergency braking. I haven’t tested that one yet, but the automatic head beam, high beam adjustment. That was so cool. I came across that on a rental car and I was like, what just happened? Oh That was great Then I wanted to stop every car that was passing me and be like stop with your high beams on you’re killing me My car’s smart enough to adjust why isn’t yours and no one would flag me down Anyway no more recalls.

[00:54:41] VinFast Investigation and Certification Issues

[00:54:41] Anthony: We have an investigation though. VinFast Auto. For those who are not familiar with VinFast, VinFast is a Vietnamese auto manufacturer. That is not as tasty as the real thing, as the best I can say. This is a preliminary investigation for their lane keeping assist system. The [00:55:00] complaints allege, they have 14 complaints?

Yeah, they have 14 complaints. For, they probably only sold 15 of these cars in the U S

[00:55:08] Michael: yeah, I don’t know if they’ve sold, they’re claiming there are 3000 vehicles involved here, but I don’t know that VinFast has even gotten that many vehicles on the road yet. I know that they shipped a thousand to America and then we’re scrambling to get them compliant with the federal motor vehicle safety standards.

And we’re dishing them out one by one as they completed them. And there’ve been, there are just a lot of concerns. We’ve had a lot of concerns around these vehicles and how they just the way there, there aren’t any other auto companies that we know of that shipped the vehicles to America and then, retrofit them to meet federal standards after they get to the country.

That’s just not how traditional auto manufacturers do it. Do it. It doesn’t seem to be a very safe way to do things. And it seems to suggest that VinFast doesn’t have the full respect for federal [00:56:00] motor vehicle safety standards that it should. So we’ve been watching them and now thankfully NHTSA is taking a look at their operations.

This is one of a number of other concerns that we’ve had. And I think we’ve covered a couple of them in the past around VinFast.

[00:56:15] Anthony: So if you’re in the market for a VinFast. Don’t.

[00:56:18] Fred: Yeah. We should point out that there’s a systemic problem in America versus Europe in particular because America does not require a type certificate for a car before it goes on the road.

It relies on the manufacturers to self-certify versus Europe where. People are required to show that the vehicle is safe before they’re allowed to put it on the road through a type certificate program. So, yeah, VinFast did a lousy job, but the system that allows VinFast to do a lousy job is also defective.

And really, even though this is aspirational, it really should be corrected by legislation that allows us. Allows Americans to put type [00:57:00] certificates in place for cars before they use this, by the way, this is my opinion, not the institutional opinion of the center for auto safety, but come on, folks, let’s get it right.

[00:57:10] Anthony: If you’re a little confused about why self certification is bad. See Boeing. And with that

[00:57:19] Michael: Actually, that’s not correct.

[00:57:20] Anthony: Self certify a lot of their parts. Thanks.

[00:57:22] Michael: No, they have a different scheme that they use, the fa that’s more like type certification. We actually, I think wish Nitso would follow that scheme that the FAA uses, but as the last few years have borne out, that system isn’t perfect either.

[00:57:37] Fred: Yeah. The Boeing situation’s different because Boeing pays. The inspectors that are supposed to be independent representatives for the FAA. So it’s a weird, bizarre, unsupportable situation where Boeing is supposed to be controlling the income of the people who are inspecting it for independent certification by the government.[00:58:00]

That’s never going to work.

[00:58:01] Anthony: Okay,

[00:58:02] Fred: I want to go back.

[00:58:03] Tesla Semi Fire: A Fiery Concern

[00:58:03] Fred: I want to go back to something we skipped over a Tesla semi fire that happened a few weeks ago. Tesla is getting into the heavy truck business. And so they’ve got this massive truck that they put out on the highway, which unfortunately went off the road.

The driver survived the driver. I think he was mildly injured, but I just want to talk about what happened. The batteries apparently ruptured and due to the going off the road and they started to burn and it took 50, 000 gallons of water to extinguish. It sounds like a big number, and to put it in perspective due to the magic of long division, I figured out that 50, 000 gallons is the equivalent of about 167 fire trucks.

It’s also important to note that even though the article said it extinguished the fire, the truth is, no, it did not, because the lithium battery fires are hypergolic. [00:59:00] They’re self sustaining. They don’t require any outside inputs like fuel or Temperature or air, the normal components of a fire. And so what the water that they squirted on it actually did was it suppressed the evolution of the fire.

And the secondary fires associated with the expulsion of combustible products from inside the batteries, so that it lit as few things in the neighborhood as possible on fire. Hopefully, it also suppressed the production of organophosphates, which are basically nerve poisons that have been detected in lithium ion fires.

And but, who really knows about that? Anyway, I just wanted to mention that if you happen to be near a lithium ion powered heavy truck, please make sure that there are 167 fire trucks nearby so that you don’t suffer any jeopardy from the fire that’s going to be coming [01:00:00] out of the batteries. A different.

[01:00:03] Michael: Yeah. And it looks like all that wonderful water that was used to extinguish that fire just washed off into the nearest forest. So I don’t know what’s in that water. I know it’s probably not good for anything around it.

[01:00:16] Anthony: Yeah. I also I want to point out that the fire departments had to dump water around the surrounding area because this fire happened in the state of California.

Yeah, that is, quite fire prone. So your next state file that I’m going to just done, I’m done.

[01:00:33] Closing Remarks and Listener Appreciation

[01:00:33] Anthony: Thank you for listening. Thank you for donating. Thank you for giving us five star ratings, telling all of your friends, getting them to subscribe and and floss, and all the things, the, all the things.

Thank you. Are we up to 14 million listeners yet?

[01:00:48] Michael: Yes, we just hit that number.

[01:00:50] Anthony: Oh man, I’m so

[01:00:51] Michael: happy about that.

[01:00:52] Anthony: Thank you. Yeah, but we’ve just lost every employee of VinFast. Until next time, bye.

[01:00:57] Michael: Bye bye. For more [01:01:00] information, visit www. autosafety. org.