Odometer rollback and automated driving fail

p>Fred: Well, good morning. Good morning, gentlemen. Good morning world. Hey everyone.

Anthony: All right. Let’s start off with our friends at Ford So I’ve talked about how I like the idea of Ford’s blue cruise and GM Super Cruise where it’s and I’m ready for this Guys, I’m gonna use correct terminology It’s operating design domain restricts the ability of this partially automated driving solution onto pre mapped highways Pretty goddamn good, right?

Michael: I like that. You got, you, you got it right. It’s operating design domain is restricted to a particular geographic areas, basically interstates and divided large divided highways.

Anthony: So what Ford and GM, their approach has been on like a Tesla, which is just like, yeah, our cars drive themselves. Ford and GM are like, nah, that’s not what happens.

We’re going to restrict where you can use this and whatnot. But unfortunately now we have. An accident the NTSB is investigating a role of Ford’s Blue Cruise in a crash. The NTSB has opened an investigation into Ford’s Blue Cruise’s hand free driver assist system. The probe was spurred by a recent fatal crash that involved a Ford Mustang Mach E equipped with Blue Cruise.

The accident occurred on February 24th in San Antonio, Texas. The initial report states the Mach E struck the rear end of a stationary Honda CR V on Highway 10. Unfortunately, the CR V driver later died.

Michael: It looks like the CRV, at least according to the driver before, the CRV was completely stopped in the middle of the road on an interstate with no lights on, which, um, that’s a very difficult situation to navigate as a driver without any of this.

Partial automation going on in the vehicle. I mean, it’s coming up onto a vehicle at nighttime. I think this was around 9 50 PM local time. At high speeds on the interstate is going to be very difficult for a driver to respond and react to and to avoid that situation. It does appear just looking at some of the footage from the post crash scenario that was on the, that was posted by one of the local news stations.

It looks like there are, was some lighting in the area. There were some overhead street light from that portion of the interstate. But, there’s still a lot of questions that are unanswered here was the driver of the Ford, distracted in a way that, prevented the driver from recovering control of the vehicle or taking over in time to avoid the crash?

Were there warnings? Did the system provide any warnings to the driver? Is this a situation that it’s even possible to provide warnings to the driver in time to avoid a collision? There’s a lot of questions there that, remain. Unanswered and both the NTSB is investigating the crash as well as NHTSA’s special crash investigations team as well to try to get at some of these, some of the answers to those questions.

Anthony: This seems like a common problem though with these systems where it’s They, I mean, Tesla runs into this problem all the time where if they keep hitting things that are static, they keep hitting things that are parked. And I don’t understand this because they’re all equipped, well, okay, the Ford and the, is equipped with a radar, so it should be able to keep pinging and saying, hey, this object in front of me keeps getting closer and it’s not moving.

I, I, what’s the, is there some sort of technological difficulty, Fred, can you guess that, that these systems have such a problem with a static item in the middle of the road?

Fred: I can guess, but it’s only a guess. And my guess is that the system is tuned so that it ignores stationary objects. So if you’re driving your car, you don’t want the car to stop when it sees the guardrail, right?

You don’t want the car to stop when it sees the fire hydrant next to the road. So there’s a prejudice in the software to ignore stationary objects. When you’re driving. You’re encountering lots and lots of stationary objects, right? So the car has to be able to discriminate between these very unusually occurring or rarely occurring stationary objects that are going to kill somebody versus the preponderance of stationary objects that are supposed to be there and are completely benign.

So I, my guess is that it’s just really damn hard for the logic to do it. I think there’s a larger problem in all of these situations though, and you’ve been in the software industry, Anthony, and what’s the tool that people use to develop software in general?

Anthony: You mean like an IDE, an integrated development environment?

Fred: Yeah. Or, and a software development kit, right? Okay. An SDK,

Anthony: yeah, sure.

Fred: So you’ve got an SDK, and everybody uses these. The problem with cars is that SDKs are basically Designed for one dimensional problems, right? So you’ve got an input and you’re looking for a single output and you know Even that is difficult enough to do as anybody who’s ever had a blue screen of death or a problem on the internet Knows quite well The problem with the cars is you’ve got a very difficult problem at the beginning Which is the perception of the object then you have another Very difficult problem, which is to take all of that digitally or digitized information and turn it into a perception of what the object is.

All right. So that’s already a very difficult problem. Then you’ve got to translate that. Those are two difficult problems into a command for the car or a series of commands because the car can do many things, right? It can turn right. It can turn left. It can slow down. It can accelerate. So you don’t have a.

a single dimensional problem. You’ve got an inherently multi dimensional problem. By analogy, if you can solve a 2D problem like driving, right, which is basically left, right, forward, and back, that doesn’t qualify you to fly an airplane, which is a fundamentally a 3D problem, right? So the software development kits aren’t really up to it.

And in order to have this car operate in spaces that people occupy, you’ve also got to have an overlay. of ethical and legal considerations that are appropriate for the vehicle and the driving habits in the area. This is a really difficult problem. Nobody has a software development kit that includes all these parameters.

And I suspect that there is a shortage of philosophers and legal experts in the engineering development. That is coming up with the algorithms and vetting the algorithms. That’s long winded, I know, but I think that’s a fundamental problem that the software world just is not up to the complexity of this problem with any kind of tools that these poor engineers are trying to use to manage self driving vehicles.

End of rant.

Anthony: Okay. So this is this is still an artificial intelligence problem dating back to the 1980s. I think when the U. S. Navy was using sonar and running it through pattern recognition to say, Hey, is that an enemy or is that a rock? And

Fred: That’s a similar problem. Wow. Okay. I mean, if you’ve got a submarine on the bottom of the ocean, right, how do you know that’s a submarine, not a dead whale?

Difficult.

Michael: So there are a couple other things on this, crash just notes that I wanted to make. One is, Ford Blue Cruise itself is not under broad investigation here. This is a crash specific investigation by the National Highway Traffic Safety Administration. The NTSB maybe is looking at this in the context of They’re broader concerns about partial automation.

We know and we know from, last week we talked about the insurance institute study of partially automated vehicle and every one but one out of 14 failed. So there’s some serious concerns here about, whether this is a Whether there, if there is no real safety benefit being confirmed by this technology, if it’s not bringing safety benefits to our roads, which so far there’s no evidence to suggest that it is, but at the same time, if, if it turns out that this crash was caused because the driver was otherwise occupied with something else going on the vehicle, not paying attention to the road, because.

The driver was relying on the Ford blue cruise to conduct that part of the driving task. Then, that’s a problem where we’re installing something into a vehicle that, you know, and I’ll say it again, it’s creating attention traps it’s luring drivers into a false sense of security. And then when a vehicle appears on the road ahead of the driver, whether or not the system picked it up or not, there’s not time to react because the driver’s not paying attention.

Paying attention to the road and not engage in the driving task. So it brings up the question for me, as we should, would it be installing this technology at all in vehicles with when there’s no proven safety benefit to the technology. And right now it’s simply sir, servicing as a creature comfort or convenience for drivers who are tired of the rather hard task of driving a vehicle.

Fred: I’m going to, I’m going to suggest there’s a related problem, which is that the A. V. companies have been getting permission to test their vehicles on public highways by saying, well, we need to get this data to improve things and make things safer, which has a certain amount of makes a certain amount of sense.

But the problem that I see is that there are numerous iterations of the software, numerous versions being put out in public and. There is no evidence that I’ve seen that these successive versions of the software are making things any safer. They’ve been at this for several years now. There should be a clear record that says we started with version X.

Now we’re at version X plus one.

Anthony: Oh, Fred’s internet’s crapping out on us.

Michael: Fred’s lost it.

Anthony: The

Fred: rationale that we’ve got to say that much safer than, so there should be data now that says the new software is better and safer than the old software. There should be, they should be able to track the progression of safety with these successive versions of software that they’ve now been using for many years.

I haven’t seen any of this data. I don’t think anybody else has either. Where is the evidence that giving people access to the roads and exposing the public to the hazards of EVs are in fact resulting in safer EVs?

Anthony: Sorry, A Vs, not E Vs. A

Fred: Vs, E Vs,

Michael: sorry.

Fred: Yeah,

Michael: you’re

Fred: right.

Anthony: Okay.

Michael: And then, going along with that, there’s a continued question that I don’t know that anyone’s answered effectively yet, for me at least.

We looked maybe a year ago at the crash where the tire comes off the vehicle next to a Kia Soul, and that vehicle is upside down, 15 feet high in the air, less than a second later, 15 Is there ever going to be a system that can. Pick up threats like that, alert the driver and expect the driver to take over and safely avoid the situation.

I don’t know if the technology could ever do that. I mean, it’s such a quick, there, there’s less than a second between the time a hazard is identified and the, the vehicle needs to be begin avoidance maneuvers immediately, and if the vehicle is attempting to tell the driver, rely on the driver, switch over to the driver’s control during that moment.

That very small space of time, it makes me wonder if it’s ever going to be possible for this type of scenario where there’s, a disabled dark vehicle that you can’t see on the road ahead of you. That’s coming up within, a split seconds. So it’s a concern overall as to, Whether even with some of the best driver assistance technology, are we going to be able to effectively avoid certain types of crashes that occurred very quickly?

Anthony: Well, I think Fred made a good point talking about is they should be able to track has safety improved over time with these different software releases. I think one of the problems. That they have to do is actually define what safety is, and define what that means. I mean, we can, sure, we can do it in broad terms, but like for all these different scenarios that come up, I think that’s such a hard problem, as to figure out, like, how do you react in this situation?

How do you react in that situation? It, we can say vaguely, like, don’t injure anybody. Okay, but, I mean, there’s so many variables.

Fred: Oh, yeah. And further to that, Anthony, there’s no recognized standard for even the metrics that need to be recorded by companies to determine whether or not the car is what the safety baseline is, nor how it’s improving.

Every company is now free to set up their own metrics and deal with them however they want to. There is no public reporting of safety metrics. In part because there is no accepted standard for what those safety metrics ought to be. That is under development, and there was recently a published document by SAE, but I don’t think any company has been using that yet, and it makes the point that nobody has used such standard metrics previously.

So there’s no way to compare one company’s development against another. Right.

Anthony: Hey, if any of these autonomous vehicle systems and companies are out there listening, I would gladly put on a philosopher hat and put a pipe in my mouth, put some leather patches on a corduroy jacket, and and charge you a lot of money to go let me think about that.

Isn’t that how the world works? No? I

Michael: don’t know.

You know what I, my degree was in philosophy in college and, I thought we just sat on a rock and tried to make the world a little better. I didn’t know about the patches and the pipe. Maybe I should try that.

Anthony: You missed out. We studied different branches of philosophy, man.

I had professors who were like, all right, I need you to write this paper about what it’s like to be a zombie. In order to do that, you gotta go get drunk first, or high. And I’m like, wait, what is happening here? Anyway, getting off topic somewhat related to this AAA has a study about, basically saying that consumers are, have fear about autonomous vehicles.

Most U. S. driver, drivers either express fear, 66%, or uncertainty, 25 percent about fully self driving vehicles. Only 19 percent of respondents actually listened to this podcast. Okay, that part is a lie, but I mean, this kind of goes into what we were just talking about. But also people are interested in semi autonomous technologies such as reverse automatic emergency braking.

Which I think I talked about a couple weeks ago saying, hey, I love this. Are people where does these partially autonomous systems that we were just talking about, like the Blue Cruise, Super Cruise, fit into this?

Michael: Well, I mean, they are, it’s self driving and advanced driving assistance are, that everybody’s trying to mixed up, mixed it up here.

In this study, it’s, it looks at semi autonomous vehicles and autonomous vehicles, but I think the, really the, Thing we’re looking at here is that they’ve been doing this study I believe for four years now and it’s basically three Measures, are you afraid of self driving vehicles?

Are you unsure about self driving vehicles or do you trust self driving vehicles and From 2021 through now, I mean, despite, some of the growth of these vehicles, these vehicle companies, some of the their presence on our roads has increased and you might expect that the numbers would slowly be going down the people who are afraid or the people who or the numbers might be going up among the people who trust these types of vehicles, but we’re seeing the opposite in many ways.

And maybe this reflects a lot of the problems. That, that Cruz has had in the past year or two. But you know, we’re looking at two thirds of the public that’s surveyed here, being literally afraid of these vehicles. And that’s up slightly from 54 and 55 percent in 2021 and 22, which is what makes me think that some of the coverage around Cruz and some of those things might’ve contributed to this.

And at the same time, we’re seeing that the folks who trust and who are unsure about self driving vehicles go down. It appears that, driver attitudes are turning somewhat against the technology, and not by large swaths. But, There’s no question at this point that the majority of Americans are skeptical about these vehicles.

I mean, you’re looking at numbers that suggest well over 90 percent are either afraid or unsure about the technology. In, and that numbers not really changed too much over time. It was 85 percent in 2022 and 86 percent in 2021. So while, manufacturers have appeared to be doing little to assuage fears.

And in fact, some of the events of the last couple of years have increased uncertainty and skepticism about the success of these vehicles.

Anthony: I mean, I still think these guys gotta go back to and learn how to crawl before they start running. Hey, let’s get things like AEB working correctly. And then, that doesn’t work correctly.

I don’t know how much I trust your car gonna drive itself everywhere. But, hey, I don’t have billions of dollars to throw at toys. If I did, man, it would be so cool. But know what scares me more than AVs? Spiders. No spiders. Drowsy drivers, distracted drivers, mainly distracted drivers, but we’re not gonna talk, we’re gonna talk about drowsy drivers.

Another triple A study showing, or actually Nitsa stated that 91, 000 police reported crashes resulted in 50, 000 injuries and 800 fatalities annually. Approximately 1 2 percent of all crashes, injuries, and death involve drowsy driving. Let me just say that again. Approximately 1 2 percent of all crashes, injuries, and death involve drowsy driving.

And we’ve all definitely had white line fever. Where you’re driving and you’re just Pull off the side of the road. I remember when they started putting in those, I don’t know what they’re called in The breakdown lane that the dudududududududu. Yeah. I mean, that was cool when they started putting that in there, because I discovered that frequency at 60 miles per hour really lulls me to sleep nicely.

It’s, it was very comforting.

Michael: Well, I mean, what’s not gonna lull you to sleep, when you were saying one to two percent of all crashes, injuries or deaths, you were saying with kind of an ominous tone as those, that was a high number. But what the AAA study actually found is that, About 17. 6 percent of all fatal crashes involved a drowsy driver.

One of the issues here is that drowsy driving is not reported accurately. It’s not investigated adequately and it’s the numbers that NHTSA comes at with based on police reports are probably a lot lower. I mean, according to the study, but maybe by a factor of 10 or more. It looks like there are a lot more people out on the roads that are driving drowsy and Well,

Anthony: I’m reading research drowsy, so that’s why I got it wrong.

I missed it. Okay, it happens.

Michael: That’s, that’s getting up near, one fifth of people. All fatal crashes are one six somewhere in that area. So Where drowsy driving was at least a contributing factor? I believe in about a third Of those drowsy driving events. We did see, blood alcohol levels that were contributing to the problem, and a lot of these events are happening in those nighttime hours where you might see more folks driving with high blood alcohol levels.

There, that is a contributing factor and alcohol does cause drowsiness. There, there’s going to be some relationship there. But, essentially, I, I think what this study shows is that there needs to be a good deal more research. Put into how to prevent this type of driving, drowsy driving.

Anthony: I think this puts into what we’ve talked about a lot in the show is that all driver’s seats need to become equipped with a caffeine syringe or an espresso machine. I mean, that will prevent these things. You sit down, boom, we get spiked right in the butt with a caffeine shot. You’ll be wide awake. No more drowsy driving.

If you support this idea and other ideas just as ridiculous as this good autosafety. org and click on donate. We will not support this idea. This is a horrible idea, but we’ll appreciate your donation. How’s that?

Fred: Hey, I’m going to quibble with this AAA study just a little bit. And this will seem like too much information, but they can’t report 17.

2%. That’s a ridiculous amount of accuracy based on the model that they’ve got. So I think that it’s supportable to say that there’s Perhaps somewhere between 10 and 20 percent, but 17. 2 percent is that kind of accuracy is an illusion that they simply cannot support. End of rant. Nerd.

Anthony: Oh, sorry. Sorry.

Okay. This is an interesting story out of USA Today. A Kansas car dealer indicted for rolling back odometers. As cases surge nationwide. Adam Newbery, 31, of Derby, Kansas, was charged with 27 counts of criminal misconduct, including odometer tampering. And I remember back, like, the analog ones.

They’re sealed up there, but there’s always some clever guy like hey, I know how to do this. And I figured in the day of di as I start saying this out loud, I realize I’m an idiot. Yeah, in the way of computers, man. Computers are perfect. There’s no way anyone could hack this. Horsewiden, I think.

Yeah. Yeah, so they’re hacking the digital odometers.

Michael: That looks like a big part of the problem here. I mean me being an idiot That’s part of it, too. But really the As we move to more to digital, speedometers odometers digital instruments in your vehicle period we’ve talked about it many times.

The automakers are not putting enough security into the systems in your vehicle to prevent hacks and whether they’re hacks for, whether it’s a hack that’s coming out of Iran to possibly cause a terrorist incident, which is our biggest fear, or one that’s, taking place in Kansas by a guy who’s intent on selling a used vehicle for a couple thousand dollars more than they might otherwise have gotten because of the vehicle’s high mileage.

It’s a huge problem and it needs to be cracked down on. And if manufacturers can’t make something as simple as software that prevents odometers from being falsified, then what does that say about the rest of the vehicle software architecture?

Anthony: So with the analog odometers, was there any sort of standard on, Hey, this is how these things have to be manufactured to some sort of specification?

Michael: Well, there were there, I don’t know that there was a standard on how they were manufactured. There was, there is a NHTSA rule on odometers that. Basically sets out parameters for disclosing odometer, disclosing mileage. I mean, one of the best ways to combat odometer fraud is by looking at accurate vehicle histories.

If you’ve got, for instance, a Carfax report or a repair ticket from a dealership that notes the vehicle’s mileage at a particular time in its lifespan, then you can double check those documents against whatever the mileage is on the vehicle. That’s. Currently displayed. So another reason why we suggest that everyone buying a used car gets some sort of vehicle report that covers repairs and cover, you want to know what’s happened to that vehicle during its lifespan, not just crashes, but you know, it’s interesting to see the repairs and whether the recalls have been performed and things like that.

And in this case, it would be really great. for you to be able to look back and say, huh, this vehicle had 97, 000 miles on it in 2022. And now it’s got 43, 000 miles on it. Something’s clearly wrong here.

Anthony: I drove backwards.

Michael: Yeah, that worked. That didn’t work in Ferris Bueller’s day off and it won’t work here.

Anthony: But is there like around those old school odometers, was there like some sort of tamper resistant seal or something like that? Was there something that yes, it could be broken, but you have to dig down on the car and say, ah, this has been broken. No idea. Nothing anymore.

Michael: I mean, there’s nothing that’s preventing.

I mean, there might have been something that was put in place to try to you. You make it harder to get into, but, just like everything if someone wants to get into something bad enough and hack it whether it’s a physical component or a software component, they’re probably going to find a way.

Anthony: Fred, as our resident, uh, felon have you ever taken apart an odometer and tried to roll it back?

Fred: I never never did that exercise. So I don’t know. I tried to do it once, but I failed. So I don’t know how to do that. Okay.

Michael: And there’s, there’s, there was a final rule for, it’s, that was issued in 2019 by NHTSA that was basically charging states with adopting electronic odometer disclosure systems.

And, that’s in place, but, it’s not preventing about half a million vehicles every year in California, it appears from the study from having this problem occur. I mean, it’s not a small problem. There are significant numbers of vehicles that are being having their odometers rolled back.

Anthony: Yeah from the article we linked to, it says, Now digital odometers can be rolled back by removing a car’s circuit board or using equipment that fastens into the vehicle’s electronic circuit. So I imagine if you remove part of the circuit board, there should be some giant errors in there. Like here, odometer reading with an asterisk next to it or something.

I’m naive.

Fred: I think this probably bleeds over into today’s style about the common access network, the CAN bus. Yeah, the CAN bus. So let’s defer that. Discussion for a moment.

Anthony: I’ll let, I say, let’s just jump right into the TOW today. Cause it’s beyond mobilized automotive cyber threats overview, the can problem brought to you by.

Fred: All right, so we’ll go ahead with that. The I think that the odometer information is probably updated through the CAN bus, like everything else, all the other data that’s being carried on the vehicle. So what is a CAN bus? A CAN bus is basically two wires that connect all the electronic components in your vehicle.

And they call it a bus because all the vehicles can talk. Over the CAN bus, so it’s like a, it’s like a superhighway for data. And one of the features of that is that it’s easily hackable. It is inherently not secure. So some enterprising folks have gone ahead and figured out how to break into this by removing the headlight on your car.

And then they inject a command that basically says, um, you have, this, the command that they inject says that the FOB that would otherwise start the car has been authorized. So whether or not it has actually been authorized, that’s the message that goes out over the bus and they have no real means of defeating this.

They can also send out a message that says, okay, unlock the doors, again, because the door locks are also attached to this CAN bus. This sounds complex, but I’m going to give you an analogy, which I like to do, and if any of you have ever seen Seinfeld, there’s A really good episode that relates to Jerry presenting at career day in his junior high school.

I don’t know if you remember that, but. Classic. At the,

Anthony: what’s the deal with homework? You’re not working on your homework. What’s the deal with homework,

Fred: kids, right? If you want to have a career and you like making people laugh, et cetera, et cetera. But part of that episode is a fire drill. And then that fire drill, the normal process of things is interrupted, and instead of kids being kids, the teacher says, Jerry, single file, and everybody gets into a line and moves out of the, starts to move out of the building.

Well, the teacher is rather like the can bus being interrupted, right? So she’s taking, she takes normal processes. And she makes them very orderly, and she has a single command that takes over all of the behavior of all the children, including Jerry. That’s what’s happening when these people break into the CAN bus.

They interrupt the normal communications, they force an overt communication, an overriding communication on everybody. And they proceed out the building, except, of course, in the case of the CAN bus in your car, they don’t proceed out of the building, they proceed into your car and out onto the highway.

Anthony: Is the CAN bus, is this a standard, or does each manufacturer do their own thing with it?

Fred: The CAN bus is a standard. And you can go out and buy CAN bus modules for, five or six bucks and stick them into your circuit and that allows you to have access to any other CAN bus. They’re really not authenticated they don’t need authentication.

We’ve talked about authentication in a couple of different venues. One is, of course, with AVs and they’re Requirement that nobody seems to appreciate to authenticate commands for the car operation and destination. We think is horribly important, but nobody else seems to be getting on to. So that’s one of the, that’s one of the mechanisms that people using to steal cars.

There was another article on a way of getting your personal information by hacking into EV charging mechanisms. And these are household EV charging mechanisms that they were talking about, but this, these comparable technology would work to break into EV chargers that are commercially located, for example.

And what they did is they opened up the cover, they physically removed a module called the Raspberry Pi computer that was running the system. And they simply popped that into a socket, looked at the data, and they were able to Personal information including passwords and locations and all kinds of things that you would expect to see in a system that it needs to be authenticated.

This is a big problem and Raspberry Pis were never designed to be inherently secure. They’re great systems, I’ve used them, I’ve programmed with them and used them in my Epic Bird Feeder design, but they’re not inherently secure. There’s lots of them out there. Now, there are ways around this, of course, and they could build more security into the Raspberry Pis for industrial use, which I guess is one direction they’re headed in.

Another direction that they could go in, and this is used by a friend of mine and certain government systems, is to automatically wipe the memory if the device is physically removed from or tampered with inside of the case. Or even to physically wipe the, or, logically wipe the memory if the case is opened without authorization.

So there’s a lot of things that can be done, but none of them are being done because of the current gold rush and people selling shovels in terms of the electric chargers that they’re putting out into the public domain. There’s yet another problem with these, which is that There’s so many lines of code that there are a lot of ways of breaking into it.

And people now use something called object oriented code. And without getting into too many details about that, one of the governing features of object oriented code is that it is inheritable, which means that one piece of object oriented code can refer to the data that’s in another piece of object oriented code.

This is a great opportunity for hackers to get in and break into, for example, the source for the inheritable code, like everybody uses GitHub to download code that they stick into their program. Well, if you are an enterprising hacker, you can get into GitHub and you can put some corrupted code into one of the source codes, the object oriented code that everybody’s bringing into their own systems.

For example, the source code for the CAN bus, right? So there’s a lot of opportunities. So I did a quick calculation to say how many opportunities exist. So I went to Wikipedia. And I looked at the Wikipedia entry for cyber security. And there’s about 1, 250 lines in there. And about 100 characters per line 125, 000 characters, 311 references, assuming 1, 000 lines for each, you’re up to 39 million characters.

And with the 10 suggested references, a couple of numbers 12 million some odd characters. So if you look at the total exposure to typographical errors in that single article, which is analogous to the exposure to cyber hacking by some nefarious means, you’re looking at something like 51, 625, 000 opportunities for somebody to break into it.

This is, this may not be a perfect analogy, but it’s apt because there are similar lines of code and similar numbers of inheritable software modules used in the software that’s driving your car or that’s running your EV charger. So the question is if you’ve got 51, 625, 000 typo exposures, how many typos actually exist?

And that’s just that one article. So somebody’s got to go in and examine those 61, 51 million some odd characters to see which ones are supposed to be there and which ones are faulty. And that’s only one of the articles in in Wikipedia. If you look at the number of articles that are currently in there, You’re talking about something like 6, 799, 000 articles that are in Wikipedia, something like 4, 500, 000, 000 words.

not counting videos, which is yet another attack surface. So this gives you some idea of the magnitude of the problem of sourcing authentic code, making sure that authentic code has only the code that you intend and has no malicious code in it. It’s a huge problem. And then if you add on the fact that you’ve got to overlay The philosophy that’s currently driving ethics in America and around the world today, that’s a big problem.

And I don’t think that the industry is really stepping up to the magnitude of even the problems that they understand, much less the need for the ethical overlays that are really going to be required to make these AVs acceptably safe on our highways.

Anthony: I, we’ve talked about before that in order for big change to happen this sounds very dark, but the right people have to be affected.

I’ll change it so it’s less dark. For some sort of safety thing to come in, the right people have to be inconvenienced. With all the software in cars, what I’m thinking is last week, for example, Apple put out an update to its operating system. And, they’re a large company, they do a lot of checks and stuff, and a whole bunch of stuff doesn’t work on their computers anymore.

And they’ve been doing this for, since the 70s, right? And so they manage to break stuff regularly. People have had it with your phones, with Microsoft Windows, like, Microsoft, I think it was even worse, was years ago, somebody managed to get one of their security certificates and send out a false update that affected some people.

Car manufacturers are not software companies. So is it gonna be that basically a whole fleet of cars are gonna have to get bricked? And blue screen of death before they go, Oh, wait, we should take this. Cause most people look at your car, even though it has all these computer systems in it, as I go in, I press a button, I can start driving.

I’m not thinking about software. I’m thinking about, I got a pedal. I got another pedal. Sometimes I have a third pedal. It’s very confusing. I got a steering wheel, got my radio. Is that what it’s going to take is some large scale failure?

Fred: Well, either a large scale failure or a small scale failure to a well known person.

That seems to be the mechanism. What percent should we nominate? I could come up with a few, but that’s out of scope for

Michael: That’s going back to dark.

Anthony: Oh, that’s making it dark? Well, what if it was entertaining? Like, we had, I don’t know. It happens to Howie Mandel and like Confetti pops out of his car.

I don’t know. That could be entertaining. I’m sorry. That

Michael: doesn’t really seem like a mechanism for change. I mean, if it’s confetti, it’s fine.

Anthony: It’s true. It is, especially when they put it in the cake.

Pardon me. I’m laughing now. So we need basically some, because legislatively it won’t move forward because legislators don’t understand, um, what things are. I’m just going to keep it there. They don’t understand things, so they won’t make a change until somebody pushes it. The auto industry is not going to push it because it’s more work on their end, so it’s going to be more of consumer advocacy groups such as the Center for Auto Safety.

Hey, thanks for listening. Have you clicked subscribe? Have you clicked five stars, liked, told all your friends, and donated? Because really, that’s the only way this is going to change. End of rant.

Fred: Thank you. It’s unlikely to end a rant, but

Anthony: My life is a rant. Ah, boy. Alright, let’s let’s jump into more interesting stuff.

Exploding bus. Let’s talk about this. There’s an article we’re linking to from The Guardian titled, Heroic Bus Driver in New Orleans Gets Students Out in Moments Before Explosions. Oh my god. This was a school bus driver, she’s got a bus full of kids driving along and she saw that. Hey. There’s flames inside the bus.

Managed to get all the kids out, get them away from the bus, and then the front end of the bus exploded. This so it’s interesting when we start, when I started seeing all the recall data and whatnot from NHTSA. You’ve got your usual cars and stuff, you’ve got your Fords and your GMs and your Teslas.

But then there’s a whole bunch of manufacturers of things you’ve never heard of, and you’re like, what is this? And a lot of these are school buses. What, what’s happening here, Michael, with this exploding school bus? Well, first of all, round of applause to the to the woman who saved all the kids.

Fred: Yeah, she did a great thing. She did a great thing. And and to be clear, this was not an electric bus. This is a conventional bus. So we’re not slamming AVs in this particular case

Michael: or EVs. And, the two things to note here, I think one is that, bus fires are Given the number of buses on the road, it’s relatively common.

I think we, the article notes that we see about one bus fire day in America. Obviously not quite this dramatic and not all of them are life threatening. But it is, it’s a significant issue. We’ve seen some really bad fires. That have affected buses in the past that were a lot more detrimental and a lot more tragic than this.

Kudos to driver in this situation. There’s not a lot of information on what caused the fire. It appears that it was, a somewhat. Slow moving engine compartment fire that, accelerated significantly once, once the children had escaped the bus. Also, I, it’s worth pointing out that the driver who saved everyone has the first name of Kia, which means it’s probably the first time that a Kia has actually saved people from fire in recent history.

Ha!

Anthony: If you enjoy the comedy stylings of Michael Brooks, see him at Chuckles in Cedar Falls, Iowa.

Michael: That’s about as far as my comedy career is going is chuckles. But you know, just another note on, buses, this is a, buses are relatively safe. We see it compared to other vehicles. We see very few in terms of data, very few.

Fatal and injury crashes involving buses each year, but when they do happen, it can be a significant tragedy and it affects the, those of us who are least able to protect themselves. And so it’s important to have people like Keo on board who can Feed the can bus in this situation and organize kids and get them off the bus in time to avoid what turned out to be what could have been a very life threatening situation.

Anthony: Okay, so different subject. This is a couple articles on lowering the blood alcohol limit. And it’s not passing. So I remember in D. C. when I lived there, the blood alcohol was like zero. It was like you couldn’t have anything if they pulled you over. Is that still the case?

Michael: It’s there is a law in DC that subjects you to a I think it’s I think it’s a different type.

It’s not a DWI. I think if it’s what they call it, but it you’re not allowed to drive in DC with really any, there is no allowable alcohol, blood alcohol limit in DC, you can be ticketed. It’s just that the punishment for driving in DC, if you’re under 0. 08, which is the legal level of punishment would be less significant, but it’s certainly a deterrent.

A lot of people don’t know about that. But what this article is Washington State’s trying to become only the second state to move the legal limit for blood alcohol concentration from 08 percent to 05%. Utah is the first, and it’s important because Utah had some seen some significant changes since they enacted that driving limit.

They’ve had, it looks like they’ve had fewer drunk drivers. Surveys have indicated that drivers are recognizing that the law has changed and they’re changing their behaviors. And the number of people killed and injured in crashes involving drunk drivers has gone down in Utah. And a lot of that the news.

probably directly traceable to this change in law. It’s something that our friends over at Mothers Against Drunk Driving have been, pushing very hard to get as a nationwide standard. And it’s something that hopefully in the future, as we start to see the alcohol Detection systems move into our vehicles right now.

There’s a lot of conversation around what the limits there should be. Can you know, can we even accurately detect 0. 08 yet is a big issue there. And it looks like that we’re going to be starting out with detection systems that detect, maybe point one point, 0. 15 percent or higher levels of alcohol.

That the technology is not quite ready to distinguish between lower levels of blood alcohol concentration, but hopefully someday we’ll get to that. So this is going, this is an issue that’s, not going to go away for a long time. And I know there are other states beyond Washington that are also considering.

The 0. 05 percent standard. And it’s something that I think we would certainly support. I mean, ideally no one is drinking and driving at all at any levels. And this is, one smaller step towards that future.

Fred: Can we get the deep state to support this somehow?

Michael: Whoa. I lost my phone number to the deep state.

I’ll have to rely on them for that.

Fred: Anthony, you got that number?

Anthony: Yeah, but I can’t talk about that. I don’t have tinfoil in my head right now. No. But look, you shouldn’t drink and drive, but when you do, just go all out. What is he saying? This is the wrong podcast for that. Now it’s glad to, I’m glad to see that my, my future home of Hawaii when I win the lottery is pushing to lower their blood alcohol content level as well.

Especially because they don’t really have straight roads there.

Michael: You and if you have any questions about whether 0. 05 percent is truly safe, look at the Netherlands and Japan they’ve put it into place nationwide and they have, seen significant reductions in traffic deaths due to this.

Limits in America, where a substantial percentage of the population apparently still thinks it’s a constitutional right to drive while drinking to the point that NHTSA had to write about that and its request for comments on that. We have a different philosophy and a, frankly, more damaging philosophy than some other countries.

Anthony: Well, my car drives itself. What if my car is drunk? Can that, can I give, what if my car runs off of ethanol? Isn’t ethanol alcohol?

Michael: Well, yeah, but you’re presuming your car has a human organ system, right?

Anthony: Don’t look under my hood, okay?

Michael: That’s creepy.

Anthony: Yeah, I apologize, people. I’m suffering from a head cold and the brightest

Michael: Yeah, you’re having a bad day.

Anthony: Okay. Oh, God. I’m just giving myself more editing work, aren’t I? No, because I’m not going to do it because, well, let’s go to recalls. How do recalls sound? I think recalls sound great. A! Fork! They’re rarely in recalls. 1, 500 vehicles. This is the Ford Super Duty High Series headlamp. Oh. So they’re saying one of the suppliers supplied them something that wasn’t lighting up well, is that correct?

Oh, the lower beam photometric requirements.

Michael: Yeah, I think what happened was, yeah, you’ve got, these are replacement headlights. These weren’t originally installed on their super duty vehicles. However, they are replacement headlights and apparently the supplier manufactured those replacement headlamps as It’s a without, they’re too, basically, they’re too bright.

They’re brighter than they should be under Federal Motor Vehicle Safety Standard 108. And Ford is essentially having to go out and find all the folks who had these installed. They identified 73 vehicles, I was, that had them installed, but they don’t know, it’s hard to trace these things. If they just sold a headlamp to someone as a replacement and that person went and installed it, or if, there was a third party involved here, I’m not sure if these were at AutoZone or some other area some other company where they’re being sold by third parties.

And that makes it impossible, essentially, for Ford to track them down. Well, not impossible, I’m sure it could be done, but it would involve some significant expense. And

Ford is recalling these. It’s an equipment recall, not a vehicle recall, so the Super Duty owners will not be receiving a notice of this, I don’t believe.

But if you do drive a Super Duty truck, it is certainly worth making sure that you’re not blinding everyone around you because you have one of these faulty replacement headlamps.

Anthony: I agree. I think all headlamps should be reduced in their photometric volume brightness because you’re trying to blind me.

Michael: You’re really going to like, adaptive driving beam headlights that are able to focus the light on portions of the road where it’s needed and not on areas where it’s not like your eyes.

Anthony: Yeah. Or bouncing off of my rear view mirror. That’s always fun. The side view mirrors. Oh boy. Moving on Hyundai.

They’re in here with 28, 439 vehicles. Well, okay. Listeners playing the home game. What do you think Hyundai is in here for? Dun! Yep, vehicle fires. It is 2019 2022 Genesis G70s, 2018 2020 Genesis G80s. I mean, if you’re gonna get a Genesis, you’re gonna get the G90. Also recalled the left handed turbocharger oil supply pipe in the subject vehicles can crack due to prolonged exposure to high ambient temperature in the engine compartment, resulting in oil leak onto the exhaust manifold.

This is, was this Hyundai before where they didn’t realize how hot the engines got? Were they the one who had a similar recall with this?

Michael: Yeah, there was a similar recall on last week that we covered. I’m not, I can’t remember if it’s just, if it was Kia last week or Hyundai last week or Genesis, they’re all pretending it was a Kia.

Fred: I remember that one. Yeah.

Michael: But there was a little more information provided here and it’s, a term I’ve never really heard outside of random teenagers on the internet, which is swag. There was insufficient swagging of the turbocharger oil supply pipe fitting. Maybe Fred’s dealt with swagging before.

Fred: That’s the term is swaging. Swage, okay. And that means that you’re using, basically you’re squeezing one tube into another and the compression forms the bond.

Michael: The internet pronunciation that I was given for that is incorrect, and it has nothing to do with swag.

Fred: Nothing to do with swag. You’re right.

It’s

Michael: what I’ve, it’s like what I’ve tried to do all my life at restaurants with straws and stuff them inside of each other to make one gigantic straw.

Fred: Exactly. Right. Okay.

Michael: No. Oh, okay. There was insufficient swaging of of the turbo charger, oil supply, pipe fitting, The straw trick that I tried was never really a good thing, it didn’t really work.

It left a gap there that opened up to let air in from the outside. I wonder if there’s something similar going on here. It doesn’t sound like you should be swaging a turbocharger oil supply pipe, based on my experience with straws.

Fred: Well, I was going to ask you how that’s going, Michael, but I’m glad you gave that elaboration.

Now, the swaging is a standard process. It just means they did a crappy job.

Anthony: Oh boy. So hey, if you need to swage and you need straws, head down to your local Piggly Wiggly. They’ve got straws. I’m sure they do. Still never been to one, but I receive three points for mentioning Piggly Wiggly first in this episode.

Next recall, Chrysler, aka Stellantis. 1, 831 vehicles. This is the 2016 Jeep Wrangler. Wrang Aw, man, the ability to speak is

Michael: Yeah, it’s a Jeep Wrangler, Thank you. Airbag clock springs. There was a recall and this is I think the third recall that’s popped up this week I saw a couple more this morning on this.

Basically, there was a recall that I think it was 16 v 288 was the number of it So it’s an eight year old recall and this has gone back and looked at it and said, hey, Chrysler, Jeep, we think you left some vehicles out of this. And the clock spring is really important because essentially without it, your airbag is not going to fire.

It’s a mechanism that supports your occupant protection systems and ensures that your airbag is going to deploy at appropriate times. And it looks like there is a similar Let’s see, there are three recalls right now, so this isn’t the only one that are involved in the clock spring failure.

That’s something that owners of these vehicles definitely want to get into the dealership and have fixed as soon as possible. It looks like that’s going to start around the beginning of May. Chrysler’s had significant issues with these clock springs over the years. And it’s something that is, is not insignificant despite the age of some of the vehicles that are covered here.

Anthony: Check your clock springs, people. Last recall, a little company called Jaguar. 2, 842 vehicles. This is the Jaguar E Pace. They’ve actually made 2, 842 of these things. They’re conducting a voluntary recall involving certain 2021 2024 Jaguar E Paces built in Austria. This is where a result of an incorrect setting in the car configuration file, the brake pad wear indicator does not display where brake pads have worn to a condition where the indicator lamp should display.

Okay, so they go to the trouble of putting this indicator lamp in there, and then they’re like, hey, we didn’t plug it in.

Michael: Yeah, and it, when it comes to brake indicators Federal Motor Vehicle Safety Standard 135 basically says, You can put them in. You, I mean, I have never owned a car that had a brake pad wear indicator.

I wait until I hear squeaking from my brakes when stopping and then get them replaced. Or look for a mouse. Yeah. I mean, that could be it too, but in these vehicles, they’ve got this fancy brake pad wear indicator and what FMDSS 135 says, if you’ve got it in there, it’s got to display.

And so Jaguar is having to go back and correct us because there’s a non compliance with the motor vehicle safety standard.

Fred: The way the wear indicator works, just for your reference, is they basically drill a hole laterally down through the brake pad and put a conductive wire in it, like a copper wire.

And when the pad wears down enough so that the copper wire contacts the disc, then you complete a circuit. And that’s how the car knows that the. That the brake pad is worn down. So it’s not a very sophisticated indicator, but it seems to work perfectly well.

Anthony: Okay. Related to brake pads, did you guys see that asbestos, the U.

S. has finally banned asbestos? And I assume they did this, I don’t know, decades ago, but they’re still using asbestos in brake pads. Did you guys know this? Go out and buy an asbestos brake pad. Filled brake pad or is this a thing? No.

Fred: I thought it was gone years ago too. I used to when I worked on a gas station, our stint at approach was to open up the brakes and then hit it with compressed air to blow out all the dust.

No. Oh, yes. And that was the standard approach. And asbestos concerns didn’t come up until many years later. Well, I’m

Anthony: glad to see that you’re still here. Geez. So far. Yeah. So far. The episode’s almost over. Lastly, you gotta, do a little Elon Musk update, so his roadster we’ve talked about before, he’s like it’s gonna go 0 to 60 in less than a second, aren’t I tough?

My mother loves me. Now he’s saying that it’s not even really gonna be a car this roadster. And it’s gonna have jet engine stuff in it, it’s gonna be really cool, it’s gonna have like, flame stripes on the side of it, and it’ll be great. And it’ll put out Doritos when you drive it, it’s the best thing ever.

And it’s gonna have a waterbed in the back. And he has a prescription for ketamine. Uh, again, referring to last week’s episode, I believe, where Fred Perkins former math professor, did basic math and said, No. No. This is not happening. So if anyone else wants to play the at home math game no, zero to 60 should, why isn’t this in the federal motor vehicle safety standards?

Like don’t sell rocket cars. I don’t think

Fred: it ever occurred to anybody to put a rocket engine under a car that’s been reserved for drag strips before. Right. And by the way, this is going to be really loud. So you’ll know if you have any, Neighbors with testosterone deficiencies or excesses from quite a long distance.

Anthony: Yeah, like, seriously, Michael, like, again, I’ve asked this a hundred times, I’m sure. Why can’t, fine, go ahead and make these cards. There’s no way these things could be street lethal. Like there’s,

Michael: I don’t know. I don’t even believe anything that comes out of Musk’s mouth anymore. So there is, it’s just seems like a bunch of puffery to me.

I mean, I’m pretty sure that there’s no way a rocket propelled vehicle is going to be allowed on America’s roads, but you know, I’ve been wrong before. We’ll see.

Anthony: Well, if you have a rocket propelled vehicle, just crack open that six pack and let that thing rip. All right, well

Fred: Let a wily coyote be your guide.

Anthony: Yeah, because I got nothing positive to say and I caught myself there. Look at that, I’ve made progress.

Michael: I’ve been a little out of the loop, too, in this episode. Ever since the word raspberry pie was mentioned, I got distracted by that.

Anthony: I’m holding a raspberry pie in my lap. Oh Hey, thanks for joining us again.

Listeners. Hopefully I’ll be back healthy next week and won’t misread things about drowsy, distracted driving. And thank you. Tell all your friends, hit subscribe, hit donate. You donate

Fred: again. Don’t drive like my moderator.

Anthony: I’m a good driver.

Fred: All right. Bye.

Anthony: And Michael says goodbye to him.

For

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