The Chronicles of Auto-Infamy: Drowsy, Distracted, and a Dash of Rust

[00:00:00] Introduction to the Podcast

[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

[00:00:28] Fred: to your favorite podcast. Good morning world. People are thinking about this.

[00:00:34] Anthony: They’re like, wait, this isn’t all things considered. I don’t know if that’s a podcast or not, but listeners, Hey, I want to point out something to you. We love our passionate listeners and whatnot. If you’re here, you probably agree with at least 85 percent of what we say.

People have pointed out times when we’re wrong, and we go, oh, you know what, maybe you’re right. I want to point out that 90 percent of what comes out of my face is generally nonsense, and it’s generally humor, and sometimes [00:01:00] that’s tough to translate. Just relax, I understand this is tough times, people adjusting to the new world order, hopefully we’re informative and a little entertaining.

Sound good?

[00:01:13] Michael: Sounds good to me. I admire your confidence, that you think you’re actually funny.

[00:01:20] Anthony: I hate this show. This is my show.

[00:01:23] Michael: Sorry, I had to start it on you so quick. I get it,

[00:01:26] Anthony: I get it. Alright, stretch.

[00:01:28] Rivian and Volkswagen Partnership

[00:01:28] Anthony: Let’s do a little, let’s do a little story that came out in I believe today’s or yesterday’s New York Times.

Rivian is teaming up with Volkswagen. So Volkswagen, everyone’s heard of it. Older autumn manufacturer. And they’ve been saying, Hey, how do we get into this Evie market? How do we get into this hardware and software stuff? And then they come across a Rivian, which is run by a very smart, very clever engineer.

And a bunch of very smart, clever engineers, and they’re like, Hey, we’re building this stuff, but how do we get money to do [00:02:00] it? Because this stuff is expensive. So this is neat, where, Volkswagen’s saying, Hey, we used to have the mindset of that, If we didn’t invent it, it’s not good enough.

And then they’re realizing, hey, maybe we should go to these Rivian folks, and I’ll put this an aside. When I once worked for a fruit company, they also had the same belief that, hey, if we don’t invent this software, hardware, it’s garbage. Then they decided, Hey, we can just buy everybody’s idea. And that’s turned out pretty well for them.

So what would you guys take on this?

[00:02:27] Fred: The company, the term synergy? Yes. What that means? Synergy is like one plus one equals three. Some . So this is, basically, a synergistic combination of two companies that are losing money that are somehow supposed to end up with a company that makes money.

So it’s going to be interesting to see how this plays out. Are you familiar with the term cynicism? No. What does that mean? I don’t know. That’s a Greek illusion, right? Yeah. Something

[00:02:56] Anthony: about the, something about Pericles. I’m not sure. [00:03:00] Shadows on a cave wall. I’m not sure either. But we digress. Yes.

[00:03:06] Michael: Yeah, it makes sense.

Volkswagen is struggling recently for market share and struggling to get electric vehicles built and on the road and Rivian has some, a lot more expertise in that area. So we’ll see what happens here. It’s,

[00:03:23] Anthony: it’s going to be interesting. All right. Now we start off with something sweet and light and easy and now let’s get right into it.

Okay. Okay.

[00:03:30] Tesla and NHTSA Investigation

[00:03:30] Anthony: Tesla and NHTSA. NHTSA, the National Highway Traffic Safety Administration which we have a death watch on. No, that’s just me. It’s just me, really. They’re calling out Tesla saying, Hey, you know who you guys keep saying your cars drive themselves? They can’t. And and NHTSA is asking the company to revisit its communications to make sure messages are consistent with user instructions.

And now that phrase, revisit its communications, in my [00:04:00] mind, I just imagine Nitza being this guy wearing socks and sandals, and being like, okay hey, class, can we if we all just just mellow out just revisit that is this Michael, you have a lot of experience with NHTSA.

Am I being cynical?

[00:04:16] Michael: So basically what happened was NHTSA’s conducting its investigation of Tesla full self driving, but they sent out a request, which they typically do in any investigation, soon after opening, they’ll send out a request. Basically the manufacturer’s saying, here’s everything we want from you.

But in this one, they also attached an email from May where they were speaking with Tesla about this issue and particularly about Tesla’s social media messaging. So Tesla came in April, apparently of this year and briefed NHTSA and the Office of Defects Investigation, and said, look, our owner’s manual, our user interface.

YouTube messaging to [00:05:00] drivers is that the vehicle is not autonomous. They have to pay attention, keep their hands and ready to go. But the agency pointed out, in the month, since they had that meeting with Tesla from April to may that they’ve come across, you know, what, 10, a dozen.

Twitter posts from Tesla, where they are actively supporting unsafe uses of the system. People that are not paying attention, not keeping their hands on the steering wheel, all these other things that we’ve discussed ad nauseum previously. So they’re catching Tesla in a lie or essentially.

Tesla’s talking out of both sides of its mouth. On the one hand, they’re saying, we’re putting out this messaging to people in our owner’s manual and online that tells them they have to pay attention. Meanwhile, they’re supporting tweets, reposting tweets online that do the opposite. It looks like that’s going to make up.

A-A-A-A-A, at least a portion of this investigation, looking into how Tesla is marketing or [00:06:00] characterizing this technology and social media, which as we know, has an outsize impact on human behavior as well. So it’s an interesting take. I, I wish that this was, a different situation.

I’m not sure what is going to happen to this investigation now that Elon is being nominated to the non existent Department of Government Efficiency.

[00:06:23] Anthony: So the head of the Department of Transportation, that’s a political appointee. The president appoints that person. Now is the head of NHTSA also a political appointee?

Yeah. Okay. So there’s going to be a new head of NHTSA coming in January, right? Yeah,

[00:06:38] Michael: maybe. We’ve seen so many acting administrators at NHTSA at this point that’s almost the norm. I think we’ve had one confirmed administrator in the past odd seven or eight years, maybe that was past six years now, I think that was actually confirmed by the Senate.

So it’s Who knows, there may be an, there’s probably an easier route to [00:07:00] confirmation of folks this time, since the Senate appears to be, or the Senate is in Republican hands. So any nominees coming out of the Trump administration should probably not have too many problems getting past the Senate.

So that’s, that’s. I don’t know what’s what the plan is for Musk’s role in the administration and how much, how much the Trump administration is going to be paying attention to his particular beliefs on vehicles versus the government efficiency question, whatever all that means, but it’s certainly troubling.

Nonetheless.

[00:07:40] Anthony: So using my example of NHTSA being a guy wearing socks and sandals and probably a skeleton being like, Hey, huh? I imagine with this new administration, he’s going to be that same guy, but with an atomic wedgie and just stuffed in a locker. Is that what we’re seeing? Am I?

[00:07:57] Michael: It’s hard.

It’s always hard to predict, [00:08:00] especially And what we’ve seen of Trump administrations, what’s going to happen in this case. I’m expecting a very industry friendly nominees to positions at the Department of Transportation and at NHTSA and perhaps people that are, have absolutely no experience in the areas of transportation, which we saw during the last Trump administration.

[00:08:23] Anthony: Oh boy. Yeah. What you’re talking about here of Elon being in charge of Dogecoin at the White House, the Department of Government Efficiency. Washington Post has an article talking about this, saying, It will be directed to drive drastic change by making recommendations on large scale structural reform and taking a more entrepreneurial approach to government.

I have so many problems with that. Mainly entrepreneurial approach to government, because government isn’t a business. It’s not a, it’s not a business I, it’s a, it’s most effective when it’s inefficient.

[00:08:55] Michael: Yeah. And how many startups fail for everyone that succeeds? Is that really how you [00:09:00] want your government to run?

It’s a very risky proposition.

[00:09:04] Anthony: Yeah. So it’s interesting a little, just the next paragraph. It talks about how Trump is giving to July 4th, 2026 for this group to finish its work and come up with its recommendations and it’s similar to the Grace Commission that Reagan appointed and the Grace Commission submitted more than 2, 500 recommendations to the White House and Congress in January 1984, most of which were never adopted.

So hey, on the plus side, these guys are going to make a bunch of noise. And if history is any clue, nothing will happen.

[00:09:33] Michael: I wouldn’t be so quick to assume that. The Reagan administration had a devastating effect on NHTSA’s funding. In fact, I would say that the agency is, has never recovered from that period of deregulation.

It’s funny and, it’s. We laugh because, Elon and Trump are both easy to make fun of for their actions and their behaviors, but ultimately, they may be putting into place systems that are going to [00:10:00] result in far less oversight by NHTSA, by the Department of Transportation, and by other government agencies that we’re engaged with.

[00:10:09] Fred: I haven’t been reading the newspapers for the last week, so I don’t really know what you guys are talking about.

[00:10:18] Anthony: Welcome to Fred’s meditation practice. It’s the only way, the only guaranteed life extending service available. Yeah. And further in the article Elon says he wants to cut the number of federal agencies from over 400 to no more than 99, which is one of these things that sounds great on paper, but it’s just nonsense.

It’s just words. Words, I’m gonna cut things and not say anything. But what you’re talking about is the devastating effect on NHTSA and whatnot is, that’s why there’s an organization such as the Center for Auto Safety, so we can be that annoying mosquito buzzing around these people’s heads that they can’t ignore as we continue to try and make roads safer for everybody.

Go to autosafety. org and click on [00:11:00] donate, please.

[00:11:01] GM’s Bicycle Alert System

[00:11:01] Anthony: Coming up next, we have oh, hey Ro, we’ll go slightly happy again. GM, I know you’re thinking, GM, always slightly happy. GM is working on a bicycle alert system for its, most of its vehicles. Not all of them, but most of them. Basically, we’ve talked about this a lot with ADAS systems where the, or ADAS?

Yeah, no, automatic emergency braking, where it has a tough time identifying people at night, people of color, people on bicycles, children. So GM has done a little PR spin about this and saying, Hey, one of my friends was, killed by on a bicycle and that’s horrible and they’re saying this is why it’s important to us to get this technology out there.

But is this more just PR fluff? Is this a silent golf clap to GM saying hey thanks for doing the obvious?

[00:11:49] Michael: I mean it’s this what GM is announcing is that they are there are three technologies that are involved here. There is the front pedestrian detection which is essentially [00:12:00] Cyclists in front of the vehicle and pedestrians as well.

That’s something that is being mandated by NHTSA to go into effect in 2029. all manufacturers are going to have that standard in their vehicles. Then they’re saying they’re going to get that into vehicles by 2025, I think, into all their vehicle standard, which is great. I’ll be interested to see if they do that.

And if those. systems meet the new NHTSA standard, which goes into effect in a few years. But in addition to that which I think most vehicles are going to be, are going to have they’re doing side bicycle detection. So essentially that’s. A lot like blind spot detection. You’re driving along the road and instead of just detecting vehicles that might be in your blind spot, their systems are going to be detecting folks on bicycles.

Which is also great. I’m sure we all love to be aware when anything is in our on the side of our vehicle that could be hit and bicyclists are a lot harder to see. In our rear view [00:13:00] mirrors. So that’s a great system. And also the situation where you’ve just parked your car on the road and you’re getting out of your vehicle and you opened your door right into the path of a bicyclist.

Fred, has that ever happened to you?

[00:13:13] Fred: I’ve never been doored. I’ve seen I’ve seen doors open and I’ve been able to avoid them so far, but I certainly know about that happening and people being injured by that.

[00:13:24] Michael: Yeah, apparently it’s a very common way that bicyclists are injured. So GM is wrapping, all three of those into a package that they’re putting on their vehicles starting next year.

It looks like all their vehicle standard without the government forcing them to. So that’s good. There are a number of other manufacturers that have started to put this Technology into their vehicles. They’re just not doing it across their lines. Volvo and Mercedes, Hyundai, Kia, Toyota, Volkswagen and some others have already started working on this.

And this is coming into play. [00:14:00] I think this is coming into play mainly because we’ve seen such a giant rise in pedestrian fatalities and there’s a renewed focus on that. By the manufacturers, because, frankly, the situation with pedestrians has gotten out of control. The numbers are grim and without some, some better technology on vehicles to detect pedestrians and bicyclists, it’s not doesn’t show any signs of slowing up.

[00:14:28] Fred: Yeah, and let’s also recognize that this technology requires a lot of sensor fusion, new software, new applications, so there’s radars looking out the side of the vehicle to implement this, there are cameras looking out the front and back, there is coupling with a brake system for automatic emergency braking, so there’s a lot of complexity involved in this.

I think that, we need to give credit for work that’s been done while still acknowledging that there is more work that could be done. But this is overall a really good thing that the [00:15:00] manufacturers, the giant manufacturer, like GM. Is focusing on this and is implementing this and hopefully the rest of the industry will come along and do similar safety projects.

This we’ve been advocating for this that people take the technology being developed for. And apply it to existing vehicles and existing traffic situations and I think this is 1 such example where that is. A V technology development is spilling over into. I think is a lot more practical and an immediate need for protecting people with existing vehicles and current technology.

[00:15:39] Anthony: The anti dooring stuff you mentioned, is that required in 2029? No. No. Okay.

[00:15:46] Michael: No, the AAB rule only applied to automatic emergency braking to avoid pedestrians in front of the vehicle and it did not include Bicyclists as part of that, and I think that was because the manufacturers [00:16:00] objected because bicyclists are moving at higher speeds and are harder for their current technology to detect.

That’s what they claimed. I’m not sure exactly true. That is obviously high speeds,

[00:16:11] Anthony: but they’re detecting a vehicle. That’s moving at high speeds. I imagine the

[00:16:15] Michael: vehicle is directly in front of them and as large and as not moving laterally, I think that’s mainly the issue there.

[00:16:23] Fred: We’ve talked about perception in the past and how difficult perception is as a very complex cognitive function that the computer has got to undertake and has got to, Do a very good job if it’s going to actually operate safely on streets automatically.

And I think that this whole problem of perception of a bicycle is something that’s difficult. If it were easy, they’d already be doing it. But the perception problem is difficult on the bicycle. It’s not really large. It’s bobbing around. It’s moving at different speeds. So again, giving credit where due.

I [00:17:00] think that. It’s a really good thing that’s happening, and hopefully there’ll be more of that.

[00:17:06] Anthony: Michael, you said with bicycles, they’re not moving laterally. Does the A B rule apply to cross traffic like at an intersection? No. It really, so

[00:17:16] Michael: If you can detect a pedestrian, we’re going to see how well it works on the pedestrian detection side.

We’re still not 100 percent sure. You know that it’s going to be, it’s not going to be perfect,

[00:17:27] Anthony: but I’m even thinking just a vehicle going cross traffic. So like I’m at a intersection, I’m going through, it’s not.

[00:17:33] Michael: The systems are designed to prevent rear collisions, so one vehicle hitting another one from behind, so they’re not equipped to capture a speeding vehicle that runs a red light in front of you.

I don’t think they’re going to be, they might capture it close to the end near impact, but they’re not going to be designed to have a 360 degree view of the roads all around you and to detect vehicles coming down the street laterally. [00:18:00] That’s more in the realm of what. Vehicle to vehicle communications or vehicle to infrastructure communication could accomplish for safety, but automatic emergency braking’s not there yet.

[00:18:11] Anthony: Okay. Good to know. Check your mirrors before opening your doors and don’t hit a bicyclist. Cause I know in New York State I don’t know if this is true anymore, but it used to be that, yeah, if you hit a motorcyclist, they hit in your door, they were at fault. But, then you’re, you don’t have a door and you’re dealing with someone who’s got a life threatening injury, probably.

Hey, check your mirrors, right? Everyone agrees.

[00:18:33] Michael: Yeah. That was like, did you ever hear, I think it’s called the Dutch reach or something. Yeah. I was just thinking of that. Yeah. Yeah. In other country, I think in Holland or the Netherlands, whatever they’re calling it these days they teach drivers and drivers education to reach for the door handle with their right arm, so they’re reaching over.

Which forces them to look at least towards their rearview mirror as they’re exiting the [00:19:00] vehicle so that they’re have an opportunity to see any bicyclists or pedestrians approaching before they open their door. That seems to be a kind of a no nonsense solution if we could if we didn’t have to rely on the technology, but fortunately, maybe we’ll have the technology there to help us as well.

[00:19:17] Anthony: I love it. The Dutch Reach. Okay. That’s something new we’ve learned today. Or at least I’ve learned today.

[00:19:23] The Touchscreen Debate

[00:19:23] Anthony: We’ve talked about this a lot before. No one likes touch screens so much. Not big fans of the touch screen. We all miss our buttons. And Hyundai has said, hey, we did some We did some user surveys some things like that.

And it turns out people hate it when they have to click through multiple screens just to adjust their air conditioning. And it’s fascinating because there’s an article from insideevs. com and they basically say the quiet part out loud is every manufacturer saw Tesla’s G Wiz factor with a touchscreen and said, Hey, we gotta be G Wiz too.

Let’s do this. Cause it looks [00:20:00] awesome in a demo. I blame Steve Jobs when he released the doc function of the Mac OS. I remember being at this event and he’s look how cool this thing looks! And we’re all like, this is such a stupid idea. This is bad usability. But Steve thought it was pretty and he got to scroll and it looked awesome on TV and hey, it looks like something out of the future and some Hollywood movie.

And that’s what all manufacturers did with touchscreens. They’re like hey. Remember when you used to push a button? Now, hey, how about clicking through eight different screens with no tactile feedback? And so thankfully, Hyundai is saying, maybe we should maybe we should put some buttons back in.

I applaud this. Yeah,

[00:20:39] Michael: They’re putting buttons back in for a number of reasons now, there’s a pretty, pretty sound consumer rejection of putting everything into the touchscreen system, particularly things that you want to access quickly while you’re driving. We’re, we’ve spoken a number of occasions about how we don’t want any [00:21:00] safety features being hidden in a menu behind a touchscreen.

There are things that have to be accessed. immediately, like wipers and headlights and other features where buttons make all the sense in the world, especially buttons that are put in a place that is recognizable from car to car so that people don’t have to relearn their locations every time they’re in a different vehicle.

Some things may be okay behind a touchscreen. If you want to do a deep dive into your fuel economy or into your trip Distances and that type of thing. That’s great. You can do that when you’re parked at home, but there’s a lot of features that just simply don’t belong under a touch screen menu.

Air conditioning is certainly arguably one you, if you’re driving, you want to make a quick adjustment to the climate and the vehicle without losing sight of the road, it’s sometimes I’ve had, I think it’s much easier to do so through a button versus, or a dial versus navigating a touch screen menu.

[00:21:57] Fred: Agreed. I think my car has got a [00:22:00] pretty good mix of buttons and touch screen. I liked the way they’ve done that. And in fact, they just came out with an update to the software that put a lot of those functions back onto the front screen, as well as alongside the buttons. So that’s my Subaru, my 2020 Subaru.

I liked the way they’ve done that. They’ve got the temperature disposed in buttons and they’ve got the Climate control mode, like whether or not you’re going to have defrost or warm your feet or whatever, that’s that’s on the touchscreen. And if you want to go into deeper considerations, then you can go down to different levels on the touchscreen, but they’ve done a good job.

So I don’t think it’s all one or the other. There’s got to be some optimal mix.

[00:22:48] Anthony: I don’t know my touchscreen. I have no idea what the functions are. I just let my phone take it over. Cause anytime I click on a button on the default Toyota stuff, it’s Oh, you don’t have this service. Oh, you don’t have this service.

Oh, you [00:23:00] don’t have this service. So I just plug my phone in and say, Hey, car, play this song. And then it just plays that song on repeat forever, over and over. Or it does variations of that song from children’s choirs to orchestras to heavy metal interpretations. And it drives me nuts. It’s it’s annoying.

It’s awful. But hey, that gets us into Gaslight this week.

[00:23:24] Gaslight of the Week: Motional’s Safety Claims

[00:23:24] Anthony: Everybody ready for some Gaslight fun? Everybody? Come on. Mr. Perkins, kick us off. I know you’re

[00:23:31] Fred: excited. Excuse me.

[00:23:34] Anthony: Whoa, that excited.

[00:23:36] Fred: My bad, sorry. So I have identified my Gaslight nominee this week as Motional, which is a company In Massachusetts, that is teamed with Hyundai to produce autonomous vehicles and it’s really like a throwback to where we were a year ago with these things before [00:24:00] the infamous incident in San Francisco, because motionals homepage says quoting from that our vehicles are never drowsy, drunk or distracted.

Their expert drivers singularly focus on the task at hand, safely traveling from point A to point B. That’s interesting. They never identify where points A and B are, so we don’t really know what the universe of safe operations for these vehicles is. That’d be useful to know, but, inquiring minds would like to know really where the boundaries of emotional AV safety are.

But it’s important in thinking about this to see the sleight of hand, because the statement Really ask the listener to project their own driving capabilities onto the vehicle when they say that they’re expert drivers. And they really wanted to presume that they’re somehow better because they’re never drowsy, drunk, or distracted.

Really, it’s an interesting sleight of hand that everybody who addresses [00:25:00] this in this way is using. But they never provide an affirmative statement of capabilities. So they’re really just asking you to project your own capabilities into the void. And, put aspirational properties that they hope to be.

Present it’s like defining the atmosphere as the absence of a vacuum absence of data is not data and the atmosphere is a lot more than the absence of a vacuum. Absence of a vacuum is 1 property. They’ve gotten the same way that this statement about their vehicles being safe really is just about a void into which you’re projecting yourself.

And let’s be real, some of the things it doesn’t have, does not have, is certification that it understands the rules of the road, like anybody who’s got a learner’s permit, doesn’t have certification that it’s a competent driver, like anybody who’s got a driver’s license, and importantly, there’s no evidence that it ever cares whether or not it’s driving safely.

Whether it cares about the safety [00:26:00] of the universe around it so it goes on, but I’m going to skip to the end of my rant here, because in the interest of time, and what they say, though, as you get into it is that they’ve Filed a voluntary safety self assessment, which is something that you can find on the website and in that they say that are quantitative assessment centers around mileage accumulation on public roads to demonstrate statistically that unknown hazardous scenarios are sufficiently rare.

They never say what’s sufficiently rare is they never demonstrate they never talk about how they demonstrate statistically that they’re compliant with their standards or what those standards are. Instead, they refer to subsection 1223 of the VSSA, and if you dive into that, it says confidence in our capabilities in known scenarios is only meaningful in conjunction with an understanding of the frequency of unknown scenarios.[00:27:00]

In AAV statistical safety testing accumulates mileage to establish whether the occurrence of unknown hazardous scenarios lies below an acceptable threshold. And then both the operational definition of unknown hazardous scenarios and the appropriate threshold remain areas of active discussion within the industry and society.

What they’re basically saying is They don’t know what the hell they’re talking about. They’ve got no basis, no statistical or numerical basis for saying what safety is, but that’s okay because it’s still under discussion. What is their underlying their assertion that this is about safe operation?

It’s really. Astonishingly vapid so skipping right to the end here. They say until the completion of safety validation to sign off and safety. We operate with a trained safety driver when testing AVs on public roads. That’s good. And then it considers the functional safety of the AV system and evaluates [00:28:00] whether scenario based testing and statistical safety testing adequately demonstrate the absence of unreasonable risk.

Again, absence of unreasonable risk is not a design standard. It’s a legal concept that just says, There’s no basis for suing us, right? A design standard is not a void. You can’t say, I’m going to design to nothing. You have to design to something. And so this is very important. We need to move away from this whole idea that absence of unreasonable risk within this industry, which is what the ALUs is a reasonable approach.

Their partners in this nonsense include Hyundai, Lyft Uber Eats via the AVSC and AVIA. But they don’t focus on affirmative safe operation and none of them do. And also a gentle reminder here that using Uber Eats. Seems to indemnify [00:29:00] Uber against damages if you’re subsequently injured while using Uber ride share services which is another interesting side effect to this.

So for all that, but basically motionals statement that their safety claims are based on nothing at all. I have nominated them for my gaslight of the week award.

[00:29:23] Anthony: Excellent. I have some follow up questions for you there, Fred. Emotional says that they’re safer than a human, right? They don’t say that.

[00:29:31] Fred: They imply it. They ask, they really threw a clever word and they ask you to infer

[00:29:36] Anthony: that. Okay, so they imply they’re safer than a human. But hey, we’re keeping a human behind the wheel, just in case. So they’ve got a human that’s better than their software behind the wheel, because if not, what’s the point of having a human there at all?

Stepping back from that even, they do the classic trifecta of, we will never drive drunk, drowsy, or distracted. Okay, I [00:30:00] can give them 100 percent all these companies that no computer will ever be drunk. Without a doubt. I think we can all agree on that. Drowsy? Drowsy is you’re falling asleep. You’re losing attention.

Everyone who’s ever used the computer knows these, they’ll get drowsy. They will definitely, especially over time, your computer slows down. It doesn’t get faster over time. It gets less reliable over time because, entropy. Distracted? Come on, every computer, everyone who’s ever seen that spinning beach ball or that hourglass thing that just freezes in front of your face.

Those computers get distracted. I don’t know what the hell they’re doing behind there. It doesn’t know what’s doing behind there. And it’s just Hey this application’s stuck. So don’t, you can have your drunk, but you’re drowsy and distracted. That is not just a human quality.

Okay. I’m sorry.

[00:30:57] Fred: No argument. Yeah.

[00:30:59] Exploring the Concept of Drunk Driving

[00:30:59] Fred: But I want to [00:31:00] explore drunk a little bit. What does it mean to be a drunk driver? It means that your reactions are slow. And you’re making decisions that are less than rational. Those are the problems with joint driving because if you were making rational decisions and you were had normal reactions, you wouldn’t say you were drunk.

[00:31:18] AI and Drunk Behavior

[00:31:18] Fred: So the question arises, does a computer ever act as though it is drunk? Does it ever make irrational decisions? We all know AI examples that are really crazy. So that happens.

[00:31:31] Anthony: The way mo going down the other side of the street. Yeah.

[00:31:34] Fred: Things, little things like that and zigzagging bone when they’re traveling behind a trailer and certainly acting drunk.

And we also know of a lot of situations where. The computers have very slow reactions. In fact, if you’ve ever used a computer, you’ve probably experienced the computers locking up, or applications within the computer just failing to respond. Even though they’re not drunk [00:32:00] chemically, They certainly can act drunk or with equivalent behavior.

So yeah, anyway, but going back to the original, they ask you to project your own capabilities into this and say, it’s just like you except better without ever clearly stating it. So

[00:32:17] Anthony: I, I love it. I want to be just like me, but better too. All right, Mr. Brooks, what’s your gaslight of the week?

[00:32:24] Gaslight of the Week: Tesla and NHTSA

[00:32:24] Michael: Mine is pretty simple, going back to a story we discussed earlier, which is a classic example of gaslighting.

In this case, it wasn’t necessarily the public that was being gaslighted, which is normally the subject of our gaslights, but it was NHTSA with Tesla coming into the NHTSA office or briefing them online. I’m not sure how that happened. But basically. Basically. A meeting was with between Tesla and this is investigative staff.

And Tesla says, look, we’re doing all this. To make sure that drivers are paying attention to their vehicles while they’re in, while they’re in full self driving [00:33:00] mode. We have messages in our owner’s manual. We’ve got messages in the user interface and the vehicle and we have messaging online on YouTube to drivers that the vehicle is not autonomous and that they have to, The driver must remain in full control all the time.

Meanwhile, Tesla on the side is going online and supporting posts on Twitter that. say the exact opposite that, that, that highlight really poor uses of the technology that, drivers are saying, look, I can take my hands off the wheel and sit back and have a little rest and not get fatigued.

And Tesla’s supporting that type of behavior on Twitter. So that’s a classic example of gas lighting. And I think it fits great as my gas light of the week.

[00:33:47] Anthony: Excellent.

[00:33:48] Gaslight of the Week: Elon Musk

[00:33:48] Anthony: My Gaslight of the Week is related to a story we just discussed too. I was so afraid Michael was going to steal mine, but it’s Elon Musk!

Gaslighter of the century. From that Washington Post article where he’s talking about [00:34:00] making government more efficient. Elon also says he wants to see federal spending cut by two trillion dollars! Keep in mind this is a guy whose entire companies are based off of dining on the government teat Tesla made what?

800 million from selling carbon credits, which is just like selling baseball cards. Bullshit.

[00:34:20] Michael: I can assure you that won’t be, that will not be cut for the budget.

[00:34:23] Anthony: That will not be cut. No he gets it. SpaceX gets how much money for launching satellites and stuff out of NASA’s pilot. Yeah, that we’re not cutting that.

That’s not getting cut. No way. What will get cut is a bunch of things he doesn’t like hey Stop calling BS on my lies the around, so the SEC, yeah, that’ll be cut, cause they I don’t like them. NHTSA, yeah, that’s gonna get cut and but the EPA, yeah, cause, I can’t just dump rocket fuel wherever the hell I want.

But that’s where you’re at. But that’s not going to add up anywhere close to 2 trillion. So

[00:34:58] Fred: Wait, come on. No, I’m sure [00:35:00] that these people are going to be very aware of conflicts of interest. And we all know that there were no conflicts of interest in the last Trump administration. We can expect that same high quality of ethical behavior in the upcoming administration.

[00:35:15] Anthony: I would like to change my gaslighter of the week to Fred Perkins gaslighter. Wait, who’s being cynical now, Anthony? I’m not sure what that word means still. Ah, okay, let’s go on thinking of a cynical California Public Utilities Commission.

[00:35:32] California’s New Reporting Requirements for AV Incidents

[00:35:32] Anthony: My favorite of the Public Utility Commissions. An article in Reuters we have a link from, in MSN.

com, titled, California Agency Boosts Reporting Requirements for Autonomous Vehicle Incidents. That’s great! California State Agency said on Thursday it is mandating enhanced data reporting requirements for autonomous vehicles, including reports for incidents where self driving cars get stuck. Thank God!

I can’t believe that word! They’re like, when they drive into wet cement, when they [00:36:00] have their existential crisis in the middle of intersections, that wasn’t included in their reporting. I am all for this. Please tell me why I’m wrong.

[00:36:10] Michael: No you’re right. That’s a End of show.

[00:36:13] Anthony: End of show.

We’re done.

[00:36:14] Michael: That’s I am right. That’s a big bullet point. There’s basically three main areas where they’re going to be beefing up reporting in California on a V. That’s the kind of the number one highlighted when it’s stoppage events is what they’re calling it. So all the A. V. Operators have to report any incident where the A.

V. S. Get stuck or have a breakdown or I think we described it as a some type of crisis of movement at some point. I can’t remember exactly, but they’re, that’s one thing. Those are stoppage events where we saw, buses, fire trucks, emergency responders of other sorts get trapped on roads behind a vehicle that was stopped in the middle of the road, having an identity crisis.

Those are going to be reported to the [00:37:00] California Public Utilities Commission. Now also they are enhancing their incident reporting. So everyone operating AV has to report on both collision and non collision incidents including, traffic tickets and other things. So they’re wait, do they

[00:37:17] Anthony: get traffic tickets?

That’s

[00:37:19] Michael: apparently. I don’t know. They have to report them. It says including citations and stoppage events. See,

[00:37:27] Anthony: that’s the thing I was confused about because they said including citations but who is citing them? And I thought California exempted these guys from basically all traffic laws.

[00:37:36] Michael: No, I don’t think so. No, I think they could still get tickets. If you’re blocking a road, for instance, maybe a citation is, a parking ticket because they parked in a no stop zone. Who knows?

[00:37:47] Fred: I think what’s happening is that these tickets are not being issued to the individual vehicles.

The citations are being given to the company as an advisory. So it’s not the sort of thing where they’re going to find [00:38:00] the individual vehicle 20 for a violation. They’re just sending a note back to the manufacturer saying please be aware that your vehicle did something bad. I think that’s what’s happening.

[00:38:12] Anthony: But do we actually know if that’s actually happening now? Cause that seemed to be, cause I remember we had the sheriff on a couple of years ago. He’s I don’t know who we’d ticket. And I spoke to a former NYPD officer and he said, yeah. He’s who would we give a ticket to? He’s there’s no way we’re going to a, the GM’s office being like, here you go, here’s a 150 parking ticket.

[00:38:32] Fred: I can’t cite the reference, but I think I read somewhere that there is a mechanism for feeding back that violation information to the manufacturer, but I’m sorry, I don’t have the details.

[00:38:44] Michael: And in some cases, most of these vehicles that are on the road now, at least have a safety driver in them. And I’m presuming that the safety driver is going to be the recipient of the citation.

[00:38:55] Anthony: Not Waymo’s, they don’t, they got no safety drivers. No. So [00:39:00] I, yeah. So listeners, if you guys know this, ideally somebody who’s the, who’s upset, he used to work for Waymo or Cruise, who wants to blow a little whistle, Bye. You don’t even have to blow a whistle, just tell us no.

There’s no parking tickets, or yes, we have a stack of unpaid parking tickets. Because I think and listeners correct me if I’m wrong, but I don’t, this has always been the thing where I’m like, who’s getting this? So anyway, I’m sorry, Michael, I was interrupting you.

[00:39:23] Michael: No, there’s only one more area.

They’re up there. They’ve got the stoppage event reporting, enhanced incident reporting, and also they’re calling streamlined collision reporting. A. V. operators are going to have to report all collisions, both to CPSC. CPUC, the Public Utilities Commission, and NHTSA, within one day of a collision.

So that, that gives them, quicker access to information about collisions. And allows them to, respond more quickly when those collisions have taken place to, to gather more knowledge before it disappears. All

[00:39:55] Anthony: Ideally a step in the right direction. And hey, if you donate, we’ll [00:40:00] find out if they’re actually giving tickets to these companies.

That’s right, that’s the only way we’re gonna do work is by soliciting donations. No, that’s not true. We’ll try and hunt this information down. But for now, it’s time to enter the world of Mr. Fred Perkins. You’ve now entered the

[00:40:13] Fred: Tao of Fred. The Tao

[00:40:14] Anthony: of Fred.

[00:40:15] Fred: And he’s gonna talk to us about

Thank you, Anthony. Thank you for that introduction.

[00:40:23] The Tao of Fred: Duty of Care and Risk

[00:40:23] Fred: So this week, it’s a little bit complicated, but I want you to try to hang on as we go through this. So there is a relationship between duty of care that we’ve talked about and there’s, and risk and also hazard, but there, there are separate things.

As I understand it, and Michael slapped me upside the head if I got this wrong, that the rules of the road are really informed by the duty of care. Every driver assumes the mantle of duty of care when they operate their vehicle legally, and they’re expressed in the rules of the [00:41:00] road. Whether or not a driver’s vehicle is safe enough, To operate is consistent with the rules of the road.

In other words, if if you do everything right, and you’re within the rules of the road, the vehicle is safe enough from the perspective of the legal system and the harm caused by violation of the rules of the road that involve the duty of care. They provide cause for law enforcement actions and judicial actions, including liability and damages.

I think I got that don’t. Is that right? Yeah, Michael’s nodding his head. Taking that a little bit further, those ideas provide a basis to consider the risk of harm. Now, risk is usually expressed as a two parameter problem. So you’ve got both the consequence of the harm, right? Something’s going to bang into something else.

But you’ve also got the probability that the harm will be encountered. It would be really bad to have a meteorite land on your [00:42:00] car, but the probability of that happening is very low, so you’re not going to worry about that, right? So risk mapping is a usual way, useful way, I should say, of managing resource allocation for prevention of harm.

So that’s, unfortunately, that’s it. The approach that the A. V. company is using to design the safety of vehicles, reliance on risk, so that allows them to say that these things are really unlikely to happen, so we don’t have to care about it. How likely is it that you’re going to run into a woman crossing a road in the middle of the night, walking a bicycle?

Oh, it’s not likely to happen at all. Yet, Ms. Hertzberg in Arizona unfortunately lost her life because that was not an anticipated event. I think that the A. V. companies prefer to consider risk rather than [00:43:00] hazards because it’s a lot easier and they convince themselves that there are certain things they don’t have to do.

Now, there’s a considerable basis of experience now on cars driving themselves. We have a lot of experience from NHTSA, or excuse me, from Tesla, and we also have a lot of documentation associated with that, both in video form and reports that they’ve given to NHTSA. All right, through the what was it? The standard general order standing general order.

There are also a lot of events that have been documented by way. Mo a lot by cruise. When an operating encounters a hazard. Risk isn’t really important because the probability of it happening is 100%, right? They’re actually there encountering the hazard. Really what the vehicles should be doing is they should be designing for hazard mitigation.

In every event that they know of, or hazard can [00:44:00] occur, or has been demonstrated to occur and then folding that into their risk evaluation. The operational AV tactical decisions. When they’re operating, they’re all made on the basis of hazard evaluation, not risk. They see something happening, they respond to it.

That’s different than designing for risk, because when you design for risk, you can convince yourself that that’s really unlikely to happen, so I don’t have to spend any resources to accommodate that. Right now, there’s a long and unfortunately rapidly growing list of what demonstrated hazards are for AVs and also hazards that people anticipate reasonably could be happening.

And my belief is that the AVs should be designed to avoid every 1 of the hazards that have been identified. And every 1 of the hazards that have been identified, either conceptually. Based on the [00:45:00] designers engineering judgment, or in practice, based upon the experience of the operating on the roads, right?

This involves conflicts with pedestrians cars going off the road. It involves. AVs beeping at each other, stopping at intersections. So it seems to me that it’s a very simple process to catalog all of these different hazard events and verify that every AV that’s allowed on the road is shown to mitigate the hazard associated with these events rather than to worry about the risk, for example.

And to be the equivalent of a human driver, AVs The computer actually has to be able to feel the pucker factor that would a human being would feel if they saw, for example, a trail of infants walking across the street behind a daycare teacher, right? [00:46:00] All tied together with a rope, you have a different human response to that than you would if you saw a string of adults crossing the road.

So that the pucker factor. The pucker factor is something that would come out of a hazard mitigation focus rather than a risk mitigation focus, right? Regression right now in the industry is really regression associated with hazard mitigation is oriented towards. What they call a minimal risk condition.

Now, we’ve talked about this before. It does not confer safety, but it gives the AV developers coverage to say I’m working towards a minimal risk conditions written right here in the regulations that I have to manage towards a minimal risk condition and then everything is okay. That’s really fundamentally wrong because you don’t want to manage risk in the AV industry so much as you want to manage the hazards.

That these AVs create, have [00:47:00] encountered, and are associated with their operation. This is a little bit of inside baseball, but I think it’s very important, because what the AV designers should be oriented towards is designing for an affirmative, acceptable safety operational standard, rather than designing towards a void that allows them to evade legal responsibility for whatever damage Their vehicle may cause evading responsibility for damage is what a focus on a risk allows them to do designing for hazard mitigation and demonstrating that they can, in fact, evade all the hazards that are known.

There’s a much better approach towards safety and allows them to, in fact, emulate what a human being would do in a similar situation when confronting that same hazard. Again, inside baseball, but I think, for the industry to succeed, [00:48:00] the hazard mitigation has got to be addressed 1st, and then the compliance with the duty of care and the rules of the road and the risk considerations should be last in this train of analysis associated with safe operation.

Does that make sense? And it could be a little confusing, but I think that it’s a little bit of a different approach than people have used in the past.

[00:48:23] Anthony: It’s a, it’s dense, but yeah. So are you talking about these hazards of, you’re saying, okay, we can collect a list of these hazards and would this be something you’d suggest for, hey, these AVs need to pass this as part of their driver certification their AV driver’s license?

[00:48:41] Fred: Yeah, I think so. For example the versus. Whatever has happened that has caused AVs to either kill or injure a human being, or to impact or cause property damage in interoperations should be cataloged. And the vehicles, [00:49:00] before they’re put on the street, before they’re allowed to be put on the street, should have to demonstrate that they can evade these or mitigate these hazards in a test track.

With reasonable statistical confidence, right? I think that’s the minimum that should be done. I don’t think that AV developers should be allowed to prevaricate and just say it’s not going to happen. It’s unreasonable if this would happen twice. We’re just going to say that the risk is very low.

Remember, risk has two parts, the probability of it happening and the consequence. Of the event, right? As long as you’re in a risk dominated analysis of safety, you can always say that it’s unlikely to happen. And this is where I think that NHTSA and state regulators should jump in and say you may say it’s unlikely, but, this person is dead, or that event happened and this damage occurred.

You got to show on a test track and you got to show in simulation, you got to show in every way possible. Or [00:50:00] practical, that this is not going to happen if we, the regulators, give you authority to operate this vehicle on the road, on the highway. When you’re operating on the highway, you’re not in a risk avoidance environment, you’re in a hazard avoidance environment, because the probability of something happening when you see it is 100%.

Because you’ve seen it, you’ve seen it developing. So I think that the, that this is very important for the regulators to consider. And before they give authority to developers to operate on the highway, they really need to show that everything known that can cause damage. By a V’s has been accommodated and the safety design.

This is not happening now. And people are still being allowed to put the vehicles on the highway by waving their hands and saying, that’s really unlikely to happen. How could it [00:51:00] ever happen twice?

[00:51:01] Anthony: They won’t be drunk, distracted, or drowsy. That’s right. Yeah. Hey, state regulators, because I’m giving up on the federal level for a short period of time.

State regulators, at least some states, pay attention to this because if you’re a state, say one of those third world country states like California and New York, You can say, hey, wait a second. Let’s not put this stuff on the road. Okay, not if you’re California. If you’re New York, if you’re New York, you can come out and say, yeah, we’re not putting this on the road cuz I want to get reelected or I don’t want to kill people.

[00:51:30] Fred: Take one of those two. Yeah, I’d vote for Massachusetts too because I happen to live here.

[00:51:35] Anthony: Yeah, but that’s some third world place up there, Massachusetts. Speaking of third world places, visit autosafety. org and click on donate. What? That was a weird transition. Why would he say such things?

[00:51:46] Michael: I don’t even know if the use of the term third world is allowed anymore, Anthony, so you might be chastised for that.

[00:51:52] Anthony: It wouldn’t be the first time. I’m just quoting our dear leader. Anyway, let’s time for some recalls.

[00:51:58] Recalls: GM, Chrysler, Tesla, and BMW

[00:51:58] Anthony: Let’s get

[00:51:59] Fred: [00:52:00] recalls. First up,

[00:52:01] Anthony: General Motors. Oh my god, 461, 839 2021 Cadillac Escalade ESV, Chevy Silverados and Chevy Suburbans, and Chevy Tahoes, and GM Sierras, and the list goes on and on.

This is a transmission control valve, may be susceptible to excessive wear over time, resulting in a gradual loss of pressure within the valve that can cause harsh shifting. In rare cases, the real world may experience a momentary lockup. Oh boy.

[00:52:34] Michael: Yeah, this is actually, it notes in the the submission by GM that this is yet another success coming out of the GM speak up for safety program which, gives their engineers

[00:52:45] Anthony: a quick way to resign.

[00:52:47] Michael: Yeah, no, it gives them incentive to speak up about problems that they notice, whether they’re on the line or, even if it’s reported by a customer. That’s a success and it’s a lot of [00:53:00] vehicles. It looks like, it’s going to be owners. It looks like a staggered notification system on this 1.

so there’s, it’s a large recall. So they’re probably not going to be able to fix everything at once.

[00:53:12] Anthony: They’re doing it based off what you paid. So the Escalade, they’re getting notice first and then the Suburbans, they’ll get notified last.

[00:53:19] Michael: It may be that if you pay the most for a big Cadillac, you might be getting the first notice, but they’re going to start notification December 9th and the last phase of their owner notification is going to start around the end of February.

So if you do own one of these vehicles, it’s. It may be, three months before you hear from GM regarding the software update.

[00:53:41] Anthony: Excellent. Next one, Chrysler 206, 502 vehicles, the 2018 to 2019 Jeep Grand Cherokee, the Dodge Durango’s is a hydraulic control unit slash anti lock brake system module.

Brake light may [00:54:00] may come on when you haven’t touched your brakes.

[00:54:03] Michael: This is also this problem allows the vehicle to be shifted out of park without holding the brake pedal, which is the brake shift interlock system that vehicles are required to have in place to prevent vehicles from rolling away.

So essentially this is a rollaway problem. The vehicles. If this condition occurs you can shift out of park without depressing the brake pedal. And so unintended vehicle movement can result from that. Vehicles can roll down driveways and other areas and parking lots and hurt people. So that is the that’s what’s happening here.

The remedy for this. under development. So we may not see it. It looks to be a software issue because it’s an it’s a heck you ABS module that’s not reading pressure properly on the circuit. I’m assuming they’re going to address this with a software fix of some kind, but they have not been able to come up with that yet.

So it may be a software issue. [00:55:00] some time before owners of these vehicles are able to get a fix. And so we’d advise everyone who owns, I think it’s the 2018 2019 Grand Cherokee or Dodge Durango, to be really careful with your shifting in and out of park for the next few months while you wait for that fix.

[00:55:18] Anthony: All right. Tesla up next. 2, 431 vehicles, the 2024 Tesla Cybertruck. It turns out if you’ve bought this vehicle, they want you to return it, and they want to slap you in the face. Cause you’re a silly person. No, this is a drive inverter, may cause it to stop producing torque. Whoa! Oh, it’s one of their MOSFETs have have failed?

Really? MOSFETs are like, you can get them anywhere, they’re, they’ve been making them forever. This

[00:55:49] Fred: really goes back to what you were talking about earlier with drunk computers, right? So computers are dependent upon the analog performance of the devices that support their computer [00:56:00] calculations and a MOSFET is basically a switch that turns electrical current off and on.

So if the MOSFET fails, it’s a, in an analog way, meaning it just doesn’t work anymore, then the inputs going to the computer are faulty and it’s a lot like being drunk, right? It just doesn’t get the information anymore that it can use to calculate responses properly.

[00:56:24] Anthony: MOSFETs, what, they first came out in the 60s?

Is that right?

[00:56:28] Fred: I’m not sure. They basically were developed to replace relays.

[00:56:33] Anthony: Yeah. They’re, they’re, you can buy like a thousand of them for a dollar. I think they’re, now, okay, that’s an exaggeration. I know someone’s going to get very upset, be like, Ah, big MOSFETs in my kitchen.

They’re very inexpensive is what I’m trying to say. They’re very kind of a standard part. They’re not.

[00:56:51] Fred: But the ones that can handle the high current for. Electro propulsion tend to be more expensive, to be fair.

[00:56:57] Anthony: They’re 2. They’re not, it’s not something where you’re [00:57:00] breaking the bank.

It’s not rocket science.

[00:57:03] Fred: It is not rocket science. It’s electrical engineering.

Ha

[00:57:06] Anthony: Ha. Anyway Tesla, the you’re gonna be noticed shortly, 9th. And this

[00:57:14] Michael: is actually January 4th is when it looks like the owner notifications are going out but they’re going to one of the other, this is the 6th recall the Cybertruck and you’re going to have to take your vehicle into the dealer for this one because it does require a physical repair.

So there will be no over the air update for the Tesla owners on this one.

[00:57:35] Anthony: Oh, this is just a a little aside for a Cybertruck owner. So a lot of people, there’s stainless steel and people think they won’t get stained and not how it works. Anyway they’ve been putting these vinyl wraps on their cars, thinking that’s going to protect it.

Where a couple of people, they just said, Hey, we’re to take the wraps off, really give a real clean down. And they take the wrap off and they go, ah, oh my God, this is rusted. What’s going on? Why is this? [00:58:00] Stuff rusted. So if you’re a a cyber truck owner I’m sorry. I, like I, I’m falling for all sorts of

[00:58:08] Fred: things.

A couple of words about chemistry. The way the stainless steel works is it develops a coating of chromium oxide on the outside that is shields the rest of the structure from oxygen. Unfortunately, when you wrap it in plastic, you’re prohibiting access to oxygen. So when water gets in. It causes it to rust because the chromium oxide coating can never develop.

So that’s why new stainless steel is shinier than old stainless steel, because the old stainless steel has got that coating. And that’s why you should never wrap the stainless steel in something that’s going to prohibit invasion of oxygen because that oxygen. Oxidation

[00:58:54] Anthony: Zappa, give me the chromium. Last up, BMW, [00:59:00] 3, 312 vehicles, the 2025 BMW.

Why do they name their cars these stupid names? X1 xDrive 28i, X1 M35i. This is some sort of SUV vehicle. The supplier produced B pillar reinforcement plate is out of specifications. Oh, and it may not have enough structural integrity to survive a crash. That’s that’s not good.

[00:59:24] Michael: No, that’s really bad.

They’re still developing a fix for this. So it doesn’t look like this will be another one where owners are probably going to have to wait for some time. But this is a concerning when your B pillar reinforcement plate is not supported properly. It looks like they’re. That the problem was that the materials that were that came in that made up the B pillar reinforcement plate were out of spec.

So they essentially, the metal or whatever they used to the B pillar support plate. was not what they were expecting. They got a bad [01:00:00] batch from their supplier, and so it makes the plates crack and fail. So that’s something that owners should keep an eye out on because, it’s a significant problem.

If you’re in a crash where your B pillar is required to support the vehicle’s occupant cabin,

[01:00:17] Outro and Final Thoughts

[01:00:17] Anthony: and that’s it for Our show today for your recalls, for our MOSFETs, for our Elon FETs and for our Boba FETs and babbling FETs. Yeah, I’m just going to keep, this is the outro music playing. Thank God.

All right. Till next time. That’s not the outro music at all. All right. Thanks listeners. Till next time. Bye. Thank you

[01:00:39] Fred: for listening. For more information, visit www. autosafety. org.