Big Cars, Bigger Problems and Driverless Dangers

[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 the 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. Welcome to the

[00:00:26] The Honking Debate

[00:00:26] Anthony: sound that no one can hear except for me, which is the loud honking in the background. I don’t know why I would share that with you people because honking is good. It’s how you let pedestrians know that you’re gonna hit them. In your giant SUV. Wait, is that why you honk?

[00:00:40] Michael: Now, honking’s good, right? But the only one that’s used appropriately, for instance some taxi driver was behind a car that was turning right across a crosswalk a few years back when I was in D. C., but there were people crossing the crosswalk and he was honking to get this car to essentially drive into those people so he could go on his merry way. [00:01:00] I guess honking can be good, but honking can also be, it reflects. The wishes of a driver who is imperfect. So honking’s imperfect.

[00:01:09] Anthony: That, that reminds me of an episode, I don’t know how many, maybe a dozen back where I told you guys about where I was going to make a turn, a left turn, but there’s pedestrians.

And so I gave, the guy behind me honking a nice little friendly gesture and he threatened me. Is he following down back streets? So anyway, if you’re listening, cause I’m not in the road right now, knock yourself out. Okay.

[00:01:32] Michael: Yeah, I’d also suggest that gestures are an imperfect reflection of your thoughts and maybe you could handle that in a better way in the future, as you age and mature.

[00:01:41] Congress and Car Safety Legislation

[00:01:41] Anthony: So speaking of aging and maturing, let’s talk about Congress. So we have a great article linking to from NPR. And we’ve all noticed for, I don’t know, 40, 50 years that cars keep getting bigger. SUVs their hood heights [00:02:00] keep getting larger. And the little car that I drive, it stays the same. It doesn’t get any bigger.

And when you crash these two big car wins. Big car crush small car. It’s bad. And so the Insurance Institute for Highway Safety said, oh yeah, we should probably update our test stuff. And so they added another 500 kilograms of weight. Which is roughly for you. At home over a thousand pounds.

[00:02:26] Pedestrian Safety and Vehicle Design

[00:02:26] Anthony: But now, lawmakers in Congress are expected to introduce a bill that would require federal standards for hood height and visibility to protect pedestrians and other vulnerable road users.

Oh my god, is this gonna happen?! We’ve been talking about this since like episode two.

[00:02:44] Michael: This is legislation that was introduced by, one of the more forward thinking congress people on safety, Mary Gay Scanlon. It’s legislation that we certainly support. Obviously, every bill that goes into congress these days, you have no idea what the outcome [00:03:00] can be.

And most safety, vehicle safety, Bills are not passed, but they can be included in the infrastructure act every four years So it’s got a chance that way as well. But the the bill that’s introduced here. It’s specifically focused at Pedestrians not just you know, the vehicle weight problem, which is also significant in car to car collisions But you know the hoods bumpers the heights The weight of vehicles are all factors that can compound the damage that’s done to the human body when it’s struck or to bicyclists and others, other vulnerable road users.

And, the piece of legislation addresses that a number of ways, it’s essentially intended to develop safety standards that protect pedestrians by, Requiring NHTSA to study weights, heights, hood and bumper designs of vehicles to determine, what can be done there and to issue rulemakings to make those vehicles safe for pedestrians.

Also, we’ve [00:04:00] discussed a lot on our podcast about, The visibility problems that come when you have these larger vehicles that have a, a smaller range of vision of objects that are lower in the road. We’ve probably talked before about just how many children you can line up and some of these in front of some of these vehicles without the driver being able to see them.

You can line up 10 or more children in front of some of these cars before drivers even able to see a child in front of the vehicle. And, we’re, the bill also requires NHTSA to put it NCAP testing out to make sure that vehicles have good visibility in addition to a visibility regulation that requires minimum standards around visibility.

So the bill does a lot. It’s a bill that we certainly support and it’s I, ultimately if passed into law and if NHTSA can get a rulemaking through, it can play a major role. In reducing pedestrian fatalities, which we’ve seen rise significantly, particularly over the last five years or [00:05:00] so,

[00:05:01] Fred: Michael, would this just catch us up with Europe or does it do something that exceeds what is already the case for your euro and cap?

[00:05:10] Michael: It’s going to be something that exceeds that to my knowledge, Europe has some hood and bumper standards and protective hood standards to protect pedestrians, but this bill allows NHTSA to go beyond that and to do, I, it would depend on, I don’t know if NHTSA is ever going to Tell manufacturers they have to reduce the weight of their vehicles.

I think that would be very tough for them to do it’s something we wish they would do but it’s something that manufacturers are going to fight because they’re interested in preserving vehicle choice for all americans Which is basically their way of saying we want to be able to sell the heaviest vehicles possible If that’s what consumers want to buy So this is you know, like many problems In America, this is one of those where, [00:06:00] some people would rather have the ability to drive massive vehicles than, have to worry about things like other people dying in the streets.

And it’s an area where the government needs to step in to put a halt to the problem and so far has been very, unwilling to do despite, many studies bearing out that, the massive kinetic energy in the today’s cars is posing a threat, not only to pedestrians, but also to people involved in vehicle to vehicle crashes.

We would hope that some of the results of this bill and the rules that could come out of it would also have a spillover effect onto the general vehicle weight problem that we have discussed Many times in the past.

[00:06:43] Anthony: So look, as somebody who plans to buy a Hummer EV, I’m conflicted because clearly I care about the environment, but I don’t care about other people.

And that’s why I want to drive a 9, 000 pound death machine so I can run over other people and not see them.

[00:06:59] Fred: There’s another [00:07:00] solution which would be to limit the drivers in the big SUVs to only people who are over six feet tall. ’cause that, that elevates your eyes and you get a much better angle.

And you could probably see kids lined up, oh, I don’t know, only six in a row rather than 10 in a row.

[00:07:17] Anthony: I see how this is. So you’re forbidding me from getting a Hummer, ev I get it. It’s your vertical

[00:07:23] Fred: challenges are your own Anthony? That’s, there’s not something I can help with.

[00:07:28] Anthony: When I first ran into you at the carnival, you were upsetting and you’re upsetting me now.

[00:07:33] Fred: No, but seriously, this is a wonderful progress in Congress. We’re very glad to see this bill and hope it gets passed into law soon.

[00:07:41] Anthony: Yeah. And surprisingly enough, the Alliance for Automotive Innovation, The auto industry trade group we’ve talked about many times here say, we don’t need laws and regulations.

We voluntarily make cars safer. Don’t you worry? We’re good here, man.

[00:07:58] Michael: I wish I’d seen that one. I might have [00:08:00] used that for my gas light of the week, but I’ve got someone else.

[00:08:03] Anthony: It’s in there. I’m not even using them for my guests. Cause it’s just too easy. They’re like, GM Cruise. Ha! So Michael, what’s the timeline on this?

So this gets, this is introduced in the house. It will Sit there for a while and ideally move forward.

[00:08:20] Michael: Ideally there may or may not be a, hearings on the issue. Something, I don’t expect a lot to be moving in Congress over the next couple of months with the election coming up.

I’m not, I think that’s a given in election years. We’ll see what happens. We may not see this language, come up in Congress again until the next infrastructure bill. It’s hard to predict what’s going to happen on Capitol Hill.

[00:08:43] Anthony: But I think it’s good for consumers.

The people listening to this show, at the very least, the next time you have to go buy a car or whatnot, take an extra second and be like, Oh, do I need that giant wall? Because it may make you feel safer, but if you can’t see in front of [00:09:00] you, is it making you safe? Just my thought for the day.

[00:09:06] Ford’s Blue Cruise System Issues

[00:09:06] Anthony: Let’s move on to Ford and their Blue Cruise system.

We’ve talked a lot about this. Blue Cruise sounds like a great idea. Pre mapped roads that allow you to take your hands off the wheels, it will drive for you. Apparently, doesn’t work very well when it’s night. Like a lot of these systems. There’s a woman in Philadelphia who was intoxicated and using a partially automated driving system when she caused her a March highway crash in Philadelphia that killed two people.

She was driving a Ford Mustang Mach E SUV and struck a stationary vehicle of a man who had stopped on the left shoulder of the road. It seems that these systems just all have a major problem working in low light conditions. This is surprising. We see this with Tesla, but they only want to use cameras.

But the Ford, as my understanding is, it uses the radar as well, and that shouldn’t [00:10:00] care if it’s day or night.

[00:10:02] Michael: And Ford also uses, geo fencing to ensure that these vehicles are only operating on a controlled access highway. So you don’t see a lot of the problems with Tesla autopilot or full self driving being able to be engaged on roads where it’s not appropriate or hasn’t been tested to ensure safety.

So even with those precautions, we’ve seen, two crashes that we know of so far involving blue crews. And this one, two people were killed. And now the the woman who was operating this vehicle in Philadelphia has been arrested for vehicular homicide or charged with vehicular homicide. And, that should be a warning, a warning to everyone who’s using Partially automated vehicle systems, adaptive cruise control, that kind of thing, that kind of thing that the burden is going to be on you when there’s a crash, there’s no state laws that are going to [00:11:00] protect you simply because you’re relying on a system created by Ford Motor Company.

And, complicating the issue here is, the fact that the woman who did this was also it looks like she was drunk or intoxicated at the time, and was using Blue Cruise to get home or to get to her next stop on a night full of bad decisions, but it’s, A real problem, and for us, I think the bigger problem, other, there’s obviously loss of life here and there’s a person who, while they were driving drunk, obviously didn’t intend for this result to happen, but the real issue is, you can, the way these systems are sold, you can, over time, become more reliant on these features and maybe, Less your you won’t, you become less willing to monitor the vehicle’s performance closely after it works for, a week or a month of doing this type of behavior.

But when it does fail, the [00:12:00] consequences can be tragic as they are here. And, as a consumer, none of us are safe. from being charged with vehicular homicide or any number of other criminal offenses. When the technology fails. So just like Tesla, it is a trap. It’s not quite as obvious a trap as some of the Tesla systems are.

Because, like Anthony referred to, The blue cruise does, they do spend additional money to put better technology into the vehicles to prevent these type of incidents from happening. Nothing is perfect and these systems certainly aren’t perfect yet. And in fact, there’s been recent research confirming that a lot of these partially automated systems that include advanced cruise control aren’t.

Going to work effectively all the time and that humans need to be ever vigilant when using these systems.

[00:12:49] Anthony: This is a good example because this is one where it’s not only just using, cameras and radar, but it’s on a limited access highways. It’s, these are [00:13:00] pre mapped roads and this accident happened in the left shoulder.

I imagine the people who map these roads know where the lanes begin and end. They have those dimensions, they have where the left shoulder is. And the fact that this vehicle drove into this road. The left shoulder shows a failure on another level with mapping or are they just, or what’s the level of granularity they have for these maps?

Standards exist,

[00:13:28] Fred: so the companies can do whatever they want. And, it’s cheaper to make a lousy map. I’ve heard discussions, technical discussions about how far over the white line on the side of the road. It’s allowable to let an AV go. Is it 1 inch? Is it 4 inches? Is it 6 inches?

My position is zero inches is the right number because there are a lot of places where, for example, a white line marks the edge of a bridge [00:14:00] or abutment, and if you go over the white line by any amount, you’re in trouble. But again, no standards exist. No standards exist for this. No industry consensus standards exist and certainly no government standards exist saying how accurately roads have to be mapped before they’re being used by either level three or level two or any kind of automated driving system.

[00:14:27] Anthony: That’s a good point. Okay. I want to jump. I apologize. I’m going to jump a little to a rabbit hole here for a second. Are there federal standards on lane markings? Is there a standard to say, Hey, the white line has to be. X feet from the center line. Michael’s not as a kid. Yes.

[00:14:45] Michael: I believe there are there state and federal standards that dictate a lot of things.

We’ve looked in this year of guard rails recently and how they’re impacted by heavier weights from electric vehicles. And as part of that, I was looking through the Virginia manual on [00:15:00] how to install guard rails. And it is. It’s an immense publication, and I’m sure the same things exist at the state and federal level, the Federal Highway Administration and others who are involved in, basically building out our infrastructure to be as safe as possible.

And when you’re driving on the interstate, you’ll notice that there it’s. Fairly uniform, unless you’re driving through South Carolina or Alabama, like I did this past week. So there are places in America, maybe Philadelphia is one of them where, maintenance hasn’t occurred or there’s road construction going on.

There could be changes that have occurred since. Forward map the area if that’s what they’re doing. So there’s always some variation there and a good system would be able to detect such variations and keep Passengers and occupants in the vehicle safe because it’s obvious and it should be obvious to anyone designing these systems that roads and infrastructure aren’t always going to be perfect.

And if these [00:16:00] cars aren’t working on somewhat imperfect infrastructure, then they’re going to be a safety risk.

[00:16:06] Anthony: Listener. You may wonder sometimes. Why you should donate to the Center for Autosafety. I said that, that’s a little abstract. I don’t know why. You’re gonna donate to the Center for Autosafety so you don’t have to read the Virginia Code on guardrails.

Michael Brooks will do it for you, okay? That’s what your donation’s going for. It’s because somebody needs to read the Virginia guidelines on guardrails. I can’t even remember the exact title of this publication, but it’s fascinating. Don’t

[00:16:40] Michael: tell our donors that.

[00:16:42] Anthony: Tell them it was painful.

[00:16:44] Michael: I tried to read Ulysses by James Joyce once, and I believe I prefer the Virginia guardrail guidelines.

[00:16:50] Anthony: Okay, so listeners, autosafety. org, click on donate. And I guess and continue Michael with his weird book fetish. Okay, [00:17:00] so this article from ABC News we have, it ends with, this is interesting, it ends with this sentence, it says, NHTSA said that from January 2018 to January to August 2023. It found 956 crashes involving Autopilot and Tesla’s full self driving systems, resulting in 29 deaths.

Are these numbers accurate? Because it’s always seemed hard to find the full it always seems that they combine Autopilot and full self driving into one series of numbers. And it never seems The numbers never seem to be quite consistent. We know it’s a lot. We know they’re dangerous systems. But is this right?

[00:17:39] Michael: I don’t know what NITS was based on, that may be what was, I believe that what was contained in the Tesla Autopilot investigation. I believe there’s only been one full self driving death confirmed, maybe two by Tesla, mainly because it hasn’t, been widely adopted. It’s not a great system at all, and it will lead to more [00:18:00] deaths if it’s not improved upon or outright banned by the government, which some of these systems I believe they should be considering at this point.

There have been 28 deaths by Takata airbag in America and Takata airbags are being ripped out of every car they can get their hands on. And I don’t understand why, if autopilot is killing that same number and there are the similar number of deaths and involving less vehicles. that are involved in Dakota, why we’re not looking at stronger actions by the agency.

[00:18:32] Anthony: Because airbags are mechanical things that people think, oh, it’s broken, I can bang on it and fix it. Whereas this stuff, they think it’s magic software.

[00:18:40] Michael: I guess the difference there is that, there’s a the airbags, no matter what you do as a driver or who you are as a human, the airbags are going to deploy and kill you in the Takata situation.

Whereas in the Tesla autopilot situation, you have humans who are actively choosing to let an automation take over. Arguably being [00:19:00] negligent or reckless in their decision making to do and to not pay attention. And so there’s some degree of fault on the part of the human in those situations, which, to me, that is really the only distinction here.

If you accept that, you almost have to accept that there is No fault on the part of those folks in Tesla’s in order to make the, make an equivalent assumption there. And I don’t think that’s possible. So

[00:19:28] Vision Zero and Urban Traffic Safety

[00:19:28] Anthony: let’s move to while we’re still talking Congress, let’s go to the neighborhood of Washington, DC, a lovely place.

I lived there for a couple of years. And I enjoyed my time there, and D. C., like a lot of places, like Hoboken, New Jersey, and New York City, has said, Hey, we’re having Vision Zero. And what this is a decade ago, the incoming mayor of D. C. pledged to eliminate traffic deaths and serious injuries in the district by 2024.

I just checked my watch, and it’s 2024. Unfortunately, the [00:20:00] number of fatalities has risen in all but two years since the initiative was launched. Ouch. So we’ve talked about how Hoboken, which is a lot less dense, a lot less, the population is much smaller, their Vision Zero has actually eliminated all pedestrian deaths, which is amazing.

They did a lot of street design and intersection design, but DC hasn’t managed to do this. What happened here, Michael?

[00:20:26] Michael: Yeah. I think the entire problem comes down to people saying and touting the phrase and the movement of vision zero, but unlike Sweden, which started in 1997 and has halved its number of road deaths.

The United States has a, we have a lot more problems getting infrastructure and road improvement measures through to actually implement the tenants behind Vision Zero. And that’s what happened in D. C. The mayor makes a promise that, 10 years ago that we’re going to eliminate traffic deaths [00:21:00] in the nation’s capital, and 10 years later nothing’s changed all that much if you look at the data.

And that’s because, they have done an, an incomplete and inadequate job of implementing all the necessary things that need to go in place with traffic signals and crosswalks and bike lanes and bus lanes and all the ways that have been discovered to reduce traffic incidents.

Between cars and cars and pedestrians and cars and bicyclists and other road users. It’s really just an inadequate implementation. If, if you can’t put your money where your mouth is and, get these things done in your city, then you probably shouldn’t on the front end have promised that you were going to be able to reduce traffic fatalities to zero in the first place, which, obviously, Is an unreasonable number, it’s vision zero.

It’s been in effect in Sweden for almost 30 years now. And, while they’ve have traffic deaths, I don’t think we’re at an era technologically where we can promise that type of thing and expect it to [00:22:00] happen, within decades within maybe even the next 50 to 100 years. I don’t know if it’s going to be possible to claim that they’re going to be zero fatalities.

Based on traffic or transportation that’s, it’s unrealistic. It’s a good goal. It’s a good ideal. And it’s something that should be, could be the basis for getting more government funding into these areas that can save lives. But ultimately it is, a symbolic thing rather than an actual really readily achievable

[00:22:35] VO: outcome.

Let me

[00:22:36] Fred: digress a little bit here. Readily achievable, yeah, is something. Achievable is something else. The New York Times reports that 127 people, including 61 pedestrians, were killed by drivers during the first six months of this year. New York has has got a problem. But another big city.

In France, [00:23:00] you may have heard of, called Paris, they had the Olympics this year, right? And do you recall seeing all the headlines about how bitterly disappointed people were with the traffic patterns and the pedestrian access in Paris this summer?

[00:23:15] Anthony: Was that because of the breakdancing competition?

[00:23:18] Paris’ Success with Vision Zero

[00:23:18] Fred: Yeah basically none of those headlines existed and it’s interesting because Paris over the last five years has been making a lot of changes towards Vision Zero.

And this includes charging triple to park larger SUVs, closing more than 100 streets to cars, getting rid of about 50, 000 parking spots, and planning to add more than 800 miles of bike lanes. What they have found as a result of all that is that currently 11. 2 percent of trips in Paris are made by bike compared with 5 percent in early 2020.

Just 4. 3 percent of trips are now made by car. The [00:24:00] combination of changes they made to the roadways and the access to public transportation and the fees that they are applying to large vehicles are having a dramatic and beneficial effect. And nobody’s complaining. It Rivoli now have more lanes devoted to bikes than cars.

It’s safe for cyclists, but it’s also a new way of designing a street where the car is no longer king. And finally, it says, looking at before and after photos of Paris, the transformation from car clogged to bike friendly is striking. New York City would have a long way to go for a similar effect. But if Paris can do that much in five years, Making a change from 5% from 11.

2 percent of trips by bike compared with 5 percent in early 2020, dramatic change, no complaints. Everybody’s healthier, air is cleaner there’s no reason why Washington, D. C., which [00:25:00] has this terrible problem, as well as New York City with all of its problems can’t do the same thing. Incremental change here does work, is beneficial, and people accept it.

[00:25:14] Anthony: Yeah, but I don’t like bicycles. They’re not comfortable.

[00:25:18] Fred: So take a subway, Anthony. I do. You’ve heard of that, right?

[00:25:22] Anthony: I do. And it’s got an interesting smell. That’s fascinating that Paris actually made that change. In all seriousness, like that’s a dramatic change. That’s a big city.

It’s fairly dense. That’s, impressive. Hey, that’s the American model, right? If the French can do it, we probably don’t understand it. We’re slightly intimidated and angry. Isn’t that how it goes? I think so. I

[00:25:45] Fred: can now go out and buy a baguette, so I guess we’re on that road. It’s called bread.

[00:25:51] Anthony: Just call it bread.

[00:25:52] Gaslighting in Autonomous Vehicle Studies

[00:25:52] Anthony: Anyway, let’s go into some gas lighting. How’s gas lighting sound for everybody? That sounds great. Hey Brad, do you want to start off? [00:26:00]

[00:26:00] Fred: Oh, sure. I’m happy to do that. So my gaslight nominee is a new publication that is coming out called the rave checklist recommendations for overcoming challenges and retrospective safety studies of automated driving systems.

That’s a mouthful. The published paper preprint is the source material. That I’m referring to here and Anthony, let me just ask you, what is the best way to assure a high quality recording,

[00:26:34] Anthony: High quality audio recording,

[00:26:36] Fred: high quality audio recording,

[00:26:38] Anthony: Low noise. A little noise is possible.

[00:26:42] Fred: There you go.

Now the engineers would say you want to maximize the signal to noise ratio. There you go. And the best way to do that is to have a very reliable signal, right? In other words, the source information that you’re trying to Manipulate has got to be as clear and [00:27:00] reliable as possible. So this study goes in exactly the opposite direction related to autonomous vehicles.

The key assumptions that it makes are that there’s going to be continuing adequate numbers of sufficiently destructive ADS crashes with a wide range of severity to produce a statistical database. And that the retrospective crash analysis can use the same actuarial techniques for AVs as for human drivers.

Now, that’s just wrong because human drivers are thought to be the same and basically unavailable if they’ve been involved in a fatal crash. So the statistical techniques to use for say, numbers of deaths per so many miles driven are developed with the idea that you can’t get to the people and they’ve all got representative and stable capabilities for driving for the most part, unless they’re [00:28:00] inebriated.

But the AAVs are completely different based upon the software configuration load they’ve got, the hardware, the sensors. They’re quite different, and they’re all traceable back to the hardware configuration that you’ve got. Whereas the humans can be considered roughly equivalent for retrospective crash analysis, the machines simply cannot.

And that’s not recognized by this study seems to me the super superficial statistical analysis of inevitable crashes will yield meaningful results for minimizing ADS safety is one of their key assumptions, but they without grounding in detailed forensic analysis that considers all the hardware software, et cetera, associated with the ADS.

And the difference is between them, you just can’t get there. You can’t make the assumption that they’re all equivalent and that they should all be put into the same statistical base. There’s all the problems with it, and I won’t go into all of them, [00:29:00] but basically, the data sources are unacceptable since they’re not objective.

In fact, this prepublication report includes something from the, which is an industry organization that does not include any countering points of view. What I like to call brownwashing and can you guess what brownwashing means

[00:29:25] Anthony: if

[00:29:28] Fred: you refer to that as the same color as bullshit, that would be fine.

But by my definition, brownwashing is a practice where bullshit is introduced without comment.

[00:29:38] Questionable Study Integrity

[00:29:38] Fred: Into seemingly respectable papers, so that it acts as though it has a reputable base. This this study suffers from that. And basically, this industry doesn’t have the same publishing integrity standards that scientists have learned to respect.

It’s a big problem. And what the industry and the [00:30:00] public really needs is a set of requirements that will maximize safety and be testable. So that specific deficiencies can be identified and remedied not to throw everything into the same batch and say let’s just look at it as though these are all the same.

There’s some ironic contradictions and I’ll just toss out 1 of them that it says that the study discussing retrospective safety studies without clarifying whether it has included. Forensic crash investigations can claim that it needs to be centered around the concepts of quality and validity, transparency and interpretation.

But in this particular report, the funding source for the study is not revealed. There are many participants. The bias of the participants is not revealed, not documented. Even though the paper recommends that potential biases of analytic decisions should be presented and discussed. So it’s internally inconsistent [00:31:00] and I think I’m running out of time, but yeah, but anyway, for those reasons that basically it’s internally inconsistent and it makes dangerous assumptions about the data that is trying to accumulate to get ADS retrospective crash statistics.

Thanks. for example, deaths per mile driven. It’s just not acceptable. And so that’s my Gaslight nomination this week.

[00:31:29] Anthony: That’s a good one.

[00:31:30] New York Times Critique

[00:31:30] Anthony: I’m going to, I’m going to take mine now, Michael. I know it’s not GM Cruise. Mine this week is another publication that had two articles just published today that are utter bullshit.

The New York Times. That’s right. The New York Times. New York Times has two articles. One is how self driving cars get help from humans hundreds of miles away. In this article, it’s talking about how different A. V. companies, robo taxi companies, they’re remotely monitoring things and they have people in some data center somewhere.[00:32:00]

Who will intervene and help it get around. And so one of the examples they show is this Zooks car the, there’s a firetruck coming towards it and wants to make a turn in front of it, but the Zooks car just keeps going. And the text they actually use in the article says, It was a reminder that driverless vehicles sometimes struggle to respond to so called edge cases.

Like passing emergency vehicles. A passing emergency That’s not an edge case. Yeah, that’s not an edge case. That’s Tuesday. What the fuck? That’s bullshit. Why are you letting them get away with that shit? That is not an edge case. The other

[00:32:34] Michael: are really angry. I’ve never heard you quite so animated.

It sounds like you and the New York times have a beef going back years.

[00:32:42] Anthony: No, not at all. The New York times is fine. They get a lot of their facts wrong. It’s weird. Like they can’t copy things directly out of a pre anyway. The other article is titled, Can Waymo’s Expanding Driverless Car Service Be a Sustainable Business?

[00:32:55] Fred: Hey, wait a minute. Before you leave, Zeus, let me just say something. I looked into that as well, and I watched the [00:33:00] videos they’ve got in there several times. And I went into California law, and it says that no motor vehicle except an authorized emergency vehicle Or vehicle of a duly authorized member of fire police department shall be operated within the block.

We’re in an emergency situation respond to by any fire department vehicle exists, except that blah, blah, blah more than 300 feet. Did anyone in that video see an officer give the Zooks permission to proceed? Because I did not.

[00:33:28] Anthony: That is extra level dorky. But that is great!

[00:33:33] Fred: And there’s something fishy going on here with what they show in their simulation because number one, the camera views clearly show the small objects like children and pets near the front or rear of the vehicle are invisible.

Nor is anything under the vehicle, such as a pedestrian dragged by crews, visible to the camera. A little bit of a problem there. But if you go to the simulation that they show, where they’re dropping breadcrumbs to show the [00:34:00] vehicle how to get around the hazard. It shows pedestrians clearly imaged on the blind side of the fire truck as the Zoox is passing it.

And there’s no way in the world the Zoox could detect them as it was going past. So how is this cartoon being made? This is really strange that they’re showing all this stuff from the Zoox that’s impossible for Zoox to image. I also wonder if the reporter actually saw this happen because it reports this was 622 p.

m. And I think reporters work nine to five, don’t they? I’m not sure. But it’s interesting. Anyway, I digress, but there you go.

[00:34:39] Anthony: I appreciate your support on my gas light. But the next article I have a problem with the New York Times is called Can Waymo’s Expanding Driverless Car Service Be a Sustainable Business?

And now I’ve talked about this a number of times in this and I’ve done just basic back of the napkin math. No, there’s no way any of these companies will ever break even. [00:35:00] They literally have to take over every city they operate in and put Uber out of business. And then in a decade, they might hit, might be might be revenue neutral.

But this article goes in here and just says things like, while robot taxi services are not profitable right now, Waymo and other autonomous vehicles companies like GM Drag Your Customer Cruise and Amazon’s We Don’t Care About Fire Ambulances and Zooks are vying for a share of a market that could one day be worth as much as 5 trillion, some analysts estimate.

What analysts? Besides Cathie Wood at ARC, what fucking analyst has ever made such bullshit numbers? New York Times, identify this person and then have them do their show them math. I want to see the math where this is going to be 5 trillion, and don’t tell me by 2075 it’ll be 5 trillion. Yeah, great. By 2075 I’ll be 20 feet tall.

This is [00:36:00] horseshit. Ah, and yeah, the article just continues and says, But there’s horrible things in the article too that surprised me, such as California law does not require Waymo to report all of its traffic data to agencies. What? This is bad. So that allows robo taxis malfunctioning in high speed freeways just to be like, Ehh, we ain’t telling you.

But the article ends with a friend of the show, Phil Kopman. Waymo has not had the big crash yet. I don’t know if they will have a big, embarrassing, problematic crash, and based upon the data, they don’t know either. So bravo, Phil! But yeah, New York Times, Why are you gaslighting us, fool?

[00:36:43] Fred: Hey, another supporting comment, Anthony.

Referring back to the previous article in the New York Times, it said that there’s one and a half people in the remote call center for every vehicle that’s out there. Now these people are probably engineers for the most [00:37:00] part. Okay, so engineers will typically earn six figures. Taxi drivers typically do not earn six figures.

So if you’ve got one and a half engineers, Instead of a taxi driver, I don’t think that’s an economic model that’s really very supportable.

[00:37:15] Anthony: That’s why your engineers are going to be located in a data center in Bangalore, India, where even then they’ll be paying more than a taxi driver locally. I would think so.

But yes. Okay, Michael. What do you got?

[00:37:30] Michael: All right.

[00:37:31] Amazon’s Driver Monitoring Issues

[00:37:31] Michael: This week I am nominating Amazon for gaslighting both, both their drivers, and contributing to what could be a problem in the future. So Amazon drivers are being, getting dings against their safety record. Because Amazon is relying on a internal cameras to monitor driving.

And apparently their signal or one of the signs they’re interpreting as driver distraction now granted, they’re trying to prevent driver [00:38:00] distraction here, which is a good thing, but what they’re interpreting are mouth movements as a, as an immediate sign of distraction without any further content.

What’s happening is their drivers are getting dinged for singing while driving. Chewing gum while driving, doing anything that requires the movement of the mouth while driving. And the problem here is that, this isn’t safety. This is over reliance on an internal camera system that can’t distinguish between human behaviors.

And so the drivers are being punished for the system’s inability to distinguish. Moving mouth on its own does not equal distracted driving and why this is a problem is, if you if we’re relying on internal camera systems to monitor drivers, and those systems are inadequate for driver monitoring as this one is.

It’s creates basically a poison pill. That’s going to pollute the well for consumer acceptance of working driver monitoring systems in the future. There’s already plenty of pushback against internal cameras and [00:39:00] vehicles. Driver monitoring may one day rely on those internal cameras. And so if the consumers revolt and the, and driver monitoring systems, aren’t allowed to record video or to use video of people in cars to make determinations as to whether they’re actually distracted.

We’re going to have a problem with the ultimate acceptance, consumer acceptance of these systems. So Amazon basically needs to do a better job of interpreting its camera feeds or eliminate them until they’re willing to pay for working, functional distraction monitoring that doesn’t punish their employees.

And Gaslight Nominee of the Week.

[00:39:40] Anthony: Excellent. And that wraps up Gaslight Nominee of the Week. Listeners, if you want to mention a Gaslight person, go to autosafety. org and click donate. And then, uh, then you can also send a message to us, too. And we’ll read them. We do read them. Anyway, let’s move on to the Tao of Fred.

[00:39:58] The Tao of Fred: AI vs. Human Intelligence

[00:39:58] Anthony: How’s that sound? You’ve

[00:39:59] VO: now entered [00:40:00] the Tao of Fred.

[00:40:02] Fred: One of my local newspapers is the Boston Globe. But before we get into this, let me just ask you, Can you think of a profession that is generally thought of as being pretty smart or maybe a couple of professions thought of? Doctors. There’s a doctor, there you go.

[00:40:17] Anthony: Zinc chloride operator. Zinc,

[00:40:20] Fred: zinc chloride operator?

[00:40:21] Anthony: Yeah, I don’t know what it is, but we’re given a list of all the jobs in the country in 10th grade. And they said one, what about one? And my zinc chloride opera was the bottom choice. So I chose that.

[00:40:32] Fred: Also okay, I was trying to go to the top end rather than the bottom end.

Maybe

[00:40:36] Michael: nuclear physicists working. Nuclear physicists. I’ll go with that

[00:40:39] Fred: one. Okay, so mechanic. I went to Google and I looked up physicists involved in lethal car accidents, and turns out there’s quite a few. And I went and I looked for doctors who’ve been involved in lethal car collisions, and it turns out there’s quite a few of them as well.

[00:41:00] There’s a doctor who wrote an article in the Boston Globe who says why AI will never replace brain surgeons. So I’m looking at, a tool, or he’s looking at a tool that says here’s how you ought to do it. Here are some references. Here’s a lot of ways to go ahead and do this brain surgery and free up this brain.

tumor that’s causing this person a lot of problems. Again, a brain surgeon’s probably a pretty smart person. They tend to be pretty smart. And it actually sounds a lot like driving, because it says, during an average brain surgery, I make several hundred, more like several thousand, small decisions, one after the other.

Each of which resembles a little physical puzzle, like the extrication of a Jenga block. Each move solves a different kinesthetic conundrum, and must be conceived and executed to perfection, lest the whole structure collapse. That’s actually a lot like driving, isn’t it? [00:42:00] But then he goes on to say, here’s why it’s never going to work.

That why artificial intelligence has a replacement of Brain surgeons will never work, but he goes on to say, imagine you’re riding in a self driving car going 65 miles per hour down a winding road overlooking a cliff. Your life now depends on the perfect functioning of a machine that couldn’t care less whether it crashes.

The guidance system behind the wheel won’t be terrified of spinning out of control, nor will its heart begin to race when it comes too close to the edge. It won’t necessarily slow down around the most treacherous corners because It has no skin in the game. While its mechanical precision might not be subject to the vicissitudes of performance anxiety, when it begins to malfunction, you won’t be able to reason with it or appeal to its sense of self preservation because there is no it.

I was actually very eloquent way of talking about the limitations of artificial intelligence in relation to life critical [00:43:00] activities. Like driving, like brain surgery, and in my mind, I think it’s very apropos to one of the fundamental limitations of the self driving vehicles or automated driving systems, or whatever you want to call them, that is not being adequately addressed at all by the developers and probably cannot be.

That’s my thought for this week.

[00:43:24] Anthony: I like that. I think you’re voting for, before we have artificial intelligence, we have to create a consciousness and have that arise so it has some skin in the game, right?

[00:43:35] Fred: Yes, sir. I’m voting for human brain surgeons and also human drivers.

[00:43:42] Anthony: I like human brain surgeons to become taxi cab drivers.

That’d be good.

[00:43:47] Fred: Again, being a brain surgeon is no guarantee of success in the traffic world. No. It’s not sufficiently smart, and so your automated driving system must be smarter than a brain surgeon. That almost [00:44:00] sounds like the name of a game show, doesn’t it? Are you smarter than a brain surgeon?

[00:44:04] Anthony: This brain surgeon wanted to do 65 miles per hour on a winding road overlooking a cliff, which I was like, I don’t think I want to go that fast. It’s that defined winding road.

[00:44:13] Fred: Anyway, that’s available in the Boston Globe last week, called Why AI Will Never Replace Brain Surgeons.

[00:44:20] Anthony: The Boston Globe.

Doesn’t the New York Times on the Boston Globe? I don’t know if I like it then.

[00:44:24] Fred: I don’t know. It could, there’s an awful lot of New York Times articles in there.

[00:44:27] Anthony: Aha.

[00:44:30] Recall Roundup: Toyota, Ford, Continental, and Tesla

[00:44:30] Anthony: Hey, let’s go into some recalls. We didn’t do any recalls last week, if memory serves correctly. This week, we’re We’ve got a bunch of them to get to.

How does that sound?

Rare entrant, Toyota, 43, 395 vehicles, the 2023 to 2024 Toyota Sequoia hybrid. And this is the tow hitch cover and bumper. There may not be enough [00:45:00] retention to keep the cover in place. Thus, the itch cover may separate from the vehicle while driving. Oh, that’s not fun. Cause these are generally made out of steel or Chrome or scrum steel.

Is that a, I don’t know if

[00:45:13] Michael: they are a tow hitch cover can be plastic for sure.

[00:45:16] Anthony: Oh, on a Toyota Sequoia hybrid. Come on. . It’s gonna be Kevlar. No,

[00:45:23] Michael: I’m not sure if it is or not, but it could be. Yeah. Alright. And we’ve seen that they’re doing a lot of recalls recently of things flying off cars.

[00:45:32] Anthony: One of my least favorite things, one of my least favorite things, the mapping software of my choice says object in the road ahead.

And so then you just panic for a while and there’s nothing, you’re like how far does ahead mean? So owners, you’ll be notified and I guess they’ll fix this for you. You’ll be notified by October 19th of this year.

[00:45:51] Michael: Yeah. It looks like they’re going to fix it. Using both a modification to the actual bumper as well as a new type of cover, [00:46:00]

[00:46:00] Anthony: exciting.

Hey, common frequent entrance to the recall roundup Ford. 90, 736 vehicles, the 2021 to 2022 Ford Broncos, the F 150s, Ford Edges, Ford Explorers Lincoln Nautilus, Lincoln Aviator. And this is a problem with loss of motive power.

[00:46:25] Michael: So they have intake valves in the engines that fail, essentially they crack and break over time, which is going to result in it appears to be an immediate loss of motive power, or you’re going to stall out in the middle of the highway, putting yourself at risk

[00:46:40] VO: your engine is going to fall apart very loud and very abruptly, this will be really bad.

[00:46:46] Anthony: That’s not good. So what they’re,

[00:46:48] Michael: what they’re doing is it looks, that the remedy is It looks like they’re going to replace the engine assemblies with for vehicles that [00:47:00] don’t meet a certain engine cycle accumulation test. So you’re going to be inspect, they’re going to inspect your vehicle first when you go to the dealership.

And they, I guess this isn’t happening to all their engines. I guess time will tell on that whether Ford’s right about that part, but for those vehicles that do fail the test, you should have a new engine assembly installed, which is not a cheap recall. And there’s 90, 000 vehicles out there.

So it’s a fairly large recall, depending on how many of these vehicles are actually defective, according to Ford’s test.

[00:47:32] Fred: For those people playing the game at home, you can look up something called a Goodman diagram, and it will. Explain to you how the number of cycles is impacting failure of metal components.

[00:47:45] Anthony: Wait, what do you mean by number of engines? What is it? What is one engine cycle? One

[00:47:50] Fred: revolution. Yeah. So after a certain number of revolutions, this valve apparently fatigues because the heat treatment of the valve stem [00:48:00] was done improperly or was specified improperly. I’m not sure which, but basically it’s too brittle and it’s not ductile enough.

So it’ll break and shatter. And then you’ve got a piece of loose metal inside your piston which is generally speaking really bad that’s what we’re dealing with here. This should be taken very seriously.

[00:48:17] Anthony: If you have a loose piece of metal in your piston, please go see your nearest doctor.

Come on, people. Who’s from out of town? Ah, that

[00:48:27] Michael: was a dad joke. I

[00:48:28] Anthony: know, I know. Look, let’s talk tires. Now, when I think tires, and eventually I’ll have to buy new tires, I think what’s the price I can afford? I don’t know if that’s the best way to do it, but let’s be real. And we have a recall from Continental Tire, the America’s LLC.

146, 000 plus Continental Pro Contact GXA. Oh crap. What a horrible name. It’s like Volvo names their tires. Yeah, there’s mainly, they were an original equipment [00:49:00] fitment for the Audi A4, A5. And approximately 10 percent of the effective tires were sold in the U. S. replacement market. And so Michael, what’s happening with these tires?

[00:49:12] Michael: So they apparently were designed to allow for higher flexing in the tread shoulder, which is I’m assuming the edge of the tread and that causes a heat buildup, which as we know from Ford’s Firestone 20 plus years ago, a lot of heat in a tire. Is a problem and can lead to separations, failure of the sidewall, belt edge separations are what happened in this recall.

If you experience a belt edge separation it may not be life threatening at the time. But if you continue to use the vehicle, then it could become a really big problem. As we’ve discussed at least. On one episode involving tires, they’re the only thing between you and the road maintaining control of the vehicle and are incredibly important and failure of one tire can lead to catastrophic consequences.

It [00:50:00] looks like this recall is going to be out in about a month for Continental tire customers, and they’re going to have a hard time finding you based on your tire purchase. So please check your tires. Make sure you don’t have Continental Pro Contact GXAOs in size 255, 35R19, 96HXL.

[00:50:22] Anthony: When you get together with your friends and family for Thanksgiving in the next couple months, why don’t you know before you go inside the house and have that pumpkin pie, look at everybody’s tires and just be like, ah, Billy Bob over here, he’s got the continental blah blah blah, get those replaced and you’ll do a good thing for people.

[00:50:38] Fred: And for those people like me who can’t remember the string of numbers that was just offered, you can go to our website and you’ll find it listed.

[00:50:46] Anthony: And Fred’s being humble, he has that number tattooed on his forehead right now. Last recall let’s do there’s a little company called Tesla, okay? They have a letter that is titled, [00:51:00] Terminology of the defect slash non compliance determination.

Tesla did a recall, what was it, a year ago? More than a year ago?

[00:51:09] Michael: No, it was 2020.

[00:51:11] Anthony: Oh my god. What year is it now? It was a long time ago. Yeah. And I guess that recall failed? And yeah, so they, yeah, please,

[00:51:21] Michael: they recalled vehicles for this trim. That was part of the vehicle structure. I think it was the roof trim.

They, similar to the cyber truck where they didn’t put the proper adhesives on the Gas pedal. And we saw the gas pedals falling off. This is a roof trim adhesion problem. And they put out a recall for this in 2020. So owners would have already had an inspection and remedy completed at their Tesla dealer.

And now they’re being asked to bring the vehicle back in because the first remedy was not working. That’s a went in and found a lot more failures for vehicles that were Addressed [00:52:00] or supposedly addressed by the 2020 recall so I think there’s about 9 000 vehicles that are going to be getting this remedy

[00:52:09] Anthony: Can we get a license to the black crow song remedy?

Do it to spend the entire budget just on using 30 seconds? No. All right. Hey listeners. Thanks for joining us We’ll be back next week not licensing music from famous rock bands until then Bye bye.

[00:52:27] VO: Thank you. Bye bye. Thanks, everybody. For more information, visit www. autosafety. org.