Partially Automated Failings and Remote Controlled Robo-Taxi’s

[00:00:00] Introduction and Welcome

[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.

Welcome listeners. To another episode of your world’s favorite podcast. That’s right. But Hey, before we start, it’s that giving time of year. Have you gone to the center for auto safety and donated autosafety. org? Click on donate. It will make you live longer, right? I can say that. Good. I did. No one objected.

[00:00:46] Fred: Good morning, Anthony. Good morning, world.

[00:00:48] Hollywood Squares and Celebrity Status

[00:00:48] Fred: Hey, I have a question for you, Anthony. Sure. Since we’re doing this podcast, are we now celebrities? Because I noticed that over the weekend that Hollywood Squares is coming back and I just wonder [00:01:00] if we could qualify for a seat in that. I could use the extra income.

[00:01:05] Anthony: Yeah, you could take up Bruce Valencia’s center square seat. It’s all you. Yeah. Yeah,

[00:01:09] Fred: Yule Gibbons isn’t around anymore either, so there’s an empty seat for him.

[00:01:14] Anthony: Hey listeners, if you can identify an obscure Hollywood Squares celebrity reference, please write in to us, contact at autosafety. org, and we’ll just giggle, that’s it, you don’t get a tote bag or a t shirt or a hat, you’ll get us going, ha.

But that’s not what the show’s about.

[00:01:32] Partial Automation and Driver Assistance Systems

[00:01:32] Anthony: This show is about partial automation. Bum, bum. The Insurance Institute for Highway Safety conducted a new study that found that drivers were more likely to stay alert while using advanced driver assistance systems, ADAS. When they were allowed to correct the steering without disengaging the partial automation.

Now, this is confusing to me because my car has some aid ass. It’s got lane keeping. It’s got [00:02:00] it’s automatic whatever the the adaptive cruise control. So I’ll do all that stuff and I can technically take my hands off the wheels. And it will nag me after a bit, but it’ll just say, Hey, shake the wheel.

And I grabbed the wheel and I shake it and it keeps going. It doesn’t disengage this stuff. So what are they showing here? That is this nudging me towards a safer habit? I thought this was, required that you had to do this.

[00:02:27] Fred: I’m not sure.

[00:02:28] Michael: No requiring risk. Go ahead, Michael.

[00:02:31] Cooperative Steering and Safety Concerns

[00:02:31] Michael: Yeah there’s, there aren’t really any requirements around this type of technology yet, at least from a regulatory perspective or forcing manufacturers to build certain things.

What the study is essentially showing is that, Drivers are safer when they’re kept in get engaged in the task by occasionally having to work with their steering wheel. They call it cooperative steering. As opposed to some of the systems that we see coming out that are trying to [00:03:00] turn your car into your living room, where essentially, you don’t have to do anything.

You can watch a movie and answer your emails and all these things that the manufacturers are advertising Essentially, they say that’s not really going to keep the drivers engaged. They’re going to be less, they’re going to be more reluctant to take over control or put their hands back on the steering wheel when they’re in sticky situations because they’re, in some ways, trusting the partial autonomy to handle those situations for them.

And that’s, and that, as opposed to that system where you’re completely disengaged in the driving task, having. At least some control of the steering cooperating with the automation to steer the vehicle makes people more likely to stay engaged in the driving task and safer. So that’s fundamentally the.

I guess the thrust of this study by the Insurance Institute, and it suggests that, maybe some of [00:04:00] these level three systems these systems like we see with Mercedes that are advertising, basically, the ability for the driver to completely disengage from the driving task play video games, watch a movie, do some work on the computer while the vehicle is operating, may be the wrong way to go, probably are the wrong way to go. I don’t think there’s any question that we think it’s the wrong way to go for technology. It’s a car. It’s not a leisure cruiser that, That can completely control itself safely yet it is we’re not even anywhere near that point and putting these systems out there before we truly understand what driver engagement means and, what type of takeover needs to be available is a bad move on the part of these companies that are pushing that technology out too quickly.

[00:04:52] Fred: I know that in my car, the lane centering system. Allows me to move the car in one direction or [00:05:00] another, and then it will restore it back to the center line. This is absolutely terrifying to me because if I use the system and hands off to try to pass a track to trailer. And I, it brings me in too close to the truck.

So I then steer to the left and it moves away and, I got a decent interval between me and the truck and we go past the truck and then I released my hands from the steering wheel. And the freaking car lurches in front of the truck that’s coming at me at 70 miles an hour because it overcompensates for the fact that I’ve adjusted away from the lane center position that it wants.

Absolutely terrifying. So if terror is the objective of the way to keep me engaged, then it’s absolutely successful.

[00:05:52] Anthony: Welcome to democracy in our current times. If terror is to keep you engaged.

[00:05:59] Michael: And one of the [00:06:00] things that, you know, that the study is looking at are vehicles where there’s a steering automation, but when the driver takes over or executes any maneuvers with the steering wheel, those automations turn off so and then have to be either reengage or turn back on.

That’s not necessarily, that may not be the way to do it. It’s, that’s not really cooperative steering. You’re essentially disengaging one system when the human takes over. So there’s a lot of kind of, there’s a lot of fuzzy ground here that seems like we need more research in this area before we start putting this stuff on the road.

[00:06:43] Anthony: Yeah. From the IHS paper. IIHS tests confirmed that this maneuver, that is grabbing the wheel and taking control, causes both the Tesla and GM systems to stop their lane centering support, while the Ford and Nissan systems continue to provide that support, but allow the driver to dictate [00:07:00] the vehicle’s position in the lane.

Once the Tesla system switches off, it must be reactivated by the driver. The GM system can reactivate automatically, but only if the driver first returns the vehicle to the center of the lane and then stops steering. Even then, there can be a lag before it resumes. This sounds like these guys are still struggling with what Boeing and Airbus dealt with 30, 40 years ago with autopilot systems.

I remember there was a famous crash of a Russian airline flying an Airbus where the guy, whatever, the pilot let his kid sit in the seat and was on the autopilot and the kid just merrrm, just turning the yoke and whatnot. And they didn’t know that, hey, if you turn the yoke all the way hard to the left, It will disengage autopilot and not tell anybody not have a warning and unfortunately that plane crashed.

So it’s, it seems like there needs to be more alerts to the user saying, hey, you’re grabbing this. I don’t know, maybe it could just even be simple of hey, do you want to reactivate this? And just a simple button you can press on the steering wheel. Yes, [00:08:00] reactivate, or no. Because I know Tesla, they’ll hide it under menus upon menus.

I don’t know how the other companies do it.

[00:08:07] Fred: I don’t know, but I think this is consistent with the industry’s approach of just throwing technology out and waiting for the best statistics to determine whether or not they think it’s safe enough to continue. This is, clearly under tested and clearly a lot of questions are reigning about how to implement this technology safely.

[00:08:26] Michael: And when things go wrong, they just blame the driver for not paying attention or not being properly engaged and avoid liability for putting that technology that’s not that may be, working against the human in many ways, putting that technology out there.

[00:08:42] Anthony: This is to the handful of people that I know work at auto companies that listen to us.

Reach out to us. Let us know. We’re not gonna out you. You’re not blowing the whistle. I actually just want to know how do you guys decide these things? We’ll completely keep it confidential. We just want to know, like really, what’s the thought process behind the [00:09:00] human factors here? Or is there none?

Is it happen that you guys hired a bunch of, tech bros from Silicon Valley who have no concept of, Hey, let’s look into the history of how things have worked and let’s do some real testing and check things out. Is that what the, that’s my sus that’s my suspicion.

[00:09:17] Michael: I don’t think there’s any question that the belt selling bells and whistles and shiny objects to people is when it is easy to do these days.

And particularly around vehicles and promising people more than they’re going to be able to get. And then not having to worry quite so much about pushback from the legal system, when you can, when you have the ability to blame the driver for over trusting. The systems that you provided them.

It’s a crazy world. We live in

[00:09:45] Fred: the reality of the social interactions within an engineer organization is that the safety engineers are thought to be just the pain in the ass. And they, and they get in the way of the quote, real [00:10:00] engineers, close quote, the people who are busy developing stuff.

So if you remember going back to the Uber crash that killed what, I can’t remember her name out in Arizona. Yeah. Unfortunately, one NTSCB made. Is that Uber did not have a, an adequate safety culture. And I think that is endemic through the industry. It’s endemic in all engineering industries, I think.

Until there’s a disaster that happens, then suddenly the safety emphasis is brought to a higher level, but the safety community really has a lower status than the engineering development community. So when the marketing guys walk through the engineering loft and see some shiny new item on somebody’s.

design screen. [00:11:00] They just rip it off and run with it and then they try to lay around safety after the fact. So I think that’s what we’re seeing here as well. The marketing guys think this is a great way to sell more cars. They stick it out there. Sooner or later, they figure the safety folks to catch up with it, but that’s not first on their list.

[00:11:19] Anthony: Yeah, I don’t understand the marketing around the stop. It was like, Hey, you can do your work while the car drives yourself. Wasn’t driving supposed to be an escape from work? I don’t know. Could just be me.

[00:11:31] Automatic Emergency Braking Systems

[00:11:31] Anthony: Hey another IIHS piece this week from National Public Radio talking about automatic emergency braking.

And the article talks about this great little test program they have going on there, but I’m going to jump into a middle of the article where it quotes I’m quoting, These systems have really infiltrated the new vehicle market, says Jennifer Morrison, the head of vehicle safety for Mazda North America.

The reason that it’s so impressive is that they’re not required. This is all [00:12:00] voluntary. Which is interesting, because right now AEB is all voluntary, right? AEB standard doesn’t come until 2029?

[00:12:10] Michael: Correct.

[00:12:10] Anthony: Yeah, I pay attention to this show. Yeah. But, and that makes an excellent point here, because they’re not being forced to use AEB.

This is good marketing, right? Cause this is an awesome feature.

[00:12:26] Michael: What they what happened? AAB has been around for a long time in one form or another. And what happened in 2016 was the national highway traffic safety administration, the manufacturers and some other safety groups, I believe.

Consumer reports, IHS.

So last year, all of their vehicles were going to be equipped with a E. B. Now that a E. B. was, You [00:13:00] know, and we’ve discussed this before, a kindergarten, elementary school AAB, it’s the AAB that we’ve talked about that performs really poorly in unlit conditions, doesn’t detect pedestrians, only works at very low speeds, not the speeds where people are actually dying.

So it’s, it was a very simple thing. system and cheap system to put into vehicles that could head off federal regulation. The feds coming in and actually trying to make a be safer, which in the last few years, and it says now put out a new standard that. does make a be actually safer, does make it detect pedestrians, does make it work at higher speeds.

And now and we’ll hear more about this for me in the gas light this week. The industry is coming out against the that’s a proposal, right? They don’t want to be told that they have to put actually working Better a be in their cars. They were totally [00:14:00] happy putting systems in their cars that didn’t work that great and meeting this voluntary agreement to head off real regulations that are going to come into play in 2029.

We weren’t happy in 2016 when this happened. We sued Netsa to try to force them to put out regulations rather than just You know, a voluntary agreement, which essentially goes nowhere. Voluntary agreements are not enforceable. The industry could do whatever they wanted at that point, and they weren’t subject to any real action against them.

They just would get a, rap on the knuckles for not meeting their voluntary agreement. It was, while it got A form of automatic emergency breaking into vehicles and a form that, has been in some cases working to protect people and not nearly as well as what the new standards would do.

But, it has worked somewhat. It’s just a very low level of protection. It, it doesn’t incentivize the industry to get better over time. It, they and [00:15:00] also it was re Yeah. At the time, in 2016, and even now, we’re seeing, automakers sell this technology as an option on vehicles.

A b is no longer an optional component. It generally comes standard on every vehicle at this point. Because the volunteer agreement, but the better automatic emergency braking is not always being put into vehicles. Not everyone is getting pedestrian automatic emergency braking. Certainly not everyone in their vehicles today is getting the higher speed automatic emergency braking or automatic emergency braking that works in low light conditions.

And so there’s still a very long way to go before reaches high school, or, gets to the point where it’s actually protecting the most people it can, whether they’re pedestrians or in cars. And there’s a long way to go and, yes. I don’t know, whether to congratulate the industry for getting it into all their cars when, it only Worked at a very low level, but that is correct.

[00:16:00] It has been voluntary and, it will be voluntary until 2029 when the standards take effect.

[00:16:07] Fred: I want to point out that 1 of the arguments the industry is using to try to delay. The regulation requiring this is that only several of the automatic emergency breaking systems currently out there actually avoid contact with pedestrians.

And the industry is saying because they don’t all avoid contact with pedestrians, then we need to have a lot more time to develop that. Really, if you think of the, from the 2016 perspective, this is allowed industry to delay implementation of technology that some companies have developed that absolutely avoid contact with pedestrians, but the other companies have not bothered to do that.

And now they’re asking for No regulations, because they’re not capable of doing what some of their competitors are [00:17:00] doing. This is, this is really double think in the classic sense. We weren’t able to develop this for because we didn’t bother and now, because some other people did it. The rest of us need to take a lot of time to do that in the meantime, thousands of people every year are dying because the systems have not been implemented in cars.

Voluntary system may have had some benefits, but it has effectively delayed implementation of life saving technology by a decade.

[00:17:32] Anthony: My take on this article, I, you guys have just pointed out, I guess I’m a victim of being gas lit. I fell for it. I’m a gaslight victim. Oh, God, no. And with that, let’s just go It’s okay.

[00:17:46] Michael: I think we all are in some way every day in the modern world.

[00:17:51] Gaslight Illumination: Industry Critique

[00:17:51] Anthony: Let’s just go right into our gaslight illumination. And, Michael, just Good. Continue on why I fell for it.

[00:17:58] Michael: Straight from this article, [00:18:00] you have the president of the Alliance for Automotive Innovation saying, that the NHTSA standard that’s coming in 29 is virtually impossible to meet at this point.

And to that I say you’ve got five years to meet it, but also there are already vehicles that meet it. There’s a Toyota Corolla or, a, not a big, fancy luxury vehicle, but, and, Pretty reasonably priced Toyota that has been able to meet this test. So we’re not talking about some super expensive tech.

We’re not talking about something that is virtually impossible as They’re saying it can be done. It has been done. It is being done. They just don’t want to put the money into vehicles to get the systems in place that can do it. They’d rather sit on their laurels make a larger profit on all the vehicles they’re putting out in the next 10 years, and continue to watch high speed crashes where automatic emergency braking could save lives happen.

We know [00:19:00] that, we know the technologies out there that can do this. It’s pretty obvious. And also, I just want to point out, in the meantime, this is the same group that is continuing to push autonomous vehicle safety and how autonomous vehicles are somehow safer than humans. But somehow, it’s impossible for passenger cars.

To work properly and braking at highway speeds of 45 miles per hour and above, I think Waymo would certainly disagree with you on that. And, virtually all of the autonomous vehicle manufacturers, and especially the ones that are making heavy trucks have to find a way to make this work.

And they already have. The technology is there. It’s just a matter of getting into all vehicles. This is at base simply a push against Yeah. It’s a push for more profits and less costs in their vehicles at the expense of safety. And that’s my gas light of the week.

[00:19:56] Anthony: Listener, do you enjoy the passion of Michael Brooks?

If so, go to autosafety. org [00:20:00] and click on donate because I was really into that. That was good. I’m gonna do my gaslight. My gaslight is the entire automotive industry for trying to sell me on level 2 plus level 3 driving systems that will take care of themselves and can disengage and allow me to do my work while I’m in a car.

Cause that’s just nonsense. Is that gaslighting? Sure. Why not? It’s a quick one. It’s short. Bread’s looking at me skeptically thinking, Jeez, you didn’t do your homework. You’re just winging it, aren’t ya? I can see your face. I know what’s going on here. I’m winging it, okay? But, I Michael did the second story, I’m doing the first story.

All right, Fred,

[00:20:35] Fred: go for it. What’s yours? Mine is a statement by Takedra Mawakana, and forgive me if my pronunciation’s off, who is the co CEO of a company called Waymo. And she wrote a holiday message, and she said Quoting here, this holiday season I’m filled with gratitude for our Waymo One riders. You share our vision for a future where transportation is [00:21:00] safer, more accessible, and more convenient for everyone.

Thank you for being part of this journey and for helping us build the most trusted driver. Wishing everyone a very safe and restful holiday season. Wow, that sounds pretty good. So let’s take it apart for so she says for our future where transportation is safer. She seems to acknowledge that there is a safety threshold that has yet to be crossed and is something that they would like to do.

So the inquiring minds wonder just what is their current safety assessment and what is their goal? How do we know where they are and what they’re going after? She also says, thank you for being part of our journey and for helping us build the most trusted driver. Note the distinction between the word safest and the phrase most trusted.

They’re really pushing for the public perception rather than the reality of safety. This is not the way [00:22:00] safety should be done, but again, safety has a low priority. in many engineering organizations. And finally or almost finally it’s no wonder she’s filled with gratitude because normally companies have to pay for test engineers and test people to test their technology.

Waymo has completely inverted this. And they now have people paying for the privilege of exposing themselves to unknown lethal hazards and unquantified risk. Waymo has managed to get untrained and unwitting test subjects to pay for the privilege of exposing themselves to dangers of this experimental transportation system.

With known uncontrolled collisions, wayward operations, exposure of passengers to threats to their personal safety by people standing in front of the cars and threatening them, and according to published reports, also to the illegal and unwise transportation across cities of children. [00:23:00] Who are not capable of taking care of themselves, right?

Because they’re children. And I also want to point out they’re not really AVs. There’s an extensive back office of remote driver assistance and service personnel that they call out whenever they, oh, run into the back of a truck or cause a traffic jam or have hallucinations and run into light poles.

These, they’re really not AVs. They’re Too many assisted vehicles that are out there doing something, so I’m going to offer the guest. I nominate to Kedra for admitting the deficiencies and Waymo’s current safety profile and Waymo’s use of unwitting test subjects who are actually paying Waymo for the privilege of helping to establish a public perception of safety at odds with the objective evidence.

In the nicest possible way and with a winning smile I will also have a reference to a, an article [00:24:00] that describes the fact that it says here, human driven vehicle accidents let’s see, no, however, accidents involving advanced driving systems occur more frequently than human driven vehicle accidents under dawn, dusk or turning conditions.

Which is 5. 25 and 1. 98 times higher, respectively, than for the human cars, human driven cars, so a long way to go for Waymo in these circumstances. We’ve seen other papers with even more dramatic comparisons, but. We’ll leave that for another time.

[00:24:35] Michael: So happy holidays to all of our test subjects.

[00:24:39] Anthony: All right.

I think it’s a tie this weekend in the gaslight illumination. Michael wins on passion and Fred you ran on clearly you wrote a paper on this, like you did research and really you put the extra effort in there.

[00:24:51] Fred: I did some homework, Anthony, unlike my high school experience.

[00:24:54] Anthony: Hey listeners, have you donated? Have you gone to autosafety. org and clicked on donate? Because once you’ve done that and [00:25:00] you’ve completed that, I want you to go on LinkedIn and follow Fred Perkins. Because it is the most entertaining thing in the world. It is the fly in the ointment. It is great because all of these Waymo people and all these AV people, they put out, basically the thing that he just read where there’s these very light on fluffy details that are great.

And then you see in the comments, Fred being like, Yeah, define safety. And every other comment is yeah, you’re great, ha, hearts, stars, yay, we all get a sticker. And Fred’s yeah, none of this is safe. Can you define what safety means? They never reply to them. It’s a shocker.

[00:25:36] Fred: It is interesting.

They do never reply. It makes you feel isolated. That’s why I want to get on Hollywood Squares. Oh, okay. I want to be popular. Do

[00:25:45] Anthony: you? Yeah. Do you? Speaking of somebody who’s super popular, let’s talk Tesla.

[00:25:52] Tesla’s RoboTaxi and Remote Control Concerns

[00:25:52] Anthony: Oh boy. From an article in TechCrunch Tesla’s gearing up to launch their RoboTaxi service.

Ha! Lie. And but at [00:26:00] the same time what they’re doing is they’re hiring a bunch of humans to sit in the back office and remotely control these RoboTaxis. That’s right, your fully self driving car is controlled by Clyde in an office park somewhere outside of Toledo, or in Buffalo, or wherever.

Tesla they did this little Hollywood Squares type thing showing, hey, here’s our RoboTaxi. And we have, look at our robots walking around and whatnot, but the robots were all remote controlled. There’s guys sitting in a back room, they’ve got joysticks moving these things around.

They’re just rich kids toys from the 80s. And so essentially that’s what the suspicion is that’s what his robo taxis are. Because Tesla’s hiring software engineers to help develop their tele operations systems that allow human operators to remotely access and control Oh, boy, do not look behind the curtain.

[00:26:49] Michael: So I think behind the curtain, you’re going to that’s that scares me to even think about allowing that, riding in a vehicle where you are. The [00:27:00] vehicle is being controlled by someone in a monitor, behind a monitor, a thousand miles away, just gives me goosebumps.

And, we’ve talked about latency and other things here, but I, even with the most advanced system available to to humankind today, the fact that, we all struggle sometimes to get a reliable web connection or to have a computer that works properly. glitching on me this week, for instance, getting into a vehicle, traveling down a highway controlled remotely by someone where all of those potential glitches and gremlins in the matrix exist scares the hell out of me.

Maybe you guys are a little more brave than I am though.

[00:27:41] Anthony: Hell no. Related to this, there was an article that just came out about Zipcar and how they had an outage and so people couldn’t use the app to get inside their cars anymore. And this is just a parked vehicle. And there’s just, an outage, which is an extreme example of latency.[00:28:00]

And people are locked into their car. Now imagine this with your robo taxi, where there’s some outage, because, outages happen all the time. And now you’re traveling down the highway at, 60 miles per hour, and Clyde in Cleveland’s I, somebody unplug the Ethernet. I don’t know, reboot.

Hold on, let me reboot the servers over here. And how many miles have passed while the car is quote unquote driving itself? Tesla. Ah. And

[00:28:31] Michael: I also wonder, one of the things, one of the weirder things, you keep seeing all these references to Tesla employees having to work 80 plus hours per week and burning the candle at both ends.

Do you really want that person being a remote operator of your vehicle? Someone who’s sleep deprived and, who certainly wouldn’t be able to drive a semi truck at that point, they’d be way past their hours for the week. Is there, there just doesn’t seem to be any consideration of safety here, on top of [00:29:00] everything seems to be.

The glitz and the shiny object. It is a robo taxi for Tesla when they haven’t shown the ability to even put together the sensors or the computing system or anything that could take on that task all their vehicles right now are require a human in the driver’s seat to control the vehicle. And, to make sure that they don’t die.

So it’s, there’s all sorts of complications here. I guess the bottom line here is that, they’re promising to have robo taxis on the road sometime next year in California and Texas, and if anyone out there is betting on that, I would encourage you to withdraw that bet right now.

[00:29:40] Anthony: Look, it’s only for the super hardcore, okay?

You gotta be super hardcore with maximum velocity, and we’re all gonna be sleeping on the floor. Okay? That’s how we do things.

[00:29:49] Tesla Cybertruck Crash Investigation

[00:29:49] Anthony: Next Tesla story, and our final Tesla story, I believe, of this week. The National Highway Traffic Safety Administration said Friday, it is looking into the Tesla Cybertruck crash in Piedmont, California [00:30:00] that killed three people and injured a fourth when the vehicle slammed into a tree and a retaining wall and burst into flames.

This is an article we’re linking to from MSN and you can see footage of the Cybertruck at the end of this. And wow, it’s it’s horrific. The Cybertruck, it looks like somebody beat up a tin can. It normally when you see car crashes, like you, there’s crumple zones and things break away from this perspective.

They have the truck flipped upside down the panels, the way they’re folding in doesn’t. Look like a normal car crash.

[00:30:34] Fred: It was horrendous. It was awful.

[00:30:36] Anthony: That’s all right. But even just the way that the car itself is dealing with the impacts, it doesn’t not look like what I’ve seen other cars in that situation, the way they crumple and collapse.

The

[00:30:46] Fred: car was forced off the road by a truck, right? And so it was hit laterally first, and then it went off the road and rolled several times. So it was a really unusual [00:31:00] accident. I feel horrible for the people who were in that. What a disaster.

[00:31:05] Anthony: Horrible. Yeah, and so they’re still investigating this and figuring out what exactly happened if this was a, what happened here.

We’ll keep you updated on that as we As more news on that breaks. But for now, Fred, are you ready for, For your towel. Oh, sure. How’s that? I did my homework. Good.

[00:31:26] Autonomous Vehicle Sensor Challenges

[00:31:26] Anthony: This is a autonomous vehicle safety pre fixer make sensor hardware swaps, less dangerous. I have no idea what that is. Take it away.

[00:31:35] Fred: All right.

You’ve all heard about artificial intelligence, right? Which is misnamed and it’s really machine driven correlation. So think of the, think of photographs, right? Where you’ve got a photograph. That has 10 million pixels, and then you have another photograph that is 2 million pixels, and the photograph is of the same subject.

But [00:32:00] how does the computer figure that out? Because the computer is going to sort through each of the individual pixels and determine which of those belong to an image. And then which of those two images have to be compared with each other? We do it with eyes, because we recognize instantly what the object is that generated the the image, right?

It’s your aunt Nellie or whatever. Computers, it’s very hard to do that, to perceive that. One of the hazards that they talk about here is that if AVs actually are put on the road in any kind of numbers, eventually, there’s going to be a need to replace some of the sensors. And so think of what would happen if you’ve got a camera that’s being used to control the vehicle, and you replace a 2 megapixel camera with a 10 megapixel camera.

All of a sudden, you’ve got numerical data sitting in a memory somewhere that’s got to be compared with [00:33:00] very different numerical data that was generated previously. And how do you do that? How do you make sure that the computer recognizes that these 10 million pixels associated with a fire hydrant are somehow the same as the 1 million pixels that are associated with a fire hydrant from a previous database?

Very difficult numerical problem. So what they’re proposing in this article is that they put in basically a digital device, basically, and another artificial intelligence engine that sits between the camera or other sensor. And the data that is stored so that the artificial intelligence preprocessor will put the data into a format and a form that allows it to be compared with other data that [00:34:00] was developed using a separate sensor with perhaps several different sensor characteristics.

Because the objective for the AI is to compare one image to another, or one sensor field to another, and in order to be consistent, it’s got to have consistent data. But if the sensor isn’t sending consistent data, there’s got to be something in between that allows the computers and the AI engines to develop this consistent history, right?

Because it’s a correlation engine, so you’ve got to have things to correlate with, and the things that you correlate with have got to be similar, To other things you’re correlating with or the butter slips off the noodle, as they say in the South. Now, this is probably a good idea as part of the system, but what it really does is it adds yet another artificial intelligence engine to this stream of processing that would have to occur.

You would have to make sure that it’s consistent, not only with the current sensors. [00:35:00] But the future sensors that might be added to it. So that’s a really tough thing to do from a development standpoint. Anticipating what technology might be bringing down the road in the future. And now you’ve got 3 artificial intelligence extensions, right?

You’ve got 1 that sits in the that sits between the sensor. And the data system, and then you have another artificial intelligence engine in the data system that does the perception associated with all of these images, right? So once you’ve got the images consistent, you’ve got to find a way to identify Antonelli from all the noise that’s in there.

And then you’ve got yet a third artificial intelligence engine system that takes the perceptions, or the perceived images, Image or the perceived data from this sensor system, whether it’s radar or camera or ultrasound or whatever, and then puts that into the driving logic. So that the vehicle knows which way to steer.

And whether to accelerate or [00:36:00] decelerate, so you’re not going to string of 3 artificial intelligence engines, each of which has to be very capable. Before the system can generate a command. For the vehicle to turn left or right or not run over the human who happens to be in front of the car. Yeah, this is probably a good idea in some sense, but it’s, I think it’s really going to have to wait for quantum computers to be put into cars because there’s just a hell of a lot of data processing that’s going to go on.

Anyway, that’s what this is all about. It’s a great article because it identifies yet another defect. And the logic associated with the assumption that AVs are automatically safer and provide a more benign driving environment than human driven vehicles.

[00:36:47] Michael: And that sounds like a really bad retrofit to me.

Why not just stick with the systems you’ve already developed the software, for? It may

[00:36:56] Fred: not be available, Michael. Company’s going out of business, right? [00:37:00]

[00:37:00] Anthony: Things, my

[00:37:00] Fred: things happen.

[00:37:02] Anthony: My assumption is, let’s say you buy a company from a car, from a company called, let’s say Tesla.

And this is a company that doesn’t have a well established parts pipeline like Ford and GM and Toyota, which they produce cars and they realize, Hey, we’ve gotta make parts for this stuff for the next decade minimum. Tesla doesn’t do that. So you get into an accident with your Tesla and stop, Fred, stop me if I get this wrong.

And let’s say my right. Side passenger mirror camera brakes, and it’s a currently a two megapixel camera and they’ve got to replace and say hey We don’t make the two megapixel anymore. You got to buy the five megapixel So they’ve got to put that five mega So this even gets more complicated because now do all the cameras in my car have to be replaced so they’re all at the same Standard or can I just replace that one?

[00:37:52] Fred: Good question, but I, I think as a practical matter, they would have to install this artificial intelligence engine in the line [00:38:00] replaceable unit, right? The camera essentially that you’re going to or the mirror that you’re going to stick on the side of the car. So that’s it. But another example would be mobilized, right?

It’s an Israeli company that produces sensing technology. That’s being promoted for use in cars worldwide. What if Mobileye goes out of business or there’s a lot of, a lot of military activity going on in Israel too. So who the hell knows what the future is going to hold for that sensor suite.

And if you’ve designed your car around that particular company’s offering. You could be in a world of hurt if that company loses the ability to provide that sensor for whatever reason. Yeah it’s complex, and this is a step in the direction of saying if it’s going to be practical, we’ve got to assume there’s going to be a wide supplier base.

There’s going to be damages, there’s going to be replacements, right? It’s going to go to Joe’s Auto Body to have the sensors replaced, and Joe’s Auto Body is not a radar developer. [00:39:00] They’re just going to buy a box and stick it on there. So what do you do in that case?

[00:39:06] Anthony: Oh, no. You regret buying a car that doesn’t have a good third parties parts supply line.

[00:39:12] Fred: Yeah. I just had a little experience with that. When the deer jumped in front of my car and I had to get my eyesight system replaced and calibrated, it was a fricking nightmare. The software. Hadn’t been updated for the replacement sensor. It took months for the guys to figure out what the hell was going on.

And then it took them weeks to put the software in. And then it had to be calibrated using the finally upgraded information. The dealer didn’t have it, the folks who were doing that had to do a lot of independent research, which they were not really trained to do. And this is just the camera system, right?

So think if you’ve got a radar, you got a LIDAR on there and, Joe’s LIDAR company goes out of business and you got to go to [00:40:00] Susie’s LIDAR company. All of a sudden you’ve got a critical system that’s going to be driving people around, it’s safety critical. And your computer doesn’t know what the hell is coming out of the LIDAR system because it was designed for Joe’s, not Susie’s.

[00:40:14] Anthony: Related to what you were talking about, Fred, Michael, as our representative of the Southern United States, does the butter slip off the noodle?

[00:40:22] Michael: I’ve never heard that phrase before, but bless your heart for asking.

[00:40:29] Anthony: Ah, go to autosafety. org, click on donate. Just put a note in there that says none of it goes to Michael.

There we go, great. And I don’t know if the butter slips off the noodle, folks. I don’t know.

[00:40:39] Luxury Car Recall: Pagani Utopia

[00:40:39] Anthony: But I do know that it’s time for some recalls. How’s that sound? Alright first one up. I love this one. The Paganini! The 2024 Is that

[00:40:49] Michael: what it is? The Paganini? No,

[00:40:50] Anthony: it’s not Paganini. Paganini was a composer. Pa Pagani?

[00:40:55] Michael: Pagani. I’m

[00:40:57] Anthony: assuming. I’m gonna say Pagani. The 2024 [00:41:00] Pagani Utopia. That’s right. It is a vehicle type. This is my favorite vehicle type. Low volume vehicles. Potential number of affected cars. If you’re playing the home game, take a guess right now. dun. The number is eight. That’s right. Eight cars. The defect on the carbon fiber passenger side dashboard panel could potentially damage the bag during deployment and affect the performance of the occupant protection during a front crash episode, increasing the risk of injury.

Now, I have never heard of a Pagina, Pag, Pagini Utopia. Has anyone heard of this car?

[00:41:37] Michael: I had never heard of it, but I looked it up because I wanted to see what it looked like. It looks a retro sports car. They, they don’t look like they look like on the older Ferrari, Lamborghini, one of those, one of those type of very classic look to them.

But the most intriguing part of this whole thing is they’re three and a half million dollars per car. [00:42:00] So they are insanely expensive. You could buy, 100 cars for that amount. That would work just as well, just not as fast and probably have working airbags. But there’s essentially, this is a recall that we’ve seen before with other manufacturers, which is they design a and build a dashboard that the airbag isn’t.

Capable of deploying properly for it through. So for the eight owners of these vehicles, you don’t have to do anything. They’re going to come pick it up from your house. I think they have five dealerships in America and they’re going to tow it away and bring it back to you once it’s fixed. They even have a team of flying doctors on call from Italy to help you out.

[00:42:43] Anthony: This is amazing. I’m looking at the car now and it looks like a McLaren was like, attacked by like a Lamborghini Countach and run through a steampunk filter or something. It is it’s bizarre and it looks like it’s part transformer. I’m looking at a photo here where [00:43:00] like the trunk lid pops forward and the back part pops backwards and the doors are those gull, not gulling, that slide up thing.

Wow. But wait, five dealerships and they’ve sold eight cars. My God, the butter has slipped off the noodle.

[00:43:16] Michael: For, 25 million or 28 million is, would be the total cost of those eight. So they’re probably making a lot of money per car, right?

[00:43:26] Anthony: Wow. Huh? Yeah. The remedy for all the vehicles is the application of Kevlar strips on all the edges of the panel in order to avoid any contact with the carbon fiber with the airbag.

That’s what I want in my car. I want Kevlar. I want carbon fiber. Bruh, it’s gonna be, I’m so fast I don’t even get hit by bullets. But, it just, I can’t drive it on roads cause it will bottom out. It’s bizarre.

[00:43:52] General Motors Tailgate Issue

[00:43:52] Anthony: Next up, we have company General Motors. Hey, last time around we had 8 recall, [00:44:00] 8 infected vehicles.

How many do you think General Motors has? That’s right, 132, 037 vehicles. The 2024 Silverado HD, Sierra HD oh, just those guys. General Motors has decided that a defect which relates to the motor vehicle safety exists, blah, blah, blah, power on latching tailgates. Water may intrude into these vehicles tailgates and come into contact.

And cause a short circuit. Oh, that’s a lame one.

[00:44:30] Michael: What’s happening is that there’s an electronic tailgate release inside the vehicle, I’m assuming you’ve got a button you can hit to unlatch your tailgate. What’s happening is that these tailgates are getting the electronic gate release switch is getting water into it.

And while your vehicle’s parked and you’re unaware of all this, the tailgate latches. Released. And so when you get back in the car, load up the load up your bed of your pickup with cargo and [00:45:00] take off down the road. The tailgate is actually unlatched can fall and everything in the back of your truck can spill out on the road and kill people behind you.

So that’s essentially what’s going on here. And it’s, yet another yet another issue with water intrusion into rear latches. I think we’ve seen this in jeeps. I know we’ve seen it in door handles as well. Some of the V. W. I. D. force, but what’s going on? You’re putting electrical components into parts of the vehicle that aren’t fully sealed from the environment.

That doesn’t seem very smart.

[00:45:36] Fred: I have a friend who grew up in Nebraska, and he said one of the recreations in high school was for a bunch of his friends and himself to get drunk, get in the back of a pickup truck with their rifles, their 22s, and then go cruising around at night in the open fields looking for jackrabbits they could shoot at.

Out of the car and I [00:46:00] thought this, there was probably no more hazardous approach to pick up trucks that they could possibly do than running around shooting at things in the dark while drunk. At high speed in an open field, but now I see that it could be even worse because now the tailgate could open.

They could all go flying off the back with their guns in the air at night while drunk in Nebraska.

[00:46:24] Michael: And I just want to I want to point out to people that own these vehicles, these, the switches with the wires that were a problem here were fixed in on March 8th of this year by GM. So they started sealing these wires with butyl tape.

I’m not exactly sure what that is, but I’m assuming it’s something that prevents water intrusion around the wires. Why does it take another, Eight months before the vehicles that didn’t have that fix were recalled. And the owners aren’t going to be able to get it, get their notification until January of next year, 10, 10 months [00:47:00] between the time GM obviously knew this was a problem and started fixing their factories to when the people that had already paid good money for these vehicles.

Till the time where they get their fix. That just doesn’t seem doesn’t seem very responsible. And I hope that no one has been injured or impacted by this problem in the meantime.

[00:47:19] Anthony: We’re calling you out, GM. And Fred, I think you should do an advertisement for either the state of Nebraska or the United States in general.

Come visit us. Get in the back of a pickup truck drunk with a gun. Start shooting at rabbits. Yeah.

[00:47:34] Hyundai Rearview Camera Recall

[00:47:34] Anthony: Next recall. Hyundai Motor Corporation, 226, 118 potential vehicles, the 2021 2022 Hyundai Santa Fe, Hyundai Elantras with a whole bunch of different letters after them, the Hyundai Santa Fe’s with letters after them, the rear view camera, oh my god, really?

Rear view camera basically insufficient solder.

[00:47:55] Michael: Yeah, that’s the rear view camera crisis really hasn’t stopped. We’re continuing [00:48:00] to see just tons of these recalls coming in and across all manufacturers, really. It’s almost like in 2018, when they were required to start putting these systems in that they did a half ass job of designing them and getting proper systems into their cars.

Or maybe they cheaped out on a lot of the components involved, but it’s, this is coming up over and over across all manufacturers. They got to get this rear view camera stuff together. I’m not sure why it’s developed into such a problem. We see many cases where they’ve been routed, they’re in the same system as the entertainment system, the radio, the touch screen here, it just looks you’ve got problems with the camera.

So this is a little different than what we’re normally seeing here, but once again problems with rear visibility, something needs to be May need to be done here to address, just the sheer volume of rear view camera issues. We’re saying,

[00:48:52] Anthony: I think the solution, this will handle two things.

One, we eliminate rear view cameras to, we eliminate water because these [00:49:00] two, this seems to be a problem. We do that. If we eliminate

[00:49:02] Michael: water, we don’t need rear view cameras anymore, right?

[00:49:05] Anthony: Exactly. There you go. Two birds, one brain

[00:49:07] Michael: or anything. We don’t need it. We don’t need anything anymore. Owners are going to hear about this.

one in mid January. It looks like you’re just gonna get a replacement of the rear view camera in the vehicle. So it may not be too complicated of a repair. You might not have to ask for a loaner vehicle while they repair, but be sure to if they’re gonna take too long.

[00:49:31] Mercedes-Benz Sunroof Recall

[00:49:31] Anthony: Next up Mercedes-Benz 33,456.

Vehicle. The 2003 to 2007. Wow. These are old Mercedes-Benz C two 30. The 2002 to 2006 C two 30 coop. This is the first one I mentioned was a sedan. Oh. There’s a whole bunch of these c if you’ve got a Mercedes-Benz, and it’s old and it starts with a c. Or an AMG, see, that’s also got a C, oh my, there’s a whole long, it’s pages and pages and pages, [00:50:00] I’m still scrolling, oh, it’s, it seems like it, oh my god what’s going on here adhesion bond what the glass panel what’s happening, this is the sunroof and the sliding roof frame may not meet specifications, so you’re gonna lose your moonroof.

Is that

[00:50:17] Michael: what? Yeah it’s, these are a lot of these vehicles are 20 years old. They’re older vehicles and NITSA has been looking at the problem with these sunroofs for quite some time. Basically the sunroof is. It’s flying off the vehicle to put it in an uncomplicated way and causing really a a safety issue for people following in cars behind them, for pedestrians, motorcyclists, that kind of thing.

But this is a recall that’s been coming very long time. And I’m sure for those vehicles, those Mercedes that are left out there, it’ll be good news for their owners. But it’s, it probably took a lot longer than it should have.

[00:50:56] Anthony: Yeah. So this adhesion process problem I’m here [00:51:00] I’ve been told is also called when the butter slips off the noodle.

All right.

[00:51:04] Porsche Headlight Software Error

[00:51:04] Anthony: Last recall Porsche, 2, 941 vehicles, the 2024 to 2025, my can electric the upper beam photometry requirements exceed maximum threshold. Oh, what’s going on here? Software error. They’re programmed to the ECE instead of FM BSS specifications. So they’re programmed to European standards instead of US standards.

[00:51:30] Michael: Yeah, it looks like they are and they, the, it looks like there’s a problem. I think the European standards apparently, allow for brighter lights in certain circumstances. And the safety issue here is that the Porsche lights are, they’re non compliant with F and D. S. 1 0, 8, which sets all the photometric requirements for, vehicle headlights and ensures that they’re not too bright. We’ve discussed that before. There’s still a lot of situations, which other cars lights will [00:52:00] be too bright. But in this case Porsche is going to have to go back and put a correct software data set into the cars. Basically the software data set that complies with the United States lighting regulations versus Europe’s.

[00:52:14] Anthony: All right.

[00:52:15] Listener Mail: Seatbelt Struggles

[00:52:15] Anthony: And now we have listener mail. And this addresses a comment I made last week where I was like, Who’s not wearing their seatbelt? And Teresa W. wrote in and said, I listened to your disbelief that people don’t wear seatbelts. Clearly you aren’t a short woman with boobs. Good observation.

Seatbelts, even on the short setting, almost immediately ride up and slice into your neck. Putting a padded cover on the belt just makes it more irritating against your neck. The only most comfortable seatbelt I’ve worn was in a convertible because it was mounted below the top of my shoulder. I try to wear seatbelts, but it’s a constant struggle.

That’s that’s awful. I feel I, I haven’t run into this problem, but I can imagine how awful this is for you. Is there [00:53:00] a solution, gentlemen?

[00:53:02] Fred: One solution is to buy a convertible, but I don’t know this in detail, but it did go to Amazon and they have a lot of adaptive devices in there that are intended for short people or shorter people to make the CPLs more approachable.

I hope that the I hope that the listener Is able to take advantage of some of those and if they in fact do the job don’t really know, but yeah it’s true. I’m neither short nor gifted with boobs, so I don’t really have a lot of intimacy with this problem.

[00:53:37] Michael: Hopefully, yeah, go ahead. It’s, this is just a problem that’s inherent to vehicles in general.

You’ll see some cars that have the ability to at least adjust the seatbelt anchorage height. Usually it’s on your, I think it’s on the B pillar to your right behind. To your left if you’re the driver and you can adjust the height there somewhat, but [00:54:00] a lot of vehicles don’t have that at all.

If you’re a rear seat, you probably aren’t going to be able to adjust your seatbelt. And if it’s coming right across your neck, or, in other sensitive areas, people are going to be resistant to using them. And so we need to make seatbelts. As comfortable as possible so that people actually use them and providing people with the ability to adjust them in ways that are safe makes a lot of sense.

It sounds right now, if I was in the same situation as the commenter, I would, you could probably be forced to shop around lots of different cars to find one where the seatbelt was actually equipped with an adjuster that could fit my body type. And that’s it’s unfortunate that there aren’t more options there as far as, there, there are a lot of people who have trouble with seatbelts.

There’s people who are too large, too small, too tall, too short, and there needs, there need to be more options. And the ability for people who are not the average size to [00:55:00] remain protected and comfortable in their car is important.

[00:55:03] Fred: So we hope the listener finds a way to accommodate that and continues to use the seatbelt because it is very important for your safety to be strapped into the vehicle.

[00:55:14] Anthony: Yeah, I’m curious, Teresa, if you can write in. I’m curious if you’ve ever tried any of these devices. But I also imagine not just for her own car, but like you go into a friend’s car and now you have that same problem because You can’t really adjust their stuff so much. Yeah, you can’t imagine this woman’s got to carry around some sort of seatbelt, adaptive device with her, wherever she goes.

That seems unreasonable. Maybe for your own vehicle, maybe there’s a solution. I don’t know. Write in, let us know. Very curious.

[00:55:41] Fred: Yeah, and what kind of car do you drive now? Maybe other vehicles would have a better, more attractive profile for you.

[00:55:52] Anthony: Who knows? But with that, everybody else has to wear a seatbelt, okay?

Theresa you get a [00:56:00] pass, maybe. Okay, everybody else put your gun down.

[00:56:04] Conclusion and Call to Action

[00:56:04] Anthony: And with that’s the end of this week’s show. Thanks for listening. Please and subscribe and most importantly, donate.

[00:56:11] Fred: for listening and don’t let the butter slip off the noodle.

For more information, visit www. autosafety. org.