Live data from Hacker News

Tesla sued in wrongful death lawsuit that alleges Autopilot caused crash

techcrunch.com

321–330 of 495 posts

Re: Tesla sued in wrongful death lawsuit that alleges Autopilot caused crash

#321
post #300

Tesla's blog post: https://www.tesla.com/en_GB/blog/update-last-week’s-accident the driver’s hands were not detected on the wheel for six seconds prior to the collision. The driver had about five seconds and 150 meters of unobstructed view of the concrete divider Looks like sensors failed to see concrete divider in nice sunny weather and car slammed in to it at 70mph. Driver was obviously over confident on system's a…

> ignored warnings to put his hands on steering If there were any. One of the claims in the lawsuit is "failure to warn" and the preliminary report of the NTSP says "the alerts were made more than 15 minutes before the crash".

The idea that the driver didn't know he was supposed to have his hands on the wheel while using autopilot seems ridiculous. Anyone who has used autopilot knows that failure to keep the hands on the wheel results in a warning.

We do know he received warnings. And we can strongly infer that he knew how to actively avoid receiving warnings. Therefore I can't possibly fathom how Tesla could be responsible for a "failure to warn".

Re: Tesla sued in wrongful death lawsuit that alleges Autopilot caused crash

#322

I would like to add over here that Tesla does indeed do sensor fusion in their cars. Their Autopilot combines radar and ultrasound with vision to decide where to drive. Commentators bringing up LIDAR are jumping the gun by assuming that this scenario isn't something that these sensors wouldn't have detected in either combination or individually to spot the anomaly. The problem over here is likely to be software due t…

While we may get to the point where sensors and a form of AI can reliably drive a car, we're so far from it that we don't even know if we're heading towards the goal post. What we need, if we want to solve the partial self driving problem right now, is not fully autonomous systems. We could start by agreeing on a way for cars to sense the road - think in the way robotic lawn mowers work where you have to bury cables.…

The problem we have right now is that:

1. laying cables wouldn't work because there are a great many historic towns and cities (eg Amsterdam) and digging up cobble streets to lay cables for self-driving vehicles simple isn't going happen when those cities have so many bigger problems for autonomous cars such as trams, cycle paths, pedestrian zones and tourists what will just walk out in front of everyone in an unpredictable haze.

2. you have districts that are so under funded that the roads aren't even maintained in their current state.

3. So lets say you scape redeveloping the roads and have radio sensors on lap posts - then you have the problem that many rural areas in even richer countries like the UK don't even have street lamps. And what about countries like Australia and America that can have long stretches for tens or even hundreds of miles without a single house, let alone street lights to put sensors on.

I agree with you that normally this kind of problem requires a technical approach but the problem we have is that redeveloping our roads would be such a monumental overhaul that it would actually be a bigger job than developing the AI to handle our existing infrastructure. Plus redeveloping our roads only solves part of the problem - you still have the issue of other cars on the road, pedestrians, cyclists, unexpected hazards like fallen trees, etc. You could fix other vehicles on the road by allowing all vehicles to communicated with each other - like a mesh network - but then you have the problem of doing that securely (so you couldn't just "virtually ram" another car off the road by telling it there's an obstruction that doesn't exist) and you again have the problem that it would be more expensive and take longer to retrofit every single vehicle with that tech than it would be to write AI. And you still wouldn't have addressed the problem of kids, animals and fallen trees.

AI might seem like a naff workaround and maybe if / when we colonise a new planet then we might be able to build roads with sensors inside tubes that restrict access (I guess a little like an underground rail network) but for the time being AI is our best bet for autonomous vehicles.

Re: Tesla sued in wrongful death lawsuit that alleges Autopilot caused crash

#323
post #278

That's what happens when you mislead your customers into believing your self-driving technology is far more advanced than it actually is. I hope Tesla loses this case and is forced to change their bullshit marketing before more people lose their lives.

But they actually didn't. They specifically tell you it is NOT a fully autonomous autopilot and that's it's more of an advanced cruise control. They also specifically tell you to KEEP YOUR EYES ON THE ROAD. If you fail to do so, you're asking for trouble. It is unfortunate the guy haf kids that will now grow up orphans, but judging from the information I've read in the Tesla blog post, it is clear that he was in the…

Not only does the car provide repeated warnings, we can pretty clearly infer that—like anyone who has used Autopilot—he had become reasonably proficient at avoiding the warnings. So not only was he warned, he was actively working to avoid those warnings.

(Edit: to those voting me down, please contribute your objections to the discussion.)

Re: Tesla sued in wrongful death lawsuit that alleges Autopilot caused crash

#324
Do they not have Rumble strips in the US?

https://en.wikipedia.org/wiki/Rumble_strip

I have been in situations in my youth where I was commuting or travelling tired, and nothing alerts you like the loud rumbling sound, and I'm sure the autopilot could detect it and bring the car to a stop if the driver hasn't been alerted by the in car warning system or the rumble paint...

Re: Tesla sued in wrongful death lawsuit that alleges Autopilot caused crash

#325

I would like to add over here that Tesla does indeed do sensor fusion in their cars. Their Autopilot combines radar and ultrasound with vision to decide where to drive. Commentators bringing up LIDAR are jumping the gun by assuming that this scenario isn't something that these sensors wouldn't have detected in either combination or individually to spot the anomaly. The problem over here is likely to be software due t…

Your comment actually emphasizes why having a lidar is better. Cameras need neural networks to detect objects. Lidars on the other hand provide distance to all the surrounding objects. The neural networks processing the point cloud may not be able to detect what the object is, but because of the distance available in the point cloud data you will know there is something ahead. Fusing radar with camera does help but r…

Radar seems to work pretty damn well for the adaptive cruise control and autonomous emergency braking installed in hundreds of models of cars made over the past 5+ years.

If autonomous systems needs more redundancy, the solution is to keep adding more cameras.

The other sensor which I think is important is data captured by cars in that fleet. This is sort of what's known as "HD maps" but fully automated and based purely sensor-derived data. So if you're driving down a highway at night in the rain, your car would know the successful paths traversed a few hours ago when it was dry and sunny.

Re: Tesla sued in wrongful death lawsuit that alleges Autopilot caused crash

#326

Tesla's blog post: https://www.tesla.com/en_GB/blog/update-last-week’s-accident the driver’s hands were not detected on the wheel for six seconds prior to the collision. The driver had about five seconds and 150 meters of unobstructed view of the concrete divider Looks like sensors failed to see concrete divider in nice sunny weather and car slammed in to it at 70mph. Driver was obviously over confident on system's a…

> Driver was obviously over confident

I wouldn’t assume that. Videos of this exact scenario prove that the reaction time required is on the order of a few seconds.

Everything Tesla says here exaggerates driver culpability, like saying he received “multiple warnings” (15 minutes ago and completely unrelated). Don’t let that hypnotize you into thinking he was being irresponsible here. There may have been other factors at play such as getting blinded by the sun during those few critical seconds of Autopilot error.

Re: Tesla sued in wrongful death lawsuit that alleges Autopilot caused crash

#327
post #225
post #151

Earlier quoted context omitted.

Tesla (and/or Musk) keep saying it's safe to let the car drive, and then when you let the car drive and it kills you, they say you're holding it wrong. If you're actually trying to increase safety, the right things to build are computers to supervise the human driver, and not systems that require a human to supervise the computer driver. If the human drifts out of the lane, nudge the car back in and beep, but don't p…

A computer supervising the driver is what got Boeing into problems with their MCAS resulting in the 2 plane crashes Tesla also doesn't say that it's safe to let the car drive. There are ample warnings and messages and feedback that say that you have to keep your eyes on the road and hands on the wheel at all times. There are likely more non-AP Tesla/non-Tesla auto accidents and fatalities than there have been AP rela…

What got Boeing into problems is management overpromising something engineering couldn't deliver, just like Tesla is doing with "autopilot."

Boeing promised a brand new and improved jet that was cheaper because it packed in a bunch of improvements without having to go through certification on a new airframe or pilot retraining. Engineering tried their best but either through negligence or "just following orders" and bureaucracy they failed catastrophically, as any engineer will tell you would happen with a non-redundant (in a 3+ vote configuration) sensor that could override pilot control inputs based on faulty readings

Re: Tesla sued in wrongful death lawsuit that alleges Autopilot caused crash

#329
post #276

Tesla's blog post: https://www.tesla.com/en_GB/blog/update-last-week’s-accident the driver’s hands were not detected on the wheel for six seconds prior to the collision. The driver had about five seconds and 150 meters of unobstructed view of the concrete divider Looks like sensors failed to see concrete divider in nice sunny weather and car slammed in to it at 70mph. Driver was obviously over confident on system's a…

> In the US, there is one automotive fatality every 86 million miles across all vehicles from all manufacturers. For Tesla, there is one fatality, including known pedestrian fatalities, every 320 million miles in vehicles equipped with Autopilot hardware. If you are driving a Tesla equipped with Autopilot hardware, you are 3.7 times less likely to be involved in a fatal accident. I didn't know Tesla sold motorbikes t…

> Why do they still do this? What's the point if not trying to mislead the public?

You're answering your own question there; they're doing it not with assumed evil intent ("let's lie about the safeties so that people kill themselves ! meuahahah") but for PR and stock reasons, but it ends up being the same result: they want people to see autopilot as "it drives for you" if you're a car buyer or an investor, but they also want people to see it as "it's just a helper, it doesn't drive you need to drive" if you're a regulatory agency, a judge or a customer that got into a crash.

As for the stats comparison removing motorbikes is a big one sure, but you could go one step further and take into account only the miles on free/highway and let's compare it again, because if the US is anything like here in France most accident are in the 15 Km around the drivers' home (probably because the driver feel "safer" because he knows the road) and that's also when autopilot is the least likely to be on. Meaning they're comparing themselves with a lot of miles where they're not competing.

And even further than that, no matter how safe/unsafe you are, if all your crash ends up with "the driver should have been handling the wheel but wasn't" then it doesn't mean all those drivers did it wrong, for me it's more a case of "you disengage/safety system is incapable of handling a case that's causing fatalities yet is common enough that all your crash report point at it".

Re: Tesla sued in wrongful death lawsuit that alleges Autopilot caused crash

#330
post #222
post #215

Earlier quoted context omitted.

Except there are multiple reports from driver that autopilot problematic in that location.

I imagine that the court or some regulatory agency could order to automate these kind of reports and use them to shutdown autopilot or make it unavailable on some routes. It would be interesting to design the ux for this. Based on the wording it could even make them more liable if accidents happen outside of roads with known issues.

Well at least Tesla drivers can and to openly report. When I had issues with phantom breaking etc on my BMW while driving on the Autobahn I never reported this since there was no sensible channel.

You just learned to live with the limitations and used it where it worked under supervision.

Post reply on HN