Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Another Tesla on autopilot steers towards a barrier

reddit.com

Post reply on HN
Print view View original

121–130 of 481 posts

Re: Another Tesla on autopilot steers towards a barrier

#121

Post by edf13 »

lvoudour wrote: post #47

I wonder what "AI updates" are going to mean from an insurance point of view. Don't know about the US, but some countries in Europe insure the driver and not the car (eg. Ireland). How far fetched is for insurers to argue that an autonomous car software update constitutes a change of driver and demand that the updates must first be validated from them before installing?

It's a good point.

Here in the UK I am insured to drive vehicle A as specified on my insurance documents. If I modify that vehicle my insurance policy is void and I need to inform the insurance co. and probably pay a premium.

In your scenario the same could be true on a software update.

Re: Another Tesla on autopilot steers towards a barrier

#122

Post by lisper »

Here is my armchair diagnosis: right before the car veers towards the barrier it drives through a stretch of road where the only visible lane marker is on the left. Then the right lane marker comes into view at about the point where the lane starts to widen out for the lane split. The lines that will become the right and left lane markers of the split left and right lanes respectively are right next to the van in front of the Tesla, and at this point resemble the diamond lane markers in the middle of the split lanes. My guess is that the autopilot mistook these lines for the diamond lane marker and steered towards them thinking it was centering itself in the lane.

If this theory turns out to be correct then Tesla is in deep trouble because this would be a very elementary mistake. The system should have known that the lanes split at this point, noticed that the distance between what it thought was the diamond lane marker and the right lane line (which was clearly visible) was wrong, and at least sounded an alarm, if not actively braked until it had a better solution.

This is actually the most serious aspect of all of these crashes: the system does not seem to be aware when it is getting things wrong. I dubbed this property "cognizant failure" in my 1991 Ph.D. thesis on autonomous driving, but no one seems to have adopted it. It's not possible to engineer an autonomous system that never fails, but it is possible to engineer one in such a way that it never fails to detect that it has failed. Tesla seems to have done neither.

Re: Another Tesla on autopilot steers towards a barrier

#123

Post by notlob »

rayiner wrote: post #33
fairpx wrote: post #13

Autopilot tech in general should be open and shared among ccompanies and universities. The more brains we can put on this the better imho. If these things happen too often, Im afraid regulators will start making laws that prevent the use of this tech, which I believe has the potential to really make life better for all... one day.

The regulators have been really easygoing so far with automated cars. I strongly suspect its because they take the optimistic comments of the tech industry at face value that these technologies are road ready. Meanwhile, I think Tesla and Uber perceive this as more of an open alpha. This will not end well.

For someone in the medtech space, it's been interesting to see how loosely (not at all?) marketing is regulated in automotive and tech in general. Were the FDA regulating automotive, there is absolutely no way that Tesla's driving assist package could be marketed with the name "Autopilot", for one. More broadly than just marketing, it's also amazing to me that Tesla could have released a feature (I know they've revised it since) that was designed for human supervision but did not require hands on the wheel for long stretches of time. What did the risk management look like there?

Re: Another Tesla on autopilot steers towards a barrier

#125

Post by greedo »

vorpalhex wrote: post #63
iainmerrick wrote: post #54

This kind of video makes me wonder: what is the "autopilot" feature actually for? Do people generally like it? (I don't own a Tesla and I've never driven one.) If you're supposed to keep your hands on the wheel, and given videos like this that show you really need to keep your hands on the wheel and pay attention, is automatic steering really that big of a deal? Cruise control, now, that really is useful because it a…

So human drivers are terrible at driving for any long period of time, especially on relatively consistent "boring" highways. Guess what computers excel at? Driving consistently on consistent highways. The Tesla Autopilot is supposed to be the always aware and paying attention portion on these cases where a human driver would be very likely to start texting or dozing off. Now, it's not fully autonomous and may well de…

Expecting "drivers" to be able to instantly switch from "somewhat zoned out" to having the situational awareness to resolve the problem that AP fails at is unrealistic.

Re: Another Tesla on autopilot steers towards a barrier

#126

Post by comboy »

swiley wrote: post #57
tjwds wrote: post #2

This comment from OP (beastpilot) is pretty frightening: > Yep, works for 6 months with zero issues. Then one Friday night you get an update. Everything works that weekend, and on your way to work on Monday. Then, 18 minutes into your commute home, it drives straight at a barrier at 60 MPH. > It's important to remember that it's not like you got the update 5 minutes before this happened. Even worse, you may not know…

And this is why if I ever get a self driving car I want: 1) root 2) control over updates 3) everything to be completely open

You do not want root on your self driving car.

Or to put it differently, I don't want to be driving on the same road as you with your rooted self driving car. You can be great sysadmin/coder, tesla guys may be too, but both of you changing random stuff without any communication with each other.. I've seen enough servers.

Re: Another Tesla on autopilot steers towards a barrier

#127

Post by kalleboo »

iainmerrick wrote: post #54

This kind of video makes me wonder: what is the "autopilot" feature actually for? Do people generally like it? (I don't own a Tesla and I've never driven one.) If you're supposed to keep your hands on the wheel, and given videos like this that show you really need to keep your hands on the wheel and pay attention, is automatic steering really that big of a deal? Cruise control, now, that really is useful because it a…

I don't own a car so I drive very rarely (once a month or less). I recently went on a week-long roadtrip in a rental Subaru which had their EyeSight system which does adaptive cruise (follows the speed of the car in front) + lane keeping. I'd say at the end of the day I was noticeably less exhausted than I usually am after a day of driving.

The Subaru system will not let you let go of the wheel for more than 15 seconds (after that it will instantly disengage), so it's more to save your effort of continual minor corrections. It also disengages as soon as it's confused in the slightest (faded lines, lines at an angle etc)

Re: Another Tesla on autopilot steers towards a barrier

#128

Post by 20after4 »

bufferoverflow wrote: post #94
gormz wrote: post #40

You can pry the steering wheel out of my cold dead hands. Might be true that on paper these things cause less accidents and less deaths, but I want to be in control of my life. If I die in a car I don't want it to be because of a software bug.

Most modern cars are drive-by-wire, entirely relying on computers and the software. And their software is huge blobs of proprietary compiled code. It's guaranteed to have bugs.

Probably not that huge, really. Embedded systems on most vehicles are fairly minimal and reliable, at least compared to modern consumer operating systems and (apparently) Tesla's systems.

Re: Another Tesla on autopilot steers towards a barrier

#129

Post by PaulHoule »

iainmerrick wrote: post #91
Naritai wrote: post #87

Earlier quoted context omitted.

To directly answer your question: people buy autopilot so they can check their phone while commuting down the freeway to work. I was talking to a group of people about the difficulties of commuting in the Santa Clara Valley, and one guy shrugged and said, "just get a Tesla, move to the left lane, and manage email until you get to work".

Yikes! I really hope that's a joke...

I see people attempting that on the roads around here without the Tesla. That is why I always honk at cell phone drivers. Sometimes I wish I had a fully-charged Marx bank in the trunk.
Post reply on HN