breser
AutoPilot Nostradamus
The post I quoted was talking about auto-pilot like systems. Not autonomous cars, they aren't the same. But sure both auto-pilot and autonomous cars will be safer, but there will still be accidents.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
The post I quoted was talking about auto-pilot like systems. Not autonomous cars, they aren't the same. But sure both auto-pilot and autonomous cars will be safer, but there will still be accidents.
perkiset
... this one goes to 11
Well, in terms of "deeply naive" I am thoroughly convinced that if I have absolute control of a system, as opposed to building a system that must interact with humans, I will have better results. But, call that only the experience of someone having been a programmer for 37+ years.
I'll execute a rare and controversial absolute: If we could program a system, end-to-end, with utter control of the vehicles involved, it would behave pretty damn well. Certainly better than our human system today. And certainly better than a controlled system that has to interact with the vagaries and inexplicable behaviors of humans. We'd have more efficient and speedier transport with lower incident of fatality. It would be adherent to a known algo and easily explained. As opposed to the highways of today, that are utterly inexplicable.
Let me start by saying that I am not an owner of Tesla (not even a perspective one). I am mostly interested in self driving technology which I why i follow this thread. So, coming back to the point, with respect to above quoted comments, you guys need to watch this video of Chris Urmson who is the head of self driving cars at Google. This is the exact problem he was referring to (Clip starts at point where its relevant but the whole video is also good and informative)
Chris Urmson: How a driverless car sees the road - YouTube
Here is a similar statement from Teller
"Teller says that Google is trying to get "stumped like the duck" again, even though there is a much easier solution for getting to market: highway driving. In the fall of 2012, Google had already given out Lexus SUVs to non-Google employees to use — and on the freeway, they worked perfectly well. "We probably could have made a lot of money selling those," Teller said. But something surprising happened: "Even though people had sworn up and down 'I'm going to pay so much attention,'" and there were cameras watching them too, "people do really stupid stuff when they're driving." Basically, "the assumption that humans could be a reliable back-up for the system was a total fallacy!" Teller said. That's why Google began to pursue a car that doesn't even have a steering wheel or a gas pedal."
breser
AutoPilot Nostradamus
You're still going to interact with humans even if every car on the road is autonomous. I don't see pedestrians or bicycles going away.
I'm pointing to real world systems that are implemented in far more ideal circumstances (same protocols for collision avoidance with a great deal of effort to make sure two different planes don't recommend evasive actions that increase the likelihood of a collision) and yet there are situations where the evasive actions recommended increase the likelihood of a crash.
But I need not even look very far to find a TACC accident:
3 day old import P85D crashed while using TACC
That particular accident happened because of an implementation detail of Tesla's TACC. Stopped objects are presumed not to be in your way. If it sees the object moving and then it comes to a stop it will consider it. I guess you could say that the car moving out of the lane where there is a stopped car is an inexplicable human behavior. But my point is it might be entirely reasonable for an autonomous car to decide to change lanes due to stopped traffic ahead of you.
I don't deny that computers ought to be able to do a better job in most cases than human drivers. But I you said that if we had auto-pilot type systems then we'd have "no problem." Now you're walking that back to "pretty damn well." They will improve safety but they will not eliminate all accidents.
To be clear I'm also not advocating that these systems need to be perfect to be adopted, perfect is the enemy of good. But I think it's important that we have realistic exceptions of the results.
dhanson865
Well-Known Member
You forgot the one from the days before windshields.
Driving over 20mph will kill you, you'll die of asphyxiation.
If the car is truly autonomous (level 4) then the car will handle pedestrians and bicycles. We're no where near that.
Not direct at you, but related to what you said:
I agree that we need realistic expectations of results. People here hear autopilot (level 2) and mentally think autonomous highway driving (level 3). No that's not what's being sold. Never was advertised that way. Is not possible with current hardware either.
Autopilot lets you not touch the pedals and (hopefully) not touch the steering wheel while the car holds you in the lane you're traveling and making sure you don't rear end the car in front of you. It wont navigate you from point A to point B. It wont change lanes for you. It wont decide where you need to go at a fork on the road. Currently TACC wont slow down for you when the speed limit changes (though I bet they can implement this in software). Etc.
As to another unrealistic expectation, people keep talking about the car automatically reparking itself at superchargers once the metal snake comes out. I don't see that happening any time soon either. The car can not detect soft objects (pedestrians, animals, etc.) and how will the car decide what's a legal parking spot?
It's one thing to have the car meet you at the door, on your private property, driving down a driveway with no other pedestrians there. It's another thing to do it at a crowded supercharger. Way down the road? Sure. In the next year or two? I don't see it happening.
Soolim
Member
Maybe as OP foretell Tesla can silence him by enrolling him on the beta AP.:wink: Or he got a sweetheart deal?
Cyclone
Cyclonic Member ((.oO))
He was scheduled for the LTE upgrade retrofit yesterday. I imagine there were some conversations at the SvC.
Andyw2100
Well-Known Member
The autopilot demonstration at the D event did include the cars changing speed as the speed limit changed. So I would expect that feature should be coming soon.
You're right on many levels, but today I'm driving a loaner with TACC. I'm freaked out. Will it come to a stop when the car in front of me is stopped? I don't know... will it? Yes, it seems that it will... but just in case, I've got my foot over the brake and ready to intervene... software works great so far, but the fear is still in the back of my mind and causing me very mild anxiety while using this feature. Can't help it.
rxlawdude
Active Member
No need to freak out. Cover your brake, and watch for the one situation where I (and others) have observed it doesn't stop as expected: when you're traveling at speed and all traffic (including that car in your lane ahead) is at a full stop, TACC seems to not recognize the stopped car until it gets so close the collision warning alarms.
Other than that, I've found it flawless.
Soolim
Member
As in, when the cruise control indicator is in grey (i.e. not tracking the vehicle ahead), be ready to apply brake before you hit a stationary object? or else the car emergency brake takes over.
Andyw2100
Well-Known Member
If the car in front of you isn't being tracked when it comes to a stop (and in your example it wasn't), TACC isn't supposed to recognize it at all. I wrote this in another thread recently:
I would recommend reading the manual for any vehicle and/or features you're unfamiliar with before operating it. That goes for a Tesla or a Honda, and I don't mean that to be snarky. If you haven't read about TACC's operating procedures and how it functions, don't use it.
I read it long ago and have followed all of the TACC threads. My comments were from the standpoint that I have trouble trusting this kind of technology, and even though I've familiarized myself with the technology, I can't change my instinctive though process of wondering if it's going to work "this time".
I was one of the people who complained about that case. Since the new update though, I've been very pleased with the car's performance in exactly that circumstance, it starts slowing down reasonably far back and comes to a stop at a safe distance with strong but not annoying deceleration. I must admit though that I still cover the brake just in case. The only time I was a bit worried was when I accidentally engaged the TACC with a car stopped in front of me. The Tesla started to accelerate, but very quickly braked heavily just as I hit the brake.
Similar threads
