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

Autonomous DeLorean drives sideways to move forward

news.stanford.edu

Post reply on HN
Print view View original

161–170 of 197 posts

Re: Autonomous DeLorean drives sideways to move forward

#161

Post by octoboto »

jsinai wrote: post #78
sillysaurusx wrote: post #35

Earlier quoted context omitted.

I would highly recommend https://www.youtube.com/watch?v=PYylPRX6z4Q as a way of framing the problem. It seems like our options for ML are currently: self-play, dataset fitting, or human feedback. It's possible to solve any problem that can fit into one of those molds. (Unfortunately the opposite is true: if a problem doesn't fit into those categories, it seems very hard to solve with any form of ML.) But, now that I…

Neural Networks are good at curve fitting which makes them useful for inference +- noise. In control systems you need to do inference to predict feedback signals +- noise. So you can use a neural network for that. Train it on input and feedback data from your samples. Your control system still needs to account for predictability, smooth behaviour etc so you have the rest of your machinery to handle that. This is not…

There's https://journals.sagepub.com/doi/abs/10.1177/027836491985942.... I think Aaron Ames' group is working on theoretical guarantees of neural net policies as well.

Re: Autonomous DeLorean drives sideways to move forward

#162

Post by dmurray »

RandomBacon wrote: post #122
charliepark wrote: post #39

Earlier quoted context omitted.

> "The trouble with pure neural net controllers is that they sometimes do something totally bogus for some data point within the normal input space." A friend's theory on Google Maps is that it sometimes intentionally sends you down a non-optimal path in order to gather data on alternate routes. This friend isn't in the ML space, so this is just their anecdotal observation. Are they right, then?

That is my theory as well. It sometimes suggests a route I don't normally take, then when I continue on the route I normally take , the ETA updates to say I will now arrive sooner. Google Maps intentionally suggests to take longer routes.

It usually suggests three routes to me, each with a time estimate, and the highlighted one doesn't always have the least expected time. I figured it might be respecting my other preferences, but it could equally be trying to balance traffic across the whole network. Though this also happens when cycling, which has a minimal impact on congestion.

Re: Autonomous DeLorean drives sideways to move forward

#163

Post by majormajor »

wongarsu wrote: post #77
charliepark wrote: post #39

Earlier quoted context omitted.

> "The trouble with pure neural net controllers is that they sometimes do something totally bogus for some data point within the normal input space." A friend's theory on Google Maps is that it sometimes intentionally sends you down a non-optimal path in order to gather data on alternate routes. This friend isn't in the ML space, so this is just their anecdotal observation. Are they right, then?

"Alternate routes" are still normal roads people travel on. Since the travel time of a path is just the sum of the travel time of all path segments you only need people traveling every road to estimate the total time of any path. Path finding is well understood with many algorithmic solutions, so there is no reason to use machine learning. Google's implementation is impressive, but I see no reason why it would benefi…

Some path times are destination-dependent in ways Google doesn't seem to handle well, like certain turns or freeway onramps vs the regular lanes next to them. You would need people "testing" those parts, not just the regular path parts.

Re: Autonomous DeLorean drives sideways to move forward

#164

Post by fourthark »

ehnto wrote: post #114
fourthark wrote: post #104

It sounded like it precomputed the route, steering, drift for the whole course first, then executed it. If the friction were variable and unknown it would get completely lost.

Which is very common, even in controlled drift events. Over the course of a single run, tires can completely change grip coefficients.

So you got the sense that it was recalculating based on the response it got?

I wasn’t clear on that from the article.

Re: Autonomous DeLorean drives sideways to move forward

#165

Post by Retric »

wongarsu wrote: post #113
krallja wrote: post #107

Earlier quoted context omitted.

Google Maps is not just path finding; it also takes real-time traffic info into account.

Which is still just path finding: estimate travel time for each segment, use travel time instead of distance as edge cost. Don't take that as a dismissal of Google maps: it's the most impressive path finder I know, working with high performance over long distances, taking into account real time information, and calculating meaningful alternative routes to choose from. It's an impressive piece of technology that actua…

Gathering data is required to do travel estimates. If Google was using handset data to update congestion information, it would want to send a percentage of users down sub optimal routes to verify they are sub optimal.

Re: Autonomous DeLorean drives sideways to move forward

#166

Post by DreamScatter »

tzs wrote: post #141

I wonder if it would be feasible to make a fleet of race cars that have self driving technology that has each car communicating with the others and monitoring the human drivers, with the car letting the human drive but able to step in if the human tries to do something that would cause a crash? I bet you could make some pretty good money with such a fleet running open races that members of the public can pay to drive…

I doubt race car drivers (who are better drivers than the average car driver) would want a computer to yank their steering or do something unexpected in a race.

Autonomous cars aren't necessarily better than race car drivers.. would be interesting if human racers and autonomous compete

Re: Autonomous DeLorean drives sideways to move forward

#167

Post by Ericson2314 »

aeternum wrote: post #151
RandomBacon wrote: post #122

Earlier quoted context omitted.

That is my theory as well. It sometimes suggests a route I don't normally take, then when I continue on the route I normally take , the ETA updates to say I will now arrive sooner. Google Maps intentionally suggests to take longer routes.

It could be that Google Maps is saving everyone time by doing this. The Braess Paradox is really interesting, and this could be an attempt to prevent it: http://vcp.med.harvard.edu/braess-paradox.html

I would imagine introducing cars to a road itself causes a braess paradox.

Re: Autonomous DeLorean drives sideways to move forward

#168

Post by DreamScatter »

zzzeek wrote: post #129

they have gotten this car to do complex and exotic movements via automated manipulations of the car's human-oriented interface, that is, a single steering wheel, four wheels that turn in the same direction, and then the brakes / accelerator. It seems though that if you are actually trying to design autonomous cars that feature greater maneuverability than what is normally possible you'd instead ditch the whole assump…

The people working on this car are designing a vehicle control system, not an actual new vehicle.

Re: Autonomous DeLorean drives sideways to move forward

#169

Post by thr0w__4w4y »

btbuildem wrote: post #145
sap9433 wrote: post #135

If self-driving tech has become so advanced, I wonder why Waymo is not delivering anything? I wonder why is Tesla taking forever to deliver self-driving, why Uber discontinued its self-driving program, Why Apple ain't serious about Apple car yet? Does this demo hide something that we should know?

I think in this demo they simply paint a path for the car to follow. The cones etc are there just to help us see whether it's following the path well. If there was a cone in the middle of the path, it would've run right over it.

Yes, this. Not trying to take anything away from this, but this car is essentially "running a program." I would love for someone to come on here from the program and publicly repudiate what I'm about to say. I bet they ran that "dance" hundreds of times, no cones or hay or obstacles, got it to the point where it was precise and repeatable enough to then set up the cones and barriers (I mean, we have robots that place parts to within 0.001" or better repeatably).

What would impress me (seriously). I go out the L.A. Colliseum parking lot, set up cones / a track that the car can definitely do (based on radiuses & turns in this video), the car "studies" the track (I don't care if a drone looks at it & takes measurements), and then the car calculates and executes a perfect path through the course the first time without hitting anything.

Again, not trying to crap on this, I'm just saying that anyone who's worked in controls, automation, engine control, etc. (I've done all these & more) isn't utterly wow'ed by this. Kudos, but we didn't just put a man on Mars.

Re: Autonomous DeLorean drives sideways to move forward

#170

Post by HeWhoLurksLate »

dharmab wrote: post #143
floatingatoll wrote: post #96

Earlier quoted context omitted.

How quickly can a car measure the available traction and respond to brief loss events? Can on/off detection response times and recovery strategies be improved for black ice (and hydroplaning)? If a car hits black ice, should it power itself down or should it max out the CPU looking for the instant traction returns on any wheel and do whatever it takes to try to slow down?

Modern traction control systems available on production cars and motorcycles have a polling rate of up to 1000Hz. You don't need self driving tech to handle most low traction situations. Black ice is an exception because once you start sliding on it there is not much you can do to regain control except ride it out and hope you don't hit anything before you get out of the ice.

My mouse has a 1000Hz polling rate. Is 1000Hz really good enough for excellent traction control? I'm sure it's fine for most situations, but the small amount that it's not good for are probably the ones that matter the most.
Post reply on HN