Who downvoted you? This is a very serious issue. A partial autopilot for cars is worse than useless: it will create distracted drivers who can't react when the car encounters something it can't handle. Self-driving cars need to be either 100% or 0%.
You need to be both looking at the road with cameras and profiling it with LIDAR. (Or terahertz radar, once that gets going.) It's not enough to just sense the car ahead. You need to be able to detect potholes, ice patches, junk on the highway, small animals, and similar problems. We could detect and avoid potholes back in 2005. Since we were doing off-road driving, that was a normal driving event.
The reason for a high-view LIDAR is that you want to see the pavement surface ahead from a reasonably useful angle and get a 3D profile of the road ahead. Google uses the Velodyne spinning-cone LIDAR scanner, which is a lot of LIDAR units built into one rotating mechanism. That's a research tool. There are other LIDAR devices more suited to mass production. Advanced Scientific Concepts has a nice eye-safe LIDAR which can operate in full sunlight. It costs about $100K, but that's because it's made by hand for DoD and space applications. The technology is all solid state, not inherently that expensive, and needs to be made into a volume product. (Somebody really needs to get on that. In 2004, I took a venture capitalist down to Santa Barbara to meet that crowd, but there was no mass market in sight back then. Now there is.)
You can only profile the road out to a limited distance, regardless of the sensor, because you're looking at the road from an oblique angle. Under good conditions, though, you can out-drive the range at which you can profile the road. That was Sebastian Thrun's contribution, and won the DARPA Grand Challenge. The idea is that if the LIDARs say the near road is good, and the cameras say the far road looks like the near road, you can assume the far road is like the near road and go fast. If the far road looks funny, you have to slow down and get a good look at the road profile with the LIDARs.
Automatic driving systems have to do all this. "Driver assistance" systems don't. Hence the "deadly valley".
That's just to deal with roads and static obstacles. Then comes dealing with traffic.
Tesla is talking about automatically putting their car into a garage. That's a good application; it's slow, and you can have sensors all around the car to avoid hitting anything. A more general system that can put a car into a big parking garage or lot is quite possible. Cooperating parking garages might have some additional bar-code markers and maybe a data link for open space info.
The whole airport car-rental thing could be done automatically, using slow-speed automatic driving to bring the car up to a pickup point at the terminal, just as the renter gets there. That may be one way this gets deployed. (I proposed that around 2003, but after 9/11, the idea of autonomous vehicles in an airport seemed politically hopeless.)
Those are some ways this might be deployed. Everyone has obsessed on automatic freeway driving since the 1950s, but that may not be the killer app.
Though, with many self-driving cars perhaps the sensors on each could share info, thereby mapping a larger (if not, endless) area.
My prediction: "crashdummy" will eclipse "heartbleed" and "shellshock"!
Humans proved that for current level driving two not particularly high-resolution cameras are sufficient. Seems like pushing in this direction will remove this expensive component?
Edit: a better solution would be to observe the behaviour of other drivers; if there is someone driving ahead of you, you can assume that the road between you and them is OK; if there's noone ahead of you, you need to drive slower and be more careful (that's how I drive at night). Once there's a critical mass of cars with cameras, cars could communicate road conditions automatically.
I would be frightened to trust the data coming from a random car in front of me. Inferring road condition from another car behavior sounds reasonable, using data supplied from it, not so much.
However, if the car three cars in front of you just broadcast "I'm doing an emergency stop right now", that's really valuable data. The human in your car won't know anything is wrong for at least a second. The human driver behind you would know about it before the human driver in front of you.
Until some asshat implements a button that sends this message for "lulz", though that would probably constitute a crime under existing law.
A computer-driven car, though, wouldn't (shouldn't) just immediately slam on the brakes because of that signal. It would tighten seat belts and start slowing down, but it also would want to avoid getting rammed by the car behind it. It can make very accurate estimates about its stopping distance and use all of it.
You get to use worse hardware, but you need several order magnitude better software.
The human eye can instantly recognize the available driving paths, the motorcyclist ahead, and project where people will walk. Software would have to parse out where the open roads are, how far that motorcyclist is and whether he can clear the intersection before the car reaches it, and what that sign on the right-hand side is—using the same information, but it has to parse it first whereas we do that almost instantly. It's a totally different game.
I expect that, after the first wave of clumsy LIDARing self-driving cars, all the car companies (Google especially) will be collecting exactly training data from the cars' sensors to build exactly this kind of model. In fact, I wouldn't be surprised if that was what the Google car was really about, in the same way Google Voice is really about collecting speech training data.
The best part of this kind of training data is that it all comes pre-annotated with appropriate reinforcements: even if the image-recognition sensors aren't hooked up to anything, they're coupled to the input stream from the other car sensors and the driver's actions. So you would get training data like
- "saw [image of stopsign], other heuristically-programmed system decided car should stop, driver confirmed stop."
- "saw [image of kitten standing in the road], other heuristically-programmed system decided car should continue, driver overrode and stopped car."
Etc. Aggregating all these reports from many self-driving cars, you could build an excellent image-to-appropriate-reaction classifier.
In a driving situation, the worst-case scenario is everybody dies.
That's just my theory on it.
The simple answer is that we don't know how to do it with stereo vision alone yet. Getting range reliably is hard, and your brain uses lots of tricks to do it.
The second answer is that we need better-than-human performance if this is to take off. So using human-type sensing might not be good enough anyway.
Lots of researchers are pushing on vision-based driving, though.
It is one thing to anticipate threats, but when the road is under snow how do computers make that intuitive leap people can?
http://en.wikipedia.org/wiki/Air_France_Flight_447#Final_rep...
"the crew lacked practical training in manually handling the aircraft both at high altitude and in the event of anomalies of speed indication; the two co-pilots' task sharing was weakened both by incomprehension of the situation at the time of autopilot disconnection, and by poor management of the "startle effect", leaving them in an emotionally charged situation;"
A great deal of the required operations for autodriving are offloaded or irrelevant in autopiloting. I know we were talking about an isolated case of the unexpected failure of an automated piloting system, but if we consider all of these additional complexities in typical road driving * millions more cars with barely-trained operators, it's reasonable to suppose that those types of problems would occur much more frequently with autodrivers than autopilots.
First, one of the pilots held the stick back all the way down to the water even though the airplane was clearly stalled. This is completely inexcusable. The automotive equivalent would be holding the accelerator pedal to the floor while aiming at a brick wall. Even worse, the guy who did this was a qualified glider pilot, and glider pilots should have extreme familiarity with stalls and stall recovery since so much of a glider's flight time is spent close to stall.
Second, the control system of the plane handled conflicting inputs from the two pilots by averaging them together. There was no indication to the pilots that they were fighting each other. Positive exchange of control of the aircraft is another really basic thing about flying, even more basic than putting the nose down in a stall. This setup made it far too easy for the pilots to be unaware of who was really flying the plane. If the control sticks moved together, it would have been obvious to the other pilot what was going on.
One could put some blame on the autopilot in causing the one pilot to forget the basics of stall recovery, but it seems to me that the real fault was in not training for it sufficiently, since that's not something you do in normal flight in an airliner anyway.
Keep in mind every car out there is currently built with a feature where you can accelerate to a lethal speed, then press a small button, and the car continues to travel at that lethal speed until the car runs out of gas, hits a brick wall, or the driver consciously disengages it. We call it cruise control.
If all we are doing is adding automatic braking and lane following to cruise control, I don't see how it can be anything other than better.
Because in the real world, people won't exercise exactly the same level of situational awareness, such that the autopilot only adds a safety mechanism. In the real world, it will replace a certain amount of driver attention. And the question is, is it always superior to human attention?
It's probably net superior, but I bet it will cause a couple of high profile accidents that a careful human could have avoided, and then people (or the government) will be scared of it.
The more valuable question is this:
Does it perform better than humans in the most frequently occurring situations that lead to incidents, are those situations a large enough fraction of all incidents such that the number of overall incidents drop significantly, and will that drop be greater than the increase in frequency of less predictable situations.
If the above turns out to be the case, then there will be fewer accidents despite less aware drivers.
edit: case in point
The admins explained it that the Microsoft news triggers the flamewar prevention they built, and it happens almost every time. Microsoft articles need admins to explicitly allow them on the front page, or they're gone in minutes.
Or I could refuse to accept the premise internet points are important and realize that downvotes aren't a reliable indication of post quality.
I'm certainly not going to be bullied by companies in to being quiet because they might take away some of my internet points.
Please RT.
Your anecdote is insufficient to support this claim. One example does not make this a common practice.