Google/Waymo is actually piloting a car from origin to destination without a steering wheel or pedals. That's a much more difficult challenge.
Put another way, no blind person is going to sit behind the wheel of a Tesla and tell the car to take him/her to the supermarket. Google's car actually does this, on real roads, today... and that's amazing.
https://www.youtube.com/watch?v=C3DbrYx-SN4
The issue is Google is doing it with (prohibitively) expensive hardware, so we don't know yet if Google's self-driving system works with cheaper Autopilot 2.0-level hardware that the other car makers are going to use as well.
Tesla had the luxury of running test drives until they had a success. This is a different thing than putting a car on the road with no steering wheel and saying bon voyage. The first is a demonstration of a single successful instance (which is probably repeatable on average); the second is a much stronger demonstration of a > 99% success rate.
https://www.aaafoundation.org/sites/default/files/2011Pedest...
That's exactly what Tesla is marketing and working on with their new self-driving features, and you can buy cars that will allegedly be capable of this today.
"All you will need to do is get in and tell your car where to go. If you don’t say anything, the car will look at your calendar and take you there as the assumed destination or just home if nothing is on the calendar. Your Tesla will figure out the optimal route, navigate urban streets (even without lane markings), manage complex intersections with traffic lights, stop signs and roundabouts, and handle densely packed freeways with cars moving at high speed. When you arrive at your destination, simply step out at the entrance and your car will enter park seek mode, automatically search for a spot and park itself. A tap on your phone summons it back to you."
It should be obvious that Tesla is going to win this race. Tesla already does fully autonomous driving, they released that video in November. And they build real, production cars. "Real artists ship", as Steve Jobs said.
If you a manufacturing ace, would you join the company with a long list of failed projects and no actual production cars, or would you join the company that is building 500,000 of the most advanced autos right now?
If I was a hardware expert, would I join Google who shies at giving a fraction of it's cash to hard tech, or Tesla who buys german automation companies?
Or plain old software. Tesla has magnitudes more data and can update neural networks to hundreds of thousands of cars at a whim. Soon millions.
The trouble with just throwing the statistics around and saying "it's fewer deaths!" is that it has to be fewer deaths in comparable situations. A tesla having fewer deaths per mile in perfect driving conditions shouldn't be compared to a possibly-drunk person in possibly-awful weather. This isn't the media creating controversy, it's people expressing skepticism when a corporation's incentives are to let a couple people die, and trying to maintain a high bar.
On the other hand, if you think fault doesn't matter and it's just one life for another, then it essentially becomes the trolley problem, which doesn't have a clear answer either.
If you can go from 30000 to 1000 just by changing to good enough autonomous cars that's a change worth doing. Then you can look at improving it to 0 and encourage that process if not through insurance liability processes through regulations that require ongoing improvement in autonomous safety.
That video, for reference, is: https://www.tesla.com/videos/full-self-driving-hardware-all-...
I'm not super convinced by that video. That drive had almost no challenges/obstacles. No bicycles, no construction, basically no pedestrians, no other vehicles behaving erratically.
Google has cars that can respond to everything from a cyclist making hand signals, to a school bus (which must not be passed), to a police car pulling it over, to a woman chasing a duck across the road: https://youtu.be/tiwVMrTLUWg?t=8m49s
I'm also not sure this is a simple case of "throw data and hardware at the problem." If it were that easy, it wouldn't have taken Google so long to get to where they are today.
https://www.tesla.com/videos/autopilot-self-driving-hardware...
So either the problem in practice is much harder than a simple video can show, or something else is up.
Though, I think the real issue is that google wants to go straight for L5 (meaning 100% automated, and you can remove the steering wheel). As compared to an L4, which is more like 99.9% automated, but you still need steering wheel for those rare edge cases.
I'm not sure there's a better spot in Austin to test autonomous cars.
That said, they aren't perfect. 1. They wait just long enough at stop signs when it is their turn, which confuses other drivers. 2. They drive well under the limit, causing small jams at peak congestion.
Given that Tesla doesn't even have emergency braking and auto-steering, I'd guess they compare quite favorably...
So you can say that Tesla is at least 1-years behind. More like 4-5.
That is extremely aggressive. Getting data is the easy part of self-driving.
We haven't heard much from Google because they've been programming and validating all the unglamorous little problems that training data and a neural net can't solve alone. 4 way stops, passing motorcyclists and stuff like that, where old fashioned hand coded features are necessary to adress situations in which specific kinds of reasoning is necessary.
Tesla has to solve all this boring stuff as well before they can offer comparable capabilities in an AV.
I don't think hand coded features are necessary. No one hand codes humans to drive.
I don't believe your hand coded features will outperform ML trained over a billion miles, even in the long tail rare events.
Sure ML is bad at rare events, but hand coded features also map poorly to such rare events, because it's unlikely you'll be able to imagine them all, or even manage to handle categories of rare events well.
Furthermore, I believe it would be possible to train an ML based rare-event detector, which simply detects the difference between "regular driving I'm familiar with" and "something funny is going on, I haven't seen this much in the training dataset", and then refer to a remote human to resolve the issue.
Successful remote control is a whole different project, very different from self-driving. In order to give the remote pilot enough situational awareness, you need plenty of streamed imagery (recreated or recorded). And you need a feedback loop with a very small latency. Not saying it couldn't be done, but it won't happen "organically" by cars becoming self-driving.
* that road with the one-foot deep pothole (it's more like a sinkhole at that point, though) in the middle of a lane
* the perpetual construction zone where there are five layers of lane markers, only one of which is the actual layer of lane markers
* the 10% grade two-lane winding road in the dark with 2 inches of snow on the road and more coming
* open highway with a constant 30mph crosswind
* dense fog with less than 100m visibility
* torrential downpour with so little visibility you can't even see the lane markings on the road right in front of you
Those are all conditions I've driven in, and I can think of several other horrible driving conditions that I haven't yet had the (mis)fortune to drive in. Just because you can drive well in good conditions doesn't mean you can drive well when things get tougher--as the driving records of many human beings can well attest.
We can likewise say there are some situations a machine shouldn't be driving.
As long as the machine can identify when it is incapable as a driver, it's only the same as the human pulling over for a nap because they can't stay alert.
Sure, if you ran a taxi service with such a system you'd have to send a regular taxi to pick up the customer, but it's only the same as any other kind of breakdown like a puncture.
And "autopilot" has been on the market for over a decade, it's OEM off-the-shelf equipment.
[0] https://web.archive.org/web/20161016050722/http://www.google...
[1] http://www.google.com/selfdrivingcar/files/reports/report-09...
EDIT: I think I triggered them all to be archived, links in [0] should work now.