Have any of these cars ever gotten stuck? Accidents are one thing, but has a car ever been able to traverse a great distance (10+km of public roads) with the 99.9999% reliability such a product would require?
You call the car. It leaves the garage and starts heading to you. Then something comes up, something like a construction zone or other odd situation. Or perhaps the car breaks down. So the car is now stopped and alone. You're sitting by the road five miles away and the car is parked blocking traffic somewhere. Who comes to help? AAA? Wil it fight back if someone tries to tow it away? Or push it onto the shoulder to free up the lane? I doubt the general public would have much sympathy for the guy who's tesla is blocking an intersection because he didn't want to pay for parking and/or bother walking to the car himself.
I cannot see drivers using this garage/parking trick very often. It's just too slow.
Reading the article, it sounds like the car isn't expected to travel more than 10 meters or so right now.
>Using Summon, once you arrive home and exit Model S or Model X, you can prompt it to do the rest: open your garage door, enter your garage, park itself, and shut down. In the morning, you wake up, walk out the front door, and summon your car. It will open the garage door and come to greet you
So from your garage to the street, or from a parking spot to out of, but adjacent to, the parking spot. Not from your home to your workplace.
Today. But Tesla and Musk have every intention to expand the range. What we have now is just a gimmick, a trick that drivers will quickly bypass once the novelty wears off. people want/expect much more very soon.
I think you underestimate the role of the driver. The human does much more than operate the vehicle. When a car breaks down the human is there to communicate and negotiate the situation. When a human stops traffic, as one did to me this morning because her puppy had gotten off leash, she is communicating with a human driver. Or in a construction zone where a workman needs to stop traffic to allow a truck to turn. An autodrive car, lacking that eye-to-eye communication we are so good at, could be very dangerous in such situations.
How will an autodrive tesla deal with a police check?
I'm not seeing anything that would suggest they are that close. Automatic, assisted driving on motorways, or very very well mapped out cities - sure. But autonomous where you can go to sleep on the back seat or indeed summon a car to you? Probably 50 if not 100 years away. And not only because of technical problems.
100 years ago, we were just getting around to inventing the light switch. I think you are seriously underestimating the pace at which this technology is advancing. It might take 10+ years for this technology to become mainstream in all likelihood. But 100 years? Not a chance.
The wright brothers made the first heavier than air human flight 113 years ago. Look what we can do now! First mobile phone call was 43 years ago and now we have such advanced tech that making calls is almost an afterthought. Once tech gets a grip its momentum is huge.
40 years ago we thought(well,some of us did) that accurate image recognition is a matter of few months of work, at most. It's 2016, and our most advanced image recognition software can't tell a zebra and a sofa in a zebra print apart. It's an insanely difficult problem. We can't even do voice recognition completely right. Automatic cars rely on that exact type of challenge - recognising patters accurately. I'm sure in certain settings they work exceedingly well. Spectacularly even. But to be safe for human transportation, they need to be 100% accurate. Not 90% or 95% accurate. They need to work in snow, rain, when a huge sinkhole appears in the ground or an infant walks into the road. They need to consider legal, moral and ethical implications lest they be allowed on the road - is it allowed to hit a pedestrian to save 4 passengers? Is it allowed to hit 4 pedestrians to save 1 passenger?
Can you at any time take manual control? If yes, how do you regulate insurance? If no, how do you tell the car that you want it in that particular spot and not any other, be it your garage or a middle of a clover field?
Yes, there already is software that can drive safely on lit, dry roads, while there is a human behind a steering wheel tracking its every move. But I still believe we are at the very least a few decades away from fully autonomous vehicles.
I actually met plenty of people who say "oh, they just need to have less accidents than humans and we are good".
No, that's absolutely not true. For the most recent example - someone made a raspberry Pi controlled insulin pump. Insulin is actually incredibly dangerous to humans if you get the dose wrong, so making an insulin pump based on hardware that does not conform with highest safety standards is just not acceptable. You know what the person behind it said when it was pointed out to them? That it doesn't matter, because raspberry Pi is still going to kill less people than the number of those who die through incorrect injections due to tiredness or simple mistakes. That's absolutely incorrect - even if such machine lowered the overall number of deaths due to incorrect insulin injections, no one would ever allow it on any market ever. No company would ever get its way out of "poor hardware choice lead to death of Mr. Smith" by saying "hey, but actually, our machine kills less people than would die naturally due to similar causes each year, so you can't hold us accountable, right??".
To me, it's the same with automatic cars - they cannot merely "have less accidents than humans". They need to have 0 accidents or they won't be acceptable. That's why the bar is high. If you are in a situation where a choice is between hitting a pedestrian or running under a semi and possibly killing everyone in your car, no one is going to blame you for doing either - our primitive brains probably are going to go with whatever seems most logical at the moment, you can blame anything on adrenaline. Computers don't have that luxury. They need to make a calculated choice - and then whoever makes them(the computers) has to live with that choice. The computer chose to hit the pedestrian - now the company who wrote its code is being sued for millions - no matter how they frame it, that's not a situation anyone wants to be in. Of course I'm going off into theoreticals here, since we don't actually have this problem yet. But I am sure it will become an actual problem and it will need to be solved one way or another before widespread adoption.
> But autonomous where you can go to sleep on the back seat or indeed summon a car to you? Probably 50 if not 100 years away. And not only because of technical problems.
This capability literally exists today, there's no way it takes 50 years to make it to market.
The capability we have today is working on fully mapped, well lit, dry roads. No one has an automatic car that could navigate in heavy snow/rain or unmapped terrain. There's very few ideas how to tackle it, unless we equip every road in some sort of trackers. I fully believe that in short time we will have all sorts of very clever cruise control systems which are essentially like Tesla Autopilot, maybe a bit more versatile. But fully autonomous vehicles? I stand by my statement that they are very far away for commercial distribution, unless you plan on selling them in Florida only, or only for use on private land.
I know, I actually read that article :-)
It says that the car could navigate the snow-covered road based on the data it collected from the same street when it wasn't covered in snow. So.....I get the idea, but that's not really "working in snow".
You call the car. It leaves the garage and starts heading to you. Then something comes up, something like a construction zone or other odd situation. Or perhaps the car breaks down. So the car is now stopped and alone. You're sitting by the road five miles away and the car is parked blocking traffic somewhere. Who comes to help? AAA? Wil it fight back if someone tries to tow it away? Or push it onto the shoulder to free up the lane? I doubt the general public would have much sympathy for the guy who's tesla is blocking an intersection because he didn't want to pay for parking and/or bother walking to the car himself.
I cannot see drivers using this garage/parking trick very often. It's just too slow.