What fatal car crashes?
The only incident I'm aware of was the one where someone drove into a semi crossing a road illegally, using a Tesla on "autopilot". Have there been others? (Genuine question).
Ford is investing $1bn in Argo AI, a startup cofounded by engineers who resigned from Google and Uber’s autonomous car projects. Competition to develop the world’s first completely autonomous car is fierce. Companies like Audi, Tesla, Baidu, Nissan, and Toyota are racing to get their brand-spanking-new technologies on the road …
I can see the day when "Self Driving Cars" are common place around towns and on motorways, but it will be a very, very long time before they'll be able to cope with the country A/B roads around here (Somerset/Dorset/Devon).
True enough, only when the only cars on the road are autonomous and they all know each others whereabouts. This is where I think Tesla's idea will be best for the short-term - autnomous but the driver still has to remain alert and with hands on wheel and in the case of cruising around bends on country lanes, foot hovering over brake!
It's not just other cars, but all the hazards you never see in towns and cities.
Downed branches after a storm, flood/standing water forcing traffic onto your side of the road. Live (and dead) animals lying in the road. Suicidal sheep who always run the wrong way. Suicidal cyclists wearing dark clothing and no lights at night. Farmer Giles with his ancient tractor spreading muck over the road instead of the field...
The list is endless and most human drivers fail to cope with these situations. So "AI" cars don't stand a chance for the forseeable future.
The list is endless and most human drivers fail to cope with these situations. So "AI" cars don't stand a chance for the forseeable future
This is a slightly strange argument. Most journeys don't take place on those roads nd nobody so far is proposing exclusively self-driving cars in the near future. However, given that human IR vision and LIDAR are nonexistent, having those subsystems as a backup on country roads is going to be a big help. IR will pick up live animals and cyclists in the road before eyesight. What has to be determined is how the system will work with the human driver, and even if it's only braking application and a dashboard sign "Unidentified object in road" that's going to avoid a few crashes. The argument that because something isn't perfect it shouldn't be done would eliminate human drivers.
It's not going to happen for a few more years.
There's the issue of computing power and having enough sensors.
Its one thing to drive in Phoenix/Scottsdale where the roads are nice and easy to figure out your lane and position, but on roads in Chicago or other large cities? LOL... good luck.
In terms of GPS positioning... its not accurate enough.
I can see absolutely nothing wrong with spending $1billion on a me too effort for a technology with some fundamental conceptual flaws, and even better spending that money with people who have so far failed to deliver this same thing for their previous employer, have no IP and in the event they do produce some IP will spend the rest of their lives fighting over it with their previous employer.
The vast amount of money just shows it has to be tech bubble induced madness affecting people I would normally think of as sensible.
The only viable move left is to jump on the bandwagon. I'll be happy with a mid- 9 digit funding package. I might even deliver something, but make no guarantees it'll do anything except slowly blink a single LED.
Should we be worried if the LED starts blinking fast or something?