BMW, Intel, Mobileye Partner On Self-Driving Cars, 'Turning Point For Automotive Industry': Reports (bloomberg.com) 59
BMW, Intel, and Mobileye NV are working to develop autonomous-car technology, reports Bloomberg, citing multiple sources. Senior executives from each company will hold an event on Friday to discuss the driverless-vehicle initiative, the report adds. From the article:Jerusalem-based Mobileye has been an early leader in providing cameras, software and other components that allow vehicles to see the world around them. BMW has been a client of Mobileye, along with General Motors Co. and Tesla Motors Inc. As automakers and their suppliers race to create systems to replace human drivers, most companies are betting on some form of artificial intelligence, which requires powerful processing.Reuters, citing one source, reports the same thing. The announcement will be a "turning point for the automotive industry," Amnon Shashua, the chairman and co-founder of Mobileye.
Re: (Score:2)
I think you meant to write Anonymous Canadian
Re: (Score:2)
How much were you paid to spread this silliness? Because everything you said demonstrates either immense ignorance or a willingness to troll on line for money.
Re: (Score:2)
Re: (Score:2)
The dumb fat Americans will never give up driving their cars and the supposed freedom.
Bollocks.
The change will be OVERNIGHT.
It's game over as soon as the dumb fat Americans figure out they can watch pron or play Candy Crush on the way to work. Dealers will probably have to ban trade-ins because they'll be unsaleable.
Different company (Score:2)
Nope.
Kinect is based on a technology by PrimeSense which couple 2 cameras.
- one regular one for colors
- one infrared one for depth
Mobileye is a manufacturer of monoscopic cameras.
Their software works on pure flat image that (apparently) where depth is infered from a projection.
Re: (Score:2)
Or does Google have smarter engineers. Seems to me Google couldn't get around a sand bag without running into a bus just a few months ago. Were they able to enter all the perceived obstacles that they thought of since then? Because surely if they had thought of everything they would have thought of a sandbag. Are they could to skip testing on ice altogether? If it can make it down a sunny Califor
Re: (Score:2)
Re: (Score:2)
Re: (Score:1)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Really? Someone is going to make a 'this product may drive through the side of a house if a cat runs in front of it' clause?
There are lots that are worse than that, and if you think that's how a self-driving car works, you are too dumb to use a computer. Please put it back in the box and return it where you got it.
Re: (Score:2)
So...you'll be turning in your cell phone later today then?
Re: (Score:2)
Why the fuck would they use orbital cameras? They just use standard cameras mounted on police cars and on the side of the road, read in your license plate and store your location in a database. If the man wants to know where you've been, he'll know.
https://www.aclu.org/feature/y... [aclu.org]
Re: (Score:2)
Boot loader (Score:2)
2. Check GPS and vision systems.
3. Check occupants for warrants and debt payments; lock out manual overrides, door handles, and re-route final destination as necessary.
Re: (Score:1)
7. Microsoft Interstate Explorer installed. Estimated time to your destination:
47 minutes...
8 hours and 10 minutes...
11 seconds...
1 hour 5 minutes...
9 days 4 hours...
47 days 14 hours 55 minutes...
1 second...
1 second...
1 second...
1 second...
Re: (Score:2)
It's also a turning point for truck drivers. (Score:2)
The first "jobs" casualty will be truck drivers.
Fleets of long-haul cargo trucks, going between a limited number of specialized terminals or loading docks, on major roadways and well-maintained commercial access roads, under lowered speed limits and other tightly-constrained legal behavior requirements, without tight constraints on transit time, are easier to automate than herds of passenger cars, and have an economy-of-scale at the owner level.
Imagine a truck on a roadway, going from one loading dock to an
Re: (Score:2)
No, you're wrong & he's right. Autonomous commercial vehicles will be among the soonest to arrive but I think companies will be careful as to which jobs & routes they get.
"it isnt going to happen fo 30 years at least." - nope, disagree. In 30 yrs time, there'll be only very few that *won't* be autonomous
Re: (Score:1)
Until the truck has a blowout (Which happens regularly), a mechanical fault happens (also happens regualrly), a senosor fucks up (happens even more), there' rain, fog, snow or dirt (Basically daily) to start.
Got a solution to the above? Let me answer for you - no you dont. Autonomus trucks are a bigger fucking pipedream than autonomous cars - it isnt going to happen fo 30 years at least.
Try 5 years, highly likely. Blowouts will be less common, because, get ready for it, automated service on tires and brakes will generally catch the problems. Also, not having drivers fall asleep and skim jersey walls or wake up driving in the median and rolling over a bunch of tire damaging junk will also reduce potential tire damage. Lower speeds will also lower tire blow out risk and brake issues. In fact, because the autonomous vehicle can't be hurried, isn't impatient, and won't fall asleep, truck accid
Re: (Score:1)
If there's no blowout, they're right.
It means they were lucky.
Re: (Score:2)
Got a solution ...?
I don't have to have a solution. The army of engineers designing this, and executives designing the deployment logistics, have to come up with a solution. But it only needs to be no worse than "a driver notices it and does something effective about it".
Having said that, I DO have some experience in the auto industry, industrial automation, remote sensing, and computer science. So let me supply a few possible solutions, just off the top of my head:
a blowout (Which happens regularly)
All
Solutions (Score:2)
Got a solution to the above? Let me answer for you - no you dont.
- One already possible solution is to have the first truck in a convoy being still maned (a la Tesla: autopilot but with human supervisor) and the human can still be in charge of the whole convoy if anything happens.
- Other possible solution would be to subcontracts handling of such problems to 3rd party companies operated by actual humans.
If the robotruck notice anything bizare, it can summon an operator to come check at it.
- More sensors, more redundancy, higher level of controls.
such robotruck could dete
Casualty already done (Score:2)
Imagine a truck on a roadway, going from one loading dock to another, as an elevator car, or a "people-mover" style unmanned train.
Some countries (hello switzerland) have already replaced huge portions of the trucks traffic with actual trains.
(Well manned train. With an operator paying close attention if anything goes wrong with the semi-automated high-speed train)
No Thanks (Score:2)
I will continue to drive my own car thank you very much. I spend the entire rest of the day staring at some damn screen or another, and the last thing I need is a self driving car so I can read Slashdot comments on my way home from the office.
Wait.... Okay maybe...
Re: (Score:2)
Fuck that. When we have self-driving cars I'm planning on picking up the highest-paying job I can find an hour away so I can game my ass off before and after work every day without any distractions from the wife and kids.
Re: (Score:2)
When we have self-driving cars I'm planning on picking up the highest-paying job I can find an hour away so I can game my ass off before and after work every day without any distractions from the wife and kids. :)
You obviously have a much better laptop and a much better data plan than I do. ;)
Re: (Score:2)
Try it now; they won't miss you and neither will we.
Ouch.. That was harsh... Not that I'm arguing with you...
Re: (Score:2)
I drive a BMW because I enjoy driving. I can't imagine owning one for any other reason.
If I can't drive it then might as well buy a Prius.
Why do we always thing AI is resource hungry? (Score:4, Interesting)
The best AI I've seen is stuff that uses techniques which are easy on resources.
The holy grail of AI is something that can run with limited resources and give you reasonably good results and from what I've seen that's not beyond the state of the art today. In fact, I personally think the advancement in AI will only take place once we forget this foolish notion that we can field brut force algorithms for stuff like driving cars where the range of 'acceptable' solutions is pretty wide given a 6 foot wide car going down a 10 foot wide lane. We don't even do that for trains yet, and you don't have to manage the steering wheel, just the throttle and brakes on those things.
Tell me, how do YOU drive a car? Do you have a better than HD set of cameras scanning the area around the car to 1/4" tolerances? Absolutely not. Why do we think we need to brute force this problem in order to do it on a computer? Something tells me we have over engineered this if Intel thinks they will be selling more processing power by being involved.
I get the devil is in the details, but watching my 16 year old learn how to drive does not tell me this is problem takes huge amounts of processing power...
Re: (Score:2)
I get the devil is in the details
Exactly.
I'm betting that the first 80% of the problems of self driving cars are already solved. But solving the last 20% will be a bitch.
And what will become of motorcycles? Will they be banned?
AI isn't science fiction any more. (Score:4, Interesting)
Now weak AI isn't just coming. It has arrived [businesswire.com]. And Moore's law was supposed to have stopped years ago but supercomputers [top500.org] and video cards [surrey.ac.uk] are still on a logarithmic slope for performance and price. The human brain is estimated to calculate between 100 petaflops and 1 exaflop [toytowngermany.com]. I know that's not a good metric but for this purpose it suffices. But as performance keeps doubling and doubling it becomes more evident that even the highest estimates are a question of a few more doubling periods. And the highest estimates assume direct one-to-one simulation of each neuron. Consider how many neurons are used for breathing, processing vision, and other things that either aren't needed in a machine or have already been done at a much lower computational cost on silicon.
It's true we don't know everything about how the human brain works. But recent progress is undeniable in terms of success stories. Jeopardy. Go. Commodity trading. Corporate resource balancing. Piloting [bbc.com]. To keep shouting that strong AI is impossible is to only betray one's own insecurity. You are not special. Your brain doesn't run on quantum magic. You have no soul. Fucking deal with it.
Re: (Score:2)
I agree with you that current results in AI are very impressive. However they are all achieved using supervised learning. Even reinforcement learning is a form of supervision. Strong AI requires mixes of semi-supervised and unsupervised learning, i.e. the system does not get all the clues before being able to generalize, and is able to learn by itself. We are not there yet.