Tesla Motors Aiming To Build Self-Driving Car Within 3 Years, Elon Musk Says
Originally published on CleanTechnica sister site Ecopreneurist.
Tesla Motors is currently aiming to have a self-driving, autonomous car built within the next three years, according to the famous Tesla CEO Elon Musk. If that goal is met, that means that Tesla will beat its rivals — Mercedes-Benz and Nissan — to the market by at least a couple of years.
Self-driving cars, also referred to as “autonomous cars,” have been in the news quite a lot over the past couple of years, but no firm dates have been put down with regard to when they will be available on the consumer market. However, with the recent announcement, and Tesla’s track record for that matter, it looks like that may now be changing. Elon Musk made the somewhat surprising announcement during an interview with the British Financial Times.
According to Musk, the new technology will be developed completely within house — not via a partnership with Google. Google has already, independently, developed its own self-driving technology, which is currently still in the process of being tested. What this means, as Stephen Edelstein of Green Car Reports makes note, is that this technology is being developed with “the same mentality that guided (the) development of the Model S.” A good sign no doubt.
With the three-year goal in mind, Tesla recently posted a job listing on its website for an “Advanced Driver Assistance Systems Controls Engineer” — essentially, just the person who could “spearhead Tesla’s autonomous car efforts.”
Intense effort underway at Tesla to develop a practical autopilot system for Model S
— Elon Musk (@elonmusk) September 18, 2013
Engineers interested in working on autonomous driving, pls email autopilot@teslamotors.com. Team will report directly to me.
— Elon Musk (@elonmusk) September 18, 2013
Approach is 360 deg flush mounted tiny cameras + radar (prob not lidar). Lot of software & hardware level image processing.
— Elon Musk (@elonmusk) September 18, 2013
More on the technology via Green Car Reports:
Musk refers to Tesla’s self-driving technology as an “autopilot,” calling a fully-autonomous car a “bridge too far” in the Financial Times interview. Drivers will reportedly be able to switch the system on and off, just like an airplane’s autopilot.
This could potentially allow Tesla to sidestep the many legal issues surrounding self-driving cars. So far, California and Nevada have legalized testing of these cars on public roads, but their status in other states is ambiguous at best.
In the race to put a self-driving car into production, Tesla’s main automotive rivals will be Mercedes and Nissan, both of which have pledged to put autonomous vehicles into production by 2020. Mercedes took its fully-autonomous S-Class prototype on a 60-mile trip in Germany last month. Nissan showed a prototype autonomous car based on the Leaf at its Nissan 360 event in Arizona.
With regard to Tesla’s competitors in the self-driving/autonomous car field, be sure to check out CleanTechnica‘s exclusive firsthand experience with Nissan’s completely autonomous version of the LEAF.
Have a tip for CleanTechnica? Want to advertise? Want to suggest a guest for our CleanTech Talk podcast? Contact us here.
CleanTechnica Holiday Wish Book
Our Latest EVObsession Video
CleanTechnica uses affiliate links. See our policy here.
Why would a normal Joe ever want such a thing? To me it is just a bunch of gimmicks waiting to go wrong. Horribly wrong.
Can anybody justify why Tesla is doing this. There must be an upside to wasting money on something not needed and risking the company in an endless quagmire of litigation…
One reason Tesla is working on self-driving is because many other car companies are. Gotta keep up and, if you’re Tesla, you probably want to stay ahead.
Why would a normal Joe want a self-driving car? Well, does Joe ever spend time in bumper to bumper commute traffic? Does Joe ever drive miles on the freeway spending his time watching out for other vehicles when watching the view or playing Angry Birds would be more interesting? When Joe is making a long drive would he rather take his nap with him or the car doing the steering?
Keeping up with the Jone’s might be it. But it’s like keeping up with the number of nuclear missiles another nation might have. I see this as a distraction from making a great car.
I’m too old to trust a self driving car. And I would not want to be required to legally watch the car in the drivers seat to make sure it does not get out of hand. I suspect once these cars become reality people will quickly object to this and when accidents do occur they will blame it all on the auto. Not the fact they were playing angry birds instead of parenting the car as legally required.
Driving in stop and go traffic? Maybe. This must be what Musk is thinking about. Remember how he rails against stop-and-go traffic staus on the 405. Says it takes him longer to commute on the 405 than it does to fly from LA to SF. (Perhaps he doesn’t have to wait hours and be searched by TSA clods if he has a private jet and hangers in great locations.)
You older than me?
I’m so looking forward to self-driving cars because I can feel my reflexes slowing and having to drive myself to town would cause me to have to move to town earlier, which I do not want to do.
And I’m finding it necessary to stop and take a break/nap on long trips, something that I didn’t have to do when younger. It’s not the hours of staying awake, it’s the hours of driving the car.
Elon hates the commute traffic. In a self-driving car he could get some work done.
People may blame any accidents that happen on the car, but the car’s “black box” will tell who made the decision to cross a double yellow line.
lol, NOBODY is older than you Bob.
If it makes you feel better I’ve always had to take naps. It is not an age thing. You are just wiser and know your limits. Something younger people often learn the hard way. And then if you are not accustomed to driving long hours you must take more breaks until you reacquire your skills. So don’t worry, you’ll never have to move back to the city. 🙂
Black boxes don’t record enough. I imagine we will have to go the Russian route. That is self driving cars will have to visually record everything from all angles. Not to protect ourselves so much from bad police but as “bug reports” back to Tesla.
My driving pattern hasn’t changed over the last 20 years.
My stamina is less and my reflexes are slower.
Black boxes will record who was in charge right before and at the time of the accident.
You are severely underestimating the ability of computers.
In 2012 there were 34,000 deaths in automobile accidents in the U.S. alone.
I would choose a computer that was meticulously designed and tested for piloting a vehicle versus people who can be distracted by vibrating phones, bright lights and/or a chick jogging in a sports bra any day
What most drivers are going to experience first is cars with collision avoidance systems. They’re already being installed in some models.
After having their car alert them a few times to dangers they hadn’t noticed, and perhaps apply emergency braking, people will start understanding that computers are more vigilant than humans.
The other night driving home I saw a deer about two feet away from my driver’s side door as I was passing it. I had been rounding a steep, uphill curve and the deer was on the inside of the curve. I did not see it until I was half way past it.
A self-driving car would have the sensors to pick up a deer standing in the dark. It would have been able to look all around at the same time rather than just attending to the outside edge of the road as I was doing.
how old are you? too old to trust technology but not old enough to realize that there are millions of people on the road with horrid driving ability. to old to realize that someone with a 2 hour commute to work could literally save 500 hours a week by instead of driving doing what ever they wanted in the car?
I am looking forward to self driving cars on account of how they will go horribly wrong much less often than people go horribly wrong. I’m not saying that a self driving car will never make an error, I’m just saying they will make a lot less than errors than humans and so be a lot safer.
I have reservations.
Driving is very complicated. I’d say more complicated than an operating system. Yet all operating systems despite being in some cases decades old have serious problems. Any of those serious problems are enough to cause fatalities if driving a vehicle. People will probably be required to monitor the computer driving the car. These people will be lulled into complacency until it is too late and they die. Or kill others. I like the idea but I don’t think it will be so smooth.
Look at the autopilot on airplanes. Though they don’t have to make many decisions they still can’t handle much outside of their normal parameters.
Look at your android smart phones. I find flaws in them on an almost daily basis. Even though I have the best hardware and latest OS.
Software is buggy. Having it lull a driver into complacency scares me. We’ll see if it is really safer. I bet there will be a huge backlash against these automated cars if they really do drive you and not just do something very minor.
Driving is very complicated. That’s why humans smack their cars into things so often. Computers can definitely be buggy, but, and this is the point, they can be less buggy than humans. If you’ve been on modern life support, flown in a modern jet plane, been on a ship at night, or worked with an industrial welding robot, then you have entrusted your life to computer systems without being killed by a bug. Obviously any production model self driving car will need extensive and expensive testing, but that will be done. My grandma is getting too old to be driving and it would be nice if she could get a self driving car to get around in and I am 100% certain that she would be safer in one of Google’s current self driving cars than controlling one herself.
That’s the theory. I think it will be a very bumpy ride. We will just have to wait and see. Hopefully we will all live through it.
Um, predator drones? They already exist, are completely autonomous, and can fly circles around most human pilots. They aren’t smart, it’s all scripted code that’s saying, “If you see this, don’t hit it by moving this flap this way…” etc, but it works.
You’re comparing two different things it seems to me. I would change your statement to “CHEAP software is buggy.”, then I’d agree with it. Microsoft Windows, OS X, Android, iOS, all of these are software designed to go on a smaller personal device that almost never costs more than $2000. There’s a big difference between a $400 phone’s software, and a $100,000 self driving car’s software. AND of course a $500,000,000, or whatever it may be, predator drone’s software. This type of fail safe, perfected software is out there it’s just not cheap. Putting this high grade of software into a consumer car without losing money is the real challenge I’d say.
So in short, the software that is as safe as you want it exists, don’t be worried about that. And to pass safety regulations and laws, the software that goes into these cars will have to at least match the reliability and safety of the high end software.
You really shouldn’t have to worry about your own safety if you go out to buy one of these cars.
However to get the cars to that standard that they must pass, I’m guessing it’ll take a little longer than the dates that these companies are optimistically looking forward to.
I’ve never seen good software at any level. No matter what the price. It’s an oxymoron.
Hence we get apple iPhones with fingerprint scan protection that is hacked within days. Not to mention all the other bugs that make it perfect for a thief. (I just listened to the “security now” podcast that gleefully tells you about all the latest apple exploits for about an hour.)
On the other end where the software is very expensive we get rockets that fly off course because part of it was designed with the metric system and the other part with some prehistoric set of goofy measurements. So much for six sigma concepts.
Everything needs to be tested. Many many times in many conditions. How do sensors work when covered with frozen rain? With mud. When the frame they rest on has been damaged and now they point at a different angle. As they age. What redundant backup systems do they have. How do they report malfunctions and get fixed? When that car at an intersection runs a light, bounces off somebody, and will now hit you in less than a second while you are sandwiched between cars? The scenarios are mind boggling and it will be a constantly evolving field for decades. Maybe 80% will be solved in 20% of the time. However I think my driving is better than anything they will come up with in the next 40 years.
“Microsoft Windows, OS X, Android, iOS, all of these are software
designed to go on a smaller personal device that almost never costs more
than $2000. There’s a big difference between a $400 phone’s software,
and a $100,000 self driving car’s software.”
I don’t that argument holds. Yes, smaller personal devices cost less individually but there are many millions of them in use. Even a small amount per unit provides an immense amount of money for writing tight code.
Where vehicle software is likely more trustworthy is that 1) the functions are far fewer than your ordinary laptop (stay within the speed limit, stay in the proper lane, don’t hit anything).
And second parties won’t be writing programs that work with the car’s computer. Most computer software problems seem to be caused by second party companies, not the operating system.
how do you think they are going to “lul drivers into complacency”/ the only way i can think of that people would be complacent with driver less cars is if they are statistically saving millions of lives a year(which if they were successful they WOULD be doing). So basically after saving millions of lives and “luling people into complacency” people would trust them too much and possibly you or someone else may die?
Well, I can certainly say why a lot of people might be interested: there are a lot of bad drivers on the road. Personally, I’d rather have the car driving itself than have to worry about the driver being drunk, or texting, or eating while driving, etc.
But you’re right, Tesla really should be really, really, careful on how they do this as they open themselves up to litigation risk if it ends up crashing. Still, they’re not the first to make self-driving cars. Google has made self driving cars that have already driven half a million miles without a single accident.
Tesla will not be first to market with a self driving car. Volvo is planning on selling one next year.
I don’t think reality will match most peoples expectations. These self-driving cars will probably be the equivalent of upgraded cruise controls. Tesla has enough software problems just trying to keep up with the promises and expectations of their existing customers using their 17″ screen.
Fair point. It’s one of those things that seems like a big leap forward at this point. But perhaps after the fact we’ll wonder how we ever got along without it. Here’s another angle on self driving cars serving as a mobile office: http://suitesearch.com/next-office-may-self-driving-car/
I don’t think the issue is technology because Google has driven their automonous cars for over 300,000 miles without accidents related to the auto pilot. (see http://en.wikipedia.org/wiki/Google_driverless_car ).
The issue will be price. Each of Google’s cars is equipped with $150,000 worth of equipment. This price will have to drop to at least $20,000 to get high end car buyers interested.
Another huge market will be for utility vehicles on controlled routes where the driver might be eliminated. That will require a new set of laws, but if the technology can be trusted (as it is on jumbo jets which run 99% on autopilot) then the savings in driver costs will make automonous cars very, very attractive.
Can’t wait until I finish this GI Bill gig (id be a fool not to) so I can launch my revolutionary companies.