Sep 17, 2013
surfside Having been interested in driverless cars for a while now (my interest began about the same time that I started learning about Tesla), I've been intrigued to hear Elon talk in the past about Tesla and the potential for driverless cars.
Click here for a new interview with the Financial Times where Elon indicates that he believes Tesla will have Auto Pilot technology capable of autonomously driving the car for 90% of total miles driven within three years. Elon has mentioned before that he believes that Google has it wrong with regard to the technology required to create driverless cars (essentially saying that their technology is too expensive); it will certainly be interesting to see how Elon intends to incorporate this "Auto Pilot" technology into its cars.
Any thoughts on whether this story will have an impact on the stock tomorrow? It certainly feels like a meaningful announcement to me, but I may be biased given my interest in driverless cars. Interested to hear other folks' thoughts.
Surfside�
Sep 17, 2013
borat99 In case people hit a paywall, here is the article.
�
Sep 17, 2013
Cameron Very interesting! To me this just says that Tesla is working on even bigger projects then most people realize.
Unfortunately, I don't know if the world is quite ready for driverless cars, and in 3 years time is especially shocking. I wouldn't be surprised if the stock reacts negatively at first, but once it's discussed and maybe more details come out about what exactly Elon is planning, we could see it rally.�
Sep 17, 2013
borat99 It'll probably a shock for me at first, but I'll be ready for it if they can show me that's safe. On the bright side, Google has already proven that it can be safe. Google's Self-Driving Cars: 300,000 Miles Logged, Not a Single Accident Under Computer Control - Rebecca J. Rosen - The Atlantic�
Sep 17, 2013
Convert2013 I do hope Tesla is doing this in close coordination with the NHTSA from the get-go, to avoid safety related compliance surprises too late in the game.�
Sep 17, 2013
ItsNotAboutTheMoney Not driverless. It's simply never going to happen except in closed systems. Auto-pilot, yes.�
Sep 17, 2013
surfside For some additional context, here is a link to a series of Forbes articles about driverless cars that is pretty interesting and provides insights into just how far reaching driverless cars could eventually be.
Surfside�
Sep 17, 2013
blakegallagher Elon had mentioned before that the tech might be ready for Gen 3. I guess now that he declared 3 years he is shooting to get it on the Model E. I think it is safe to say the Model S and X will get many more safety features relatively soon.�
Sep 17, 2013
callmesam
Chunka Mui has been incredibly prescient when it comes to autopilot.
Elon once called his system a Super Smart Active Safety System.
Wonder if he'll call it SSASSYS or sassies.�
Sep 18, 2013
Causalien If you look at the job posting for driverless. It seems to tatget Sebastien Thrun of Google. A doctorate level candidate.�
Sep 18, 2013
T B I'm a bit concerned about having about one second warning time when the autopilot decides the traffic situation is too complex for it (the remaining 10%) and I have to take over. In a split second I have to (put away the phone I was tinkering with) reach for the wheel, examine the traffic situation around the car, and react correctly to it.
But somehow I trust Mr Musk and his engineers they will solve it.�
Sep 18, 2013
marvinat0rz You won't be allowed to play with your phone when your car is on autopilot - I'm certain you'll have liability even when the autopilot is active. So you have to pay attention at all times, but you won't have to actively drive the car. Given that you do this, it will only be a question of grabbing the wheel and taking over. Until we get to a situation of driverless cars sometime in the semi-distant future, the driver will always have the legal responsibility and always have the attention on the road.
Airline pilots sometimes land the aircraft on autopilot, but you can be damn sure they're not playing Angry Birds on short final. Angry Birds might be played in cruise, when the weather is nice and there are no expected problems. If things gradually start to look more challenging, the iPad has to go. This is in fact a flawed analogy; traffic is a lot more hectic than the cruise portion of an airline flight. On an airplane, conditions will almost never change in less than 20 seconds. In a car, things can really change on a second-to-second basis.�
Sep 18, 2013
bonaire Autopilot cars on roadways full of non-Autopilot cars seems dangerous. I think autopilot works when all cars are working concurrently under autopilot.�
Sep 18, 2013
Mario Kadastik Yeah, but that has the chicken and egg problem
�
Sep 18, 2013
RationalOptimist I've taken a ride in one of Google's driverless cars. It's a truly mindshifting experience. It hurled itself around a track, wheels squealing, pinpoint accuracy. Far, far, far better driver than I could ever be. People are going to be truly shocked by this technology. Marketing it as 'autopilot' rather than driverless is genius, and will get acceptance much earlier. If Elon can get a version of this done to include in Gen3, look out world.�
Sep 18, 2013
bonaire How does it work in a dust-storm or heavy fog? What if James Bond throws out an oil slick and smoke bomb?�
Sep 18, 2013
LASpark I think for 2016 they are probably referring to active cruise control combined with lane keeping, giving you hands free freeway driving.�
Sep 18, 2013
Norbert Tesla enters race to build self-driving car - Yahoo Finance
Elon talking about Tesla having 90% self-driving cars in 3 years: I think implied is the intention to stay cutting edge also in car technology, not just electric drive trains, and that Tesla is a tech company. That may also be, partly, a response to recent marketing talk by the older auto companies about their electric car plans.
- - - Updated - - -
Trader (technicals) bullish about Tesla:
Trader: "Tesla is Apple, pre-iPhone" | Talking Numbers - Yahoo Finance�
Sep 18, 2013
gregincal I don't think the situations they are talking about are the emergency situations. The car should always be able to avoid hitting another car and can react better and faster than a human ever could. I would imagine something like merging traffic when there is a lane closure would be difficult, but you would have plenty of time to take over.
Anything too complicated for the computer is also going to require human drivers to slow down.�
Sep 18, 2013
Robert.Boston I find it wryly amusing that Elon is talking about auto-pilot when the Model S doesn't even have parking sensors!�
Sep 18, 2013
Mario Kadastik Erm, want to be a bit more specific? There are parking sensors in the options list and I'm getting them�
Sep 18, 2013
brianman And this is partly why some of us don't want autopilot cars:
- - - Updated - - -
Yes it does...�
Sep 18, 2013
Kipernicus I think he meant "MY Model S doesn't even have parking sensors!"
I for one am looking forward to the day when I can get onto highway 5 and take a nap while the car takes me to the next supercharger. Or read a book while slogging through the 405. I'm very encouraged with what Mercedes has on the new S class and Volvo's road-train demos.�
Sep 18, 2013
brianman Renting a tow truck can get you that today!�
Sep 18, 2013
gregincal It has parking sensors, but it doesn't have adaptive cruise control, which is the first step towards autopilot (maintaining the proper distance behind the car in front of you).�
Sep 18, 2013
brianstorms I don't relish the idea of autopilot. I don't want to relinquish that much control to the car or to the company. And even if it is 99.99% reliable, if I drive 12000 miles a year, that means it works properly only 11,998.8 miles out of those 12000. Not good enough. I may be an early adopter with the Model S, but I am a staunch Luddite when it comes to autopilot or self-driving cars.�
Sep 18, 2013
Lloyd Tess..... Wake me when we arrive in Las Vegas at a good place for breakfast.
"I will Lloyd. What would you prefer for Breakfast? I show a great buffet with good ratings for Eggs Bennedict. I know you like that!"
That will do fine Tess.
"Good night Lloyd. Would you like me to play some music while you go to sleep?"
Thank you Tess....... Music trails........�
Sep 18, 2013
mershaw2001 I disagree with you, brian. I don't care to use the autopilot myself, but I am looking forward to the day that others will have autopilot. I'm sure there are plenty of people out there who have limited sight and limited mobility and yet are forced to take command of the car in order to drive. Imagine if you took all those drivers who are slow to respond or slow to react and you made them all predictable autopilot drivers? You and I could still cruise the highway, and do so in even more safety, until we decide we aren't confident enough in our driving ability.
Imagine if you could take all the drunk drivers off the roads, too.�
Sep 18, 2013
brianstorms Technology innovations all come at a price. ICEs brought mobility to millions, but covered the earth in pavement and snarled traffic, and filled the air with pollution. Cellphones brought communications to millions, but as we have learned, at the price of every shred of personal privacy. Autopilots and self-driving cars, like the existing Model S and devices like the Kindle, will undoubtedly "phone home," track one's every movement, which again, most will argue the old chestnut, "I've nothing to hide" which is not the point. Autopilot/self-driving cars are one more removal of privacy and individual control. There will be good aspects, like a bad guy stealing car, well, call the authorities and, presto, the car gets disabled remotely. But there will be bad aspects too, like authorities unlawfully deciding to detain you and utilizing a "back-door" that the car company had to install allowing the govt to override the car's ability to move forward, or, worse, override the destination you programmed into the car, and instead it takes you to, oh, some undisclosed location.
Sure, claim tinfoil hat rebuttals all you want. But we live in the post-Snowden era, and all bets are off, including with cars, in my opinion.�
Sep 18, 2013
brianman "Lloyd?"
"Lloyd, are you there?"
"Crap. I'm never going to hear the end of this from the neighbor's car."
"Hello, Tesla roadside? Yes, this is Lloyd's car. We got hit by a BMW."
"Yes, yes. We're fine. Lloyd's a bit sleepy though, and that BMW was shredded."
- - - Updated - - -
Shall we talk about low-flow toilets? And forced transitions to different light bulb technologies? Don't even get me started on healthcare and religion. Government loves to do things "for" you.�
Sep 18, 2013
NuclearPowered Elon Tweet: Intense effort underway at Tesla to develop a practical autopilot system for model s.�
Sep 18, 2013
PeterW Tweet from Elon
Intense effort underway at Tesla to develop a practical autopilot system for Model S (@elonmusk)�
Sep 18, 2013
sleepyhead Another tweet from Elon:
Engineers interested in working on autonomous driving, pls email [email�protected]. Team will report directly to me. (@elonmusk)�
Sep 18, 2013
MartinAustin Elon Musk has put out three tweets in the last hour, all about autonomous driving. He must really have a bee under his bonnet about it!
The latest one appears to indicate they have narrowed their investigations down to a solution. So it's not just "trying to develop it."�
Sep 18, 2013
Tasmanian Devil As much as I respect Mr.Musk and what he has achieved: The idea of autopiloting is absolute rubbish. People WANT to drive their cars. The freedom a car provides is one of
the great feats of this invention. I can understand that autopiloting is considered. But it is a major mistake to believe people want it and it is a major design error to pursuit
this goal NOW with priority. But Mr.Musk isn't the only one...looks like a form of superflous design fever (Morbus Fisker) running through the auto industry right now.
Ask ANYONE around You: People want to control their own car. If You want to design a fully automatic car, the automobile itself has to be completely redesigned, like with
a bathroom and a bed, You know ? Why design an automatic vehicle and still forcing the driver to sit (uncomfortably) in the drivers seat all the time...
Strongly recommend against priorising this now for Tesla...�
Sep 18, 2013
qwk I'm sure that the "autopilot" will be an option that is user selectable. For those who want to drive, this option isn't going to stop them.�
Sep 18, 2013
Discoducky T-Dev, I agree, but then again I don't drink coffee and yet it's a crazy big industry. Some people will want this and governments will eventually mandate it for safety considerations. I don't understand how this fits in with 'sustainable transportation' and I really hope this doesn't take away from my new Model X and E that I want to buy.�
Sep 18, 2013
EnergyMax Totally agreed.
Except. When I make a 400 mile drive up the 5 to visit family. Driving at 65mph for >45min in a sitting is _not_ fun. I'd much rather have autopilot take over the controls.
In ordinary daily life, no autopilot for me.�
Sep 18, 2013
surfside Another interesting Tweet from Elon describing how they plan to implement the auto pilot:
"Approach is 360 deg flush mounted tiny cameras + radar (prob not lidar). Lots of software & hardware level image processing."
The plot thickens...�
Sep 18, 2013
ItsNotAboutTheMoney Huh? Reduce accidents and you reduce energy use. Steer perfectly and you reduce energy use. Avoid unnecessary braking and you reduce energy use. Avoid unnecessary acceleration and you reduce energy use. Park smoothly back into a spot and you reduce energy use. Precise computer control can and does increase efficiency. It's dumb computer control like basic cruise that doesn't.�
Sep 18, 2013
NuclearPowered If Tesla is hiring to do the development of the AP all in house, it makes me wonder if a potential partnership with Google went bad? Potenatial disagreements over the technology and such.�
Sep 18, 2013
brianman Removing the humans reduces energy use as well....�
Sep 18, 2013
surfside Out of curiosity, have you read the series of Forbes articles? I personally think when these features eventually become the norm, people will wonder how on earth they managed without a car with auto pilot.
Ask anyone that has a commute of over 20 minutes (which is a meaningful portion of America), and I think you would find that most people would love to be able to use auto pilot to get to and from work every day.
- - - Updated - - -
This wouldn't surprise me at all given Elon's previous comments about Google's approach to driverless cars.�
Sep 18, 2013
gregincal Yeah, just like the silly cruise control and automatic transmissions ideas. Boy those never became popular. :wink: I definitely agree with Musk that talking about driverless cars puts people off and seems like science fiction. But extending adaptive cruise control to also do the steering is something people can understand.�
Sep 18, 2013
Sandman14 I disagree Devil. Auto enthusiasts want to drive their car, and enjoy doing so. But a lot of people just want to get their mini van to school and the grocery store. Those people will use an autopiloted car. If/when it becomes safe and affordable.�
Sep 18, 2013
ggies07 I have a 50 min to an hour commute to work, so that's close to two hours a day in my car. I would LOVE to have this feature. No more stress on the way home.�
Sep 18, 2013
Jackl1956 I like the idea of collision avoidance features actively engaged while I'm driving.�
Sep 18, 2013
Convert2013 I think the Auto-pilot thing may get the market confused on priorities...Elon can discuss it but needs to keep it low-profile. I see 3 tweets from him today on the auto-pilot subject.
Focus should be on margins, new products, production, superchargers, etc... the basics first.�
Sep 18, 2013
vin5xxx Commuters are going to love autopilot. Enthusiasts will just switch it off. People who like to drink wine at dinner are going to love autopilot as well.
I think now that Hyperloop has been published Elon has some free time on his hands to head up the autopilot project. Affordable autopilot is going to be HUGE!!!!!!�
Sep 18, 2013
Sunnyday I think this automatic driving stuff is cool, and could improve safety, but as a shareholder I'd rather see their immediate energy focused on the supply chain, and specifically batteries. Would like to see an announcement of a megagigawhatever factory. Increased battery supply and lower battery cost are probably the two biggest issues standing in the way of great execution in the years ahead. A battery deal would probably send shares up 10-15% in minutes. Automatic driving is a nice to have, but batteries are a need to have.�
Sep 18, 2013
MartinAustin Totally agree!�
Sep 18, 2013
JRP3 Agree completely. Seems very much like a gimmick, and one that personally just doesn't excite me that much. On the other hand Tesla talking about it probably gives the stock a little bump.�
Sep 18, 2013
Theshadows Maybe I'm too geeky but auto driving cars are super amazing and how much time could be saved? Imagine being able to work or read while stuck in bumper to bumper stop and go traffic. That wasted hour spent driving could become productive. What about the possibilities of 10-20 years down the road, am auto drive lane like the HOV lanes now, only going 100mph.
If Tesla Motors nails this and patents it! Look out! The value of the company will go to Mars!�
Sep 18, 2013
JRP3 I can understand your point, but that type of driving simply is not part of my life, by design. I'd probably go postal :cursing:�
Sep 18, 2013
Theshadows Sometimes I have to drive 3 hours on the highway one way, work for 5, then drive home 3 hours. Sometimes the drive home includes a half hour nap because I'm drowsy, I would love to keep on heading home and take a quick nap, then after my snooze, pull out my laptop and do some work.
I love driving too and can understand what you are saying, but there are times I would love autopilot just like I love cruise control now. And to have the patent to this technology. That is extremely valuable. Think long haul truckers, UPS, USPS, and FedEx delivery drivers, they could prep their next delivery while the truck backs down the driveway. The list is endless of people that could use this technology to cut costs and increase productivity.
Ps: this should probably be in another thread, I went ot again.
- - - Updated - - -
Oh yea, and to be able to have two or three glasses of wine at dinner instead of one.�
Sep 18, 2013
JRP3 I had the same thought, but figured since you can still get a DWI if you're in your car and it's parked, autopilot probably wouldn't get you off the hook.�
Sep 18, 2013
brianman So is the list of lawsuits if anything goes wrong ever (or someone fakes a failure well).�
Sep 18, 2013
bareyb Yeah. People will start blaming every crash they get into on it, just like they were when Toyota had their "sudden acceleration" recall. All of a sudden, there was a rash of Toyota owners having the issue (even when they weren't). Unfortunately, I think we'll see self driving cars about the same time we are going to see flying cars. Neither will happen and for the same reasons. Too much potential for disaster. If there IS a self driving car, I don't see it being fully automatic and would probably require some form of interaction from the driver to make it work. I don't think we'll be able to take naps and read... Just my guess...�
Sep 18, 2013
brianman People are (well at least 1 reported case so far) already blaming the 17" to try to get out of DUI.�
Sep 18, 2013
bonnie Think about what driverless technology will do for an entire new market ... we will no longer need to take the car keys away from aging parents. Instead, we will be giving them keys to a Tesla with driverless technology. Anything we can do to allow people to age with dignity is a good thing ... and driverless technology is one of those things. And it's a double win, since it also happens to open up a new market.
To be clear, I'm not talking about a car that acts as a taxi (though that would be useful). I'm talking about a car that provides assistance to a driver who may not be as sharp as they once were, to allow them to safely navigate the roads without putting themselves or others at risk.
Sometimes it's not about what WE want. Sometimes it's about what others need.�
Sep 18, 2013
bareyb Great... that's all we need. I can also guarantee you that some group sponsored by the oil or car companies will try to get the entire GUI banned at some point saying it's too dangerous. What we'll end up with a is a bunch of "Lawyer screens" and a crippled GUI while every kid in the country can still have a fully functioning Cell Phone turned on, while they drive and text.�
Sep 18, 2013
blakegallagher
As I shareholder I am not focusing on full autopilot aspect of this just yet but I think it will be huge for sales when they incorporate all of the features (and more) that other high end luxury sedans have. I am sure before they incorporate autopilot we will get lane departure warning and adaptive cruise controls which will go a long way to increase demand and margins.
I would much prefer a battery announcement too but I am confident that Elon is thinking about it.
I would love some sort of Joint Venture with Panosonic to build a big battery factory as close to the Fremont factory as possible�
Sep 18, 2013
stevezzzz I'm in favor of an autopilot for long-distance highway driving, particularly when coupled with a comprehensive trip-planning application. I can tell you as a pilot that having an autopilot doesn't reduce your engagement, it just frees you up to consider the bigger picture and evaluate your progress against the flight plan. You can always turn it off and hand-fly the plane when you want.�
Sep 18, 2013
bareyb I'm looking forward to a (relatively) inexpensive battery swap program one day. Tesla must have something in the works as they've proven they can do it in under 90 seconds.�
Sep 19, 2013
Theshadows Googles track record so far is flawless as long as humans don't try to intervene.�
Sep 19, 2013
ggies07 That's one of the reasons Carlos Ghosn from Nissan is trying to develop the tech for their cars as well. It's a great positive example for our society. I'm not sure what I will be like when I'm 70-80, but if something like this would let me get out just as much as when I was younger, that would be great.�
Sep 19, 2013
Ludus I get the impression Elon has concluded that the GoogleX self driving car project didn't have much advantage and was too focused on just pushing the limits of the technology rather than developing commercial product. Like for a lot of other car companies the idea is that this is really series of technologies that can be introduced as separate features....some of which are just software upgrades. Instead of all or nothing, the car is either conventional or robotic, the view is that it just gradually gets more robotic features as Tesla gets more and more real world performance data streaming in.
i think this means that the cars would have a full sensor suite and be capable of driving without a passenger quite a bit before these features are switched on for ordinary users.
Hard core driving enthusiasts don't enjoy every driving task. I think a lot of people who love to drive down the coast highway for fun would switch on autopilot on I5. I'd hope one of the early true driverless modes would be an auto-valet enabling (at least in designated places) just getting out of your car and letting it go park itself or return from the lot to pick you up at the door when you phone it.
i don't think this is a major distraction, the Tesla platform is the best suited in existence for this technology because every aspect of operation is natively under software control already. It's just a new layer of software to act on it's own based on sensor input rather than waiting for driver inputs. Most IC cars are still controlled through rube goldberg mechanical linkages that aren't as easy to adapt to software control.�
Sep 19, 2013
dsm363 I would love this even of they only allowed it for highway cruising (disabled below 45mph maybe). Would make longer trips nicer.�
Sep 19, 2013
ItsNotAboutTheMoney Yes, the full sensor approach makes sense. Equip the car will lots of sensors with the intention of doing advanced work, and then gradually develop software that uses all of the inputs. Once there's enough to make it worthwhile start building the car with the sensors. That way, they can got the supplier resolved in one go.�
Sep 19, 2013
aronth5 Many in the world may not be but certainly anyone who views driving as an impediment to staying current with their social media will gravitate to it very quickly. For that alone it will be a huge win and safer for the rest of us with fewer distracted drivers. Everyday I see drivers looking down at their smart phones or tablets while they drive and it drives me crazy.�
Sep 20, 2013
Causalien I believe that Autopilot with 5 cameras covering 360 degrees is possible, the only roadblock is figuring out the cost structure that will make this work. One high frame rate camera in front, 4 lower frame rate camera in each corners of the car. Keep cost below $5000 and use image processing to determine distance instead of a Lidar. I don't claim to be an expert in this field, but I've done some work in image processing with the 360 degree camera system that Elon mentioned. Hopefully someone better than me can find some fault and we can debate about possibilities.
Hardware setup
If I remember correctly, the two major players in image processing hardware space are Cognex and Matrox. Cognex has similar offerings at a lower cost, but Matrox has a better API.
�One capture card with 8 input feeds ~$2000
�One high frame rate camera ~$1000
�Four low frame rate cameras ~$2000
�Bumper sensors piggy backing on airbag sensors ~$0
�Shared CPU with car console ~$0
Five cameras with one high frame rate in front and 4 lower frame rate cameras at the four corners of the car with overlapping field of view. Autopilot can piggyback on the sensors used by airbags and use them as a collision detector for last line of defense, saving some cost. The biggest variable in the cost, is in how much camera we need based on the requirement. Obviously, the more camera we slap on, the cheaper the camera and image quality if we want to keep the cost to $5000.
The front high frame rate camera's main job is to determine relative speed and prevent a frontal collision, this is where things gets a bit fuzzy, we need to figure out what frame rate is necessary for this to function properly and what transfer delay from capture to the cpu's memory is acceptable. For a surgeon's medical console, I know that the system needs to be 60hz, 10bit color and 30 msec maximum delay between capture to display. So we can start with that as a reference design.
Roadblock 1: Relative distance of moving cars
The next problem is sensing distance. This is the problem that Lidar solves and a camera based system might have problems with. With cameras, it is hard to determine how close an object is, because an object can just be big and cover a large pixel area of the sensor without being near the car. The one way I can think of is to use image processing techniques to process this information. Here is the trick I think will solve the problem of how to determine relative distances of cars nearby: License plates. License plates are state issued and have a standard size. It is also something that is on every car and has a white background which is very easy to be picked out by image processing software.
Roadblock 2: Pedestrians
Pedestrians present a different challenge as they are mostly immobile when looked at from the point of view of the car. The most important thing about pedestrians is determining their distance to the car. Assuming they are immobile, we can use the difference of size between subsequent frames and the car's own speed to determine how fast the pedestrian is approaching the car. Again, each person is different and their height will affect the judgment of this process, but a person can probably be estimated based on how far off the vertical axis they are in the image sensor and how much vertical pixels they occupy.
Cost breakdown
I am only familiar with Matrox's cost, so using the Matrox QxT line of product where 8 inputs can be received and captured by one PCIe card connected to a PC, we should be able to reduce the cost of electronics, pushing the majority of the system's cost onto the hands of the cameras themselves. Here's the tricky part that I believe can make or break the bank.
If my memory serves me correctly, an Dalsa camera normally paired with these capture cards in an industrial setting can cost up to $10,000 with the lens included. Multiply that by 8 and we have the autopilot system that's equivalent to the cost of a Model S. As a consumer, it does not make sense to buy that... unless you are in Norway. From a back of the napkin calculation, Model S owners might be willing to shell out $5,000 for technology based on the evidence that most people choose the tech package that cost around $3,750. $5,000 on a $75,000 Model makes sense since that's only 6% of the overall cost. $5,000 on Gen 3 that cost $35,000 might not make sense for people as that represent a full 14% of the cost.
So let's limit the scope to model S for now. With Matrox's Qxt, the cost of each capture card is around $2,000. Assuming we are using the on board CPU to keep cost in line, we have $3,000 left over to buy 5 cameras and we are left with off-the-shelf type consumer type cameras. Personally I think it is fine, because the on board CPU can't handle the load of processing 8 high def cameras anyway.
Disclaimer
All opinion and information are for Tesla Motors Club discussion only. Please do not quote me anywhere else, nor use the information for any type of formal debate. I don't claim that I know anything, nor do I claim that any of the cost here are accurate. These are back of the napkin estimates and information pulled from my memory from something I did a couple of years ago.�
Sep 20, 2013
vin5xxx Elon states that their autopilot system will cover 90% of driving situations. I would guess he is probably referring to commuter travel and long-distance travel. Maybe also neighborhood travel. That would not require a complicated array of cameras, sensors and computers. The picture shown above would probably fall into the category of the 10% that autopilot would not work in. So probably the driver would need to take control in parking lots, inner cities, construction zones, etc. But for vanilla driving mostly on the highway I think the system could be built rather inexpensively.�
Sep 20, 2013
brianman I'd love to see "Autopilot does Autocross".�
Sep 20, 2013
sleepyhead The whole idea of autopilot is for it to work especially in parking lots. This is where you have the most accidents.�
Sep 20, 2013
vin5xxx Having autopilot navigate through a parking lot would be extremely more technical than autopilot driving in normal traffic. Pedestrians, shopping carts, lack of right-of-way signage, etc. all create too many variables for a computer to figure out. In order to get an inexpensive system out in 3 years I believe it will only operate the car in vanilla driving scenarios such as I described above. That cheaper and simpler system makes sense. Again, the intent is for 90% of driving scenarios.�
Sep 20, 2013
vin5xxx What I am describing is a 1st Gen autopilot system. I'm sure future generations would be able to navigate parking lots, etc. I would classify google's autopilot as a future gen system because of its higher LIDAR costs.�
Sep 20, 2013
aronth5 I would tend to disagree with this. As vin5xx states parking lots, inner cities, construction zones are where the drive will need to take over. At least with the first generation autopilot systems.
Highways and local routes should be the goal initially.�
Sep 20, 2013
brianman Many people have many different reasons and goals for "autopilot" and "self-driving" and such. Assuming you have a monopoly on "the reason" or "the idea" is folly.�
Sep 20, 2013
confy Given that Google's self-driven cars are still in beta, it would be very surprising if Tesla can come up with a road-ready self-driving feature in three years. The general self-driving problem is something researchers have been working on for around two decades now. Perhaps something like the Benz S-class's autopilot is within reach, but not the sort of system Google is trying to build.�
Sep 20, 2013
Causalien My reasoning is this:
Even on highways and autoroutes, the primary failure case to avoid at all case, is hitting a pedestrian or a car. Because once you do hit one, a tsunami of politicians will vie to shutdown any type of autonomous driving. "Robot kills Human" will make for some great headlines and gain you more votes from your constituents.
Therefore, designing the AI to avoid collision with human and cars will have to be built right away. Just because you are on a highway, doesn't mean that you can stop detecting for collision with a human. So if that system is always active, then it means that you might as well design it to be able to work in cities and normal roads where pedestrians are present.
What I am getting from you guys is that $5000 is too much for this autopilot package?�
Sep 21, 2013
Mario Kadastik I think the main challenge will be in crappy weather conditions. I was driving today in deep fog that had a total whiteout ca 30-40m from the car I assume (was able to see ~1.5 spans of the street lamp posts). That's extremely hard to decipher using a camera based system unless it has also filters and sensitivity to IR etc. Second crappy weather would be serious rain with splatter and dirt getting onto the cameras making the pictures blurry and out of focus in varying regions of the view. In good daylight conditions it's not that hard, but it's the crappy weather that has to be worked on...�
Sep 21, 2013
Tasmanian Devil I am still very sceptical about the whole concept. If Elon would prove me and the other sceptics wrong my stock would be happy so no problem here. I don't have to be in the right. But it is imperative that an autopilot system
is 100% fool proof. Saying it will apply to 90% of all situations is probably not going to cut it. Who decides about those 10% ? How are they handled. You have to take into account human behaviour at the moment You go down
to 99% fro 100%. That opens the door wide for dubious argumentation in case anything serious happens. If You do it this way it's bound to be a sue-fest.
Also I see a big problem in the interface, i.e. the point where the driver takes over or the other way around, where he gives control to the car. When will this happen ? I can easily imagine situations, where things get
a little narrow or hectic and the human driver suddenly loses trust in the system, takes over and causes an accident doing so.
I still see a Pandora Box of problems here, as much as I like Tesla to succeed.�
Sep 21, 2013
WarpedOne I suspect this AP is something close to battery swapping. We have it but it is not available/usable by you.�
Sep 21, 2013
jerry33 I'd suggest that the autopilot would alert when the conditions were beyond its capability. Something like this:
CURRENT MODE: Autopilot off
No indicators.
CURRENT MODE: Autopilot on standby
Conditions within autopilot capacity. Autopilot ready indicator green
Conditions not within autopilot capacity or marginal. Autopilot ready indicator red
In standby mode it would operate similar to Mobil Eye where it alerted the driver but would take no independent action.
CURRENT MODE: Autopilot on
Conditions within autopilot capacity. Autopilot engaged indicator green
Conditions marginal but within autopilot capacity. Autopilot engaged indicator yellow and audio alert (Driver can choose to leave it on or not)
Conditions not within autopilot capacity. Autopilot engaged indicator red and audio alert (Autopilot turns off)
The first release would have only limited times when the autopilot could be engaged. Perhaps partly GPS based. So rather than trying to figure out if it was in a parking lot by sensor readings it would do it by location as well as sensors. On standby, it would still give the driver clues.�
Sep 21, 2013
mrdoubleb I think Elon is thinking outside the box, like he usually does.
Remember, before Tesla, engineers would have laughed at the idea of building an EV out of laptop batteries. (Yes, i know these are not exactly laptop batteries, but you get the point). The industry standard was using industrial-grade, relativley low capacity but highly reliable components and chemistry, which, of course, also meant range was low, yet prices were high. Then came Elon, and he changed the paradigm.
With self-driving cars, one of the biggest issues is that they are trying to make them work in every single situation, every weather condition. Even Airbus and Boeing don't do that, AFAIK pilots only engage autopilot in good weather, at cruising altitude, for things like hours of flying in a straight line. This Google-approach means, they have to use lasers, high performance computers, etc. which make the car look and cost as much as an aircraft carrier.
I bet Tesla's system will warn and disengage when there is too much fog, too heavy rain or snow, ice, unexpected construction work on the road etc. and only work in "normal" conditions. On the flip side it will be simple and (relativley) inexpensive. A also think it is only intended for the next gen Model S, not the Model E, as it will still be a costly extra. I could be wrong about that, though.�
Sep 21, 2013
aronth5 Agree. Like anything else it will be done in phases and will improve as each generation of self-driving capabilities is available.�
Sep 21, 2013
vin5xxx Jerry33, mrdoubleb, aronth5 - I think we are all on the same page. This first generation of autopilot will be quite simple. I think many people here are imagining a further complex generation that will be introduced down the road.�
Sep 21, 2013
rdalcanto LOL! The guy can start with spare parts and build the best Space/Rocket program in the world. He can put satellites in orbit and dock with a tiny station in space. I think he can safely get a car to grandma's house....�
Sep 21, 2013
brianman IMO, "Auto pilot" has a connotation of "it does the easy boring stuff that puts you to sleep". Building this technology is far simpler and more likely be reliable than park-to-park self-driving.�
Sep 22, 2013
AudubonB First responses to Causalien's initial nuts-&-bolts post:
1. License plates do NOT uniformly have a white background, so that's a very minor tweak.
2. Although I like the "cars move fast so there's one algorithm to deal with them; people are essentially immobile so another algorithm takes care of them" argument, now you have to throw into the mix:
*motorcycles - small, fast and with a reputation, deserved or undeserved, for jinking. And with far smaller license plates, too
*bicycles - small, slower than most cars but not so immobile as pedestrians, and with a reputation, deserved or undeserved, for jinking. And with infnitely smaller license plates, too. On the other hand, their tires, etc., might become the default sizing=distance determinant
*joggers/sprinters - pedestrian-sized objects with a reputation for jinking
*and then there are dogs, cats....or where I live, caribou and moose (talk about immobile!)�
Sep 22, 2013
ItsNotAboutTheMoney Do you use cruise control? To me it's absolutely no different in principle and in fact auto-pilot would be better than current dumb cruise control systems that are found in the vast majority of cars. Why are you worrying about this, when you should be railing against the millions of accidents waiting to happen right now?
Auto-pilot would include the adaptive cruise that's found in cars with advanced safety systems (and is by far the most popular component). The car has a set of parameters within which it is expected to operate and will run on auto-pilot as long as it remains within those parameters. Auto-pilot's parameters will never ever be outside human norms, because otherwise the car would get into situations where it would be dangerous for the driver to take over.
The only question is how much software will be able to do.
What I really like about Tesla's approach is that they're starting with the end goal in mind. The end goal defines the hardware requirements, which should mean that they'll be able to sell systems (ha, cars) with the hardware and limited software capability and then upgrade the software to introduce new features. In the traditional model, manufacturers use technology to sell hardware upgrades, but the risk of that is that when you lag the competition you are forced to cut profits to compete. With the Tesla model you have a primary focus of leading on the hardware, with the knowledge that the software can catch up. This is part of the reason I always expect delays from Tesla.�
Oct 1, 2013
Ludus Elon released that video of him playing with a leap motion controller and various display systems using high end Siemens CAD. The other components were quite expensive but the 3D motion tracking package with two cameras, infrared, and the processng chip (Leap Motion) is about $80 at Best Buy. Check it out. That provides real time precise tracking of all 10 fingers...an order of magnitude better performance than xbox connect...the other model.
For actual product I think he expects the cameras/sensors and hardware processing to be very inexpensive and unobtrusive, something like Leap Motion adapted for longer distance and deployed in a fixed 360 array. Maybe a dozen or more cameras with integral motion capture processing chips and infrared arrays together costing less than $1000.�
Oct 1, 2013
jeff_adams I saw a video recently where someone flew a pilotless F-16. Not just takeoff and landing, but full acrobatics.
It wouldn't surprise me to learn that Elon saw it and that inspired him to make this a higher priority.�
Không có nhận xét nào:
Đăng nhận xét