In his Master Plan, Part Deux, some of Tesla CEO Elon Musk’s visions make a lot of sense, like a compact SUV based on the Model 3 platform, and a pickup truck, which will presumably have more than enough torque to haul around a big load (and plenty of space for a big battery pack).
That’s not at all what I want to talk about. Instead, I want to talk about some of the more speculative parts of the Musk vision: the self-driving vision. (I first wrote about some of the dystopian aspects of self-driving cars for TTAC in 2010.)
Musk wants you, the owner of a Tesla, to click a button and have your self-driving car go off and Uber itself during the day while you’re at work, earning you money. It will come back to you at the end of the day, ready for you to use again. This vision is going to have a very harsh collision with reality.
Let’s assume you’re not the sort of person whose car is strewn with personal effects, so you’re not worried about your cars’ passengers stealing your things. Even then, you still have to worry about passengers who smell bad, vomit, or otherwise damage your car. Even if you’ve got good insurance to absorb the monetary damages, you’re now living without your car while it’s being repaired, no doubt availing yourself of the very car-sharing service in which your own car once participated.
My day job is conducting research in computer security, so I want to delve deeper into things that can go horribly wrong. It’s how my brain works. Bear with me on this.
Let’s say you’re interested in stealing cars. If you could click on an app and summon a car — which would arrive without a driver — then you’re a very happy camper. How might that work? First, you create a fresh account on a burner phone using a stolen credit card number. You travel to some abandoned warehouse district, then you summon the driverless car. When it arrives, you promptly box it in with other cars. The Tesla’s Autopilot will be far too polite to fight its way out, and now you’ve got the car trapped. Of course, you brought a flatbed tow-truck along, so you can hoist the car up and clamp it down.
Your next mission is to disable the radios so Tesla’s central command can’t track the car. Each car will have antennae hidden in different locations, so you’ll have to know what you’re doing, but all the cars of a given make and model have the same antenna package, so you know what to do before you head out to the chop shop. Easy.
Needless to say, Tesla will be keenly aware of your gambit and will keep around honeypot cars, loaded with extra radios and GPS trackers and such, so you think you’ve got it made, but they’re tracing you all the way home. This sort of back-and-forth one-upmanship is something that happens in a variety of security scenarios, and it’s wildly unclear who eventually wins the game.
Now consider a less futuristic, more presently plausible scenario: the self-valeting car. I don’t want somebody else polluting my precious bodily fluids by messing up my car, and that includes the evil valet parking attendants who set up orange cones to block me from using any of the perfectly reasonable parking spots at my favorite restaurants. Instead, my car can drop me off and go find its own parking spot. If we’re only a few miles from home, the car could just head home alone, open the garage, pull in, and chill. If we’re far away, Tesla could have an arrangement with a nearby pay lot. Sounds great, right? Well, all the same theft scenarios present themselves. You can imagine roving bands of car thieves, just looking for empty cars in motion. Block ‘em in and haul ‘em out.
When a group of researchers from UC San Diego and the University of Washington studied this in 2011, they conceived of a sophisticated thief who could remotely hack a car and issue it guidance commands, which was entirely feasible with the GM cars they happened to be looking at. (GM didn’t fix this until 2015.) With that car, a remote high-tech thief could remotely unlock and start the car, but it would take a local human to drive it away. With the next-generation Tesla Autopilot, the local accomplice would no longer be necessary, nor might it raise eyebrows to see a car start itself up and drive away without anyone behind the wheel.
As before, we get into a game of one-upmanship, where Tesla must defend its fleet against creative attackers, and Tesla customers have to hope that Tesla will do a decent job of it. Tesla, in fact, seems to be doing a reasonable job of pushing regular software updates over the air to its cars, so it’s at least in a position to fight the good fight. But consider other car manufacturers. If we can’t trust Volkswagen to correctly meet U.S. diesel emission specs, can we trust them to actively defend their cars from electronic attacks, even a decade after they’ve left the showroom floor? What sort of requirements would it take to keep them vigilant?
It’s tempting to make this a regulatory function, but VW diesels passed their regulatory tests. As another example, electronic voting machines, with which I’ve been involved for 15 years, are all “certified” against voluminous standards by ostensibly independent third parties. Despite this, any security expert can drive a truck through the holes, and older machines are still widely used without any software updates for their known flaws. (If the election’s really close, I’ll be busy this November.)
My conclusion is that certification and testing are not going to cut it for truly autonomous, self-driving cars. We need something better, which leads to a somewhat surprising solution: Tesla and other automakers are going to be forced to move to in-house insurance companies. “Buy our car, and we warrant it against theft. No deductible.” This aligns all the incentives properly. You want to sell me a fully autonomous car, then warrant it against theft and other sorts of electronic attacks (e.g., an adversary might want to use Tesla’s online network to track your physical location).
This matters to any car owner, and it will really matter when Tesla gets into the autonomous 18-wheel tractor trailer business. When you’ve got a semi with potentially millions of dollars of inventory in back, it needs to resist tampering and theft long enough for a (non-autonomous) police car to be dispatched.
[Image: Tesla Motors]
“Tesla and other automakers are going to be forced to move to in-house insurance companies.”
Holy sticker shock, Batman! Even assuming they partner with an insurance company to provide this single product rather than cover it internally (likely), that’s going to be pretty expensive.
Manufacturer-provided insurance in new in the US, but very common already in the UK and elsewhere. https://www.carwow.co.uk/best/cars-with-free-insurance
Interesting.
Well, could be, or the manufacturer could want the car insured to the point that when it’s wrecked, the owner has an easy time buying a replacement.
I’m sure they’ve looked at this line of business. Maybe some kind of regulatory thing here that prevents it?
It might be a necessary plan. Reuters is reporting another unexplainable autopilot crash – this time in China.
Enough of these autopilot crashes and insurance is gonna become difficult to obtain for these cars.
Not really. The insurance company just has to say that autopilot has to be disabled because it’s an unassumable risk.
Enough of that and Tesla will just stop making the car with this feature.
If actuarial data supports such an action (i.e. concrete data shows Autopilot is worse than Joe Sixpack at driving) then yes that is a likely outcome.
On the other hand, two accidents doesn’t mean its time to pull the plug. Insurance companies will warm to Autopilot *if* they conclude it’s safer.
Right now we can all speculate that its either safer or more dangerous, but until the numbers come in, it’s all conjecture (and good media material).
In the car world, Tesla production numbers are pretty low and the repair/replacement costs are pretty high so I wouldn’t be surprised if they look a little risky to insure. Who knows what the insurance companies actuarial tables look like for Teslas? I would have to guess a fair number of them have been crashed just because they’re fast cars and drivers can run out of talent pretty quickly.
If their security protocols are good, the policies should be inexpensive. It’s the only way I can think of that would make such security protocols actually effective.
I would consider the possibility that these “plans” are intended to create buzz and convince the public that Tesla is a visionary technology on the cutting edge of innovation. This is a marketing ploy, not a business plan.
I would agree with that, PCH,
All automakers are running scared of the idea that Uber could essentially dis-intermediate them. Tesla at least has leading positions in the technology capabilities that will keep them close to their customers.
I don’t think that you do agree with me.
I’m saying that it ain’t gonna happen. It’s just talk.
I agree that is is just talk now, originating from Missouri (i.e., the ‘show me’ state). I don’t think it’s an impossible scenario, but I hope that Tesla is spending the bulk of its energies on bringing the Model 3 to market successfully.
You think that this is a serious plan.
I don’t.
I don’t see why it’s out of the realm of possibility. At least as an *option* for owners. it’s not like he’s saying it’d be compulsory.
Tesla has done a fantastic job of convincing some of the market that it is a forward-thinking innovative technology company even though it is in the business of manufacturing old tech.
This is just part of that effort. The point is to talk about it for the sake of it so that the Tesla fan club gets all gooey just thinking about how wonderfully futuristic it is. It makes them want to be a part of it because Tesla becomes a symbol for the daring visionaries who they personally want to be.
Early adopters like to feel as if they are on the cutting edge, and Musk is telling them what they want to hear. Whether or not it actually happens is completely irrelevant; you keep saying it in order to reinforce the story.
Yeah if I had money down on a Model 3, I’d regard every other of Tesla’s pet projects as a threat to its timely release. All their focus should be on getting the Model 3 to market because if it’s delayed there will be a lot of angry people.
And there’s no way in hell I would let my unmanned Tesla wander around town picking up strangers. I can’t imagine anybody wanting that. People who can afford to buy a Tesla aren’t looking for extra pocket money.
To me this looks like more of the technical side-show distracting folks from the reliability and delivery problems on the production models. But as they migrate into the heart of the market volume-wise, they’re going to have to produce.
And btw B&B, nobody answers my oft repeated question…
who makes the first retail delivery?
Tesla 3
Elio
Mitsu PHEV
Mitsubishi PHEV, the Elios snake oil at this point, the 3 will be so low qualiry it’ll disappoint its pre-order saps.
The PHEV will launch but no one will notice.
Agreed 100%. Shareholders are losing patience with the oracle. Tesla got a good bump from the second tech bubble but if they want to exist long term they have to transition to becoming an actual company that makes stuff.
If they can get in to disconnect the antennae can’t they just hack it with the OBD port?
In a related story:
“You can imagine roving bands of car thieves, just looking for empty cars in motion. Block ‘em in and haul ‘em out.”
Are roving bands of car thieves (complete with shoulder spikes and mohawks I assume) just going to ignore all of the cars NOT in motion? If they have a flatbed they can grab such a car today…without even using 3 other cars to box it in! Better hurry up and steal all the cars you can today while they are all stationary!
To be any more serious of a problem than today’s car theft or ubering, some of the scenarios in this article require enough of a downfall of society to halt any technological progress required to enable mass market self driving cars in the first place!
OK, so…I’m supposed to be all right with my Tesla going off doing its’ own Uber thing all day long while I’m at work?
Nope. Not OK at all, on any number of different levels.
1) I don’t want to get in my Uber’d Tesla after a hard day at work and find the back seat full of empty beer cans, or the remnants of joints, or “precious bodily fluids” (tip of the hat to Kubrick). Yecch. No thanks.
2) Speaking of Kubrick, what happens when my Uber’d Tesla decides to go all HAL 9000, and hits a mother of six as she’s walking down the street? I’ll tell you what happens: I own the car and I’m liable. Again, no thanks – I don’t care if I’m insured or not, I don’t want to deal with that.
3) It’s obvious to me that Autopilot ain’t exactly fully developed, so even if I have no legal or financial liability (in theory), I’m not comfortable with my car driving itself around all day.
And that’s strike three.
I’m sure this is marketing BS, but I’m not interested in any of it…not one bit.
You do realize this is voluntary. No one would force you to whore out your car as a taxi, right?
Put differently, then: I’m not comfortable with the option of doing this.
(Yes, I know no one’s forcing me…but that doesn’t make it any better of an idea.)
Let’s say your Tesla costs you $800/month. But you earn $50/day as a taxi, and then pay a kid $10 every evening to clean it. You come out even, i.e., free Tesla. Still opposed?
The $10/day cleaning kid would have to have running shoes on. Here’s why.
It’s 11:30am and you get a notification from Tesber that your most recent fare has asked for a full refund and given you zero stars out of five. You discover that the 10am fare had a party in your car, left open beer containers in the car along with a used condom on the steering wheel. The interior is peppered with cigarette burns. Of course the 10am fare cannot be contacted, the fares info turns out to be fake.
Sure the hidden cam you installed captures it all and makes for some very interesting viewing, but you are still left with the hassle of getting you car fixed and claiming on the uninsured motorist part of your policy.
To me, one of the advantages of owning a car is that I get access to it whenever I want. If I want to run an errand or go out for lunch, my car is in the parking lot at work. If I’m home and I want to run to the store or get a craving for ice cream, I can hop in my car and go. I don’t have to worry that I can’t use my own car because someone else rented it.
Plus, the demand for cars is going to be the highest at the same time I want to use mine – rush hour, after-work errands, weekend trips. It’s going to be lowest at the same times I’m not using mine – when I’m at work or asleep.
Antennas can be stymied by throwing a foil blanket over the whole car.
I agree with the writer’s concerns, but another roadblock to this dream is cultural behavior. Very few of us are willing to let a good friend borrow our car for a weekend, or a month, let alone rent it out daily for anonymous plebes to use.
Not gonna happen.
“Very few of us are willing to let a good friend borrow our car for a weekend, or a month, let alone rent it out daily for anonymous plebes to use.”
You say that, but consider five years ago. If you asked people “Would you like to be a taxi driver part time?” they’d inevitably have said no. Now look at Uber and Lyft.
It’s all about the marketing and framing. Even if it’s not profitable, people will do it if it’s trendy and there’s a buzz.
I’m not letting anyone borrow my car except for family. Family consists of my parents, my wife’s parents, and our sisters. Everyone else is on their own.
I was just about to ask to borrow your C-Max this weekend. I need to transport bird seed, sand, and wet dogs.
Have it back by Monday.
“have your self-driving car go off and Uber itself”
I don’t know what Uber is but if this is what it sounds like I hope my car is at least going to be discrete about it.
This whole article sounds like it came from The Onion. And I can see the follow up headlines:
Tesla leaves owner after dispute over sub-standard parking conditions
Striking Teslas found under overpass drinking high octane gasoline
Tesla owner arrested for pimping
Your Tesla has the greatest possible enthusiasm for this mission, Dave.
Uber is a US based car service that is used around the world in large cities, SCO owns the car and works as a part time taxi, I use a app on my phone and ask for a ride to the airport for example, you have a UBER app on your phone , it tells you where I am, I get a message saying SCO is on his way and a picture of you and what kind of car you have, you pick me up , drive me to the airport, I pay UBER via a credit card and they pay you the driver a percentage of the fare which is usually much less than a taxi.
So during the day, your $100,000 Tesla is a Johnny Cab? No thanks.
My reasons for being skeptical about the ‘rent out your car while you are at work or the gym’ idea are much more mundane.
Many people use their cars like trash cans, McDonald’s wrappers in the passenger foot well etc. Asking these people to keep a neat and tidy car to rent out is a behavioural non-starter.
For those of us that don’t use our cars as trash cans, ever store something you’d rather no one else have access to in your car? Security badge, check book, keys to the house? Another behavioural non-starter.
Oh I have a solution for the trash issue. When you have something you don’t want in the car, throw it in my yard!
I live on a somewhat main street, and at least three times a week someone throws trash into my grass. Can, cup, whole bag of restaurant garbage, and even a Coke bottle half full of Coke and cigarette butts. Almost threw up with that one.
People are disgusting.
Too true, its quite common to see cars filled with old greesy Fast Food bits thrown about, or the driver smokingvaping away.
I don’t want that in my cars, not even in my older junkers.
At Corey: You should setup a hidden camera and just make a video of all the trash drive bys at your place.
Would that help anything, I’m not sure the police would be willing to pursue fining people even if I got an image of their plate.
Yep. That about sums it up.
Way back when we lived next door to a fish and chip shop and opposite a pub. Real handy if you want a beer and some fast food. The other patrons however thought our front yard was a great place to deposit fish and chip wrappers, cigarette buts, you name it. We made it especially convenient in that the front wall was the perfect height for sitting on. That way you could come out of the pub rip roaring drunk at 11:30pm, get some fish and chips, sit on our wall to eat all while talking and laughing very loudly as drunk people do and then leave all your crap in our front yard upon departure.
No but it’d be entertaining, just a sped up video of people mindlessly littering. Maybe it’d fo something but thats a crapshoot.
I do not see part time taxi service but I can see drive to work and then home, pick up said kid, drive them to school, back to work for me so I can use said car for going to lunch, back to school to get said kid, drive him to practice get him from practice to home, and back to my office so i can drive home.
Fully autonomous cars are simply today’s tecno-porn for today’s tecno-geeks, like flying cars and mag-lev trains and whatever else Popular Science envisioned us all using in The Future. They aren’t going to come about without a massive improvement in our current infrastructure, processing power (and “decision making skills”), and regular-folks’ social acceptance.
Groundbreaking opinion, I know.
Simple – legalize drugs and use the newly-freed officers to enforce crimes with a victim.
Not Musk’s most brilliant thinking lately, I agree. I looked it up in the Tesla forum, mot many takers for sharing their priced possessions. The more posh and the pricier, the less likely people want to share their car. You need a more tool-like vehicle. Google is still closest with its robo pod, although that particular one saps all the fun out of going from A to B in an ‘auto-mobile’.
Or…
Tesla modifies their chargers to have bi-directional communication, and lets them talk to the nearest power grid operator. There are seven main grid operators in the U.S. and another three or four in Canada, so this is a manageable population.
Using existing standards and technologies, the Tesla chargers can respond to signals from the grid operator to use the car batteries for small bursts of charging or discharging, to help keep the overall grid frequency constant. This is called frequency regulation and grid operators already pay big generators to provide it.
Batteries are ideal because they have very fast ramp rates. One car by itself doesn’t do much good, but if you have a population of several thousand, all synchronized by a back-end server, it now becomes very feasible and puts barely any strain on each individual vehicle’s battery.
Benefits: your car is parked, no risk of trash or other downsides of renting your car out to strangers, it helps with the overall grid stability, and you’ll get paid (not much, admittedly). The University of Delaware did a proof of concept several years ago and it worked quite well.
Not as sexy and “wow”-inspiring as Mr. Musk’s idea, but it is useful, will benefit both the car owner and the community (grid stability), has been proven to work, and is a lot less risky. Call me stupid, but I think it might be more appealing to the general population.
No sir, I don’t like it.
“Your next mission is to disable the radios so Tesla’s central command can’t track the car. Each car will have antennae hidden in different locations, so you’ll have to know what you’re doing, but all the cars of a given make and model have the same antenna package, so you know what to do before you head out to the chop shop. Easy.
“Needless to say, Tesla will be keenly aware of your gambit and will keep around honeypot cars, loaded with extra radios and GPS trackers and such, so you think you’ve got it made, but they’re tracing you all the way home. This sort of back-and-forth one-upmanship is something that happens in a variety of security scenarios, and it’s wildly unclear who eventually wins the game.”
———————————–
Easiest way around that one is to build a box van on the rollback (hardly impossible) and build a Faraday cage into it (also pretty easy). Take it into one of the warehouses or off to a shipping port, load it into a Faraday shipping container and off it goes; assuming it’s not ‘chopped’ on site. Really, the concept has already been demonstrated in the way so many other high-end cars are stolen and shipped overseas already.
I suspect that if (when) self-driving cars happen, they will increase traffic congestion so much that many drivers of conventional cars will be forced off the roads in frustration for large portions of every day.
This will create a huge additional demand for Uber type services from people who own normal cars but just can’t bare to drive in the worst of the traffic. Or from those same people who know they won’t be able to find parking when they get to their destination because of the increase traffic. This is definitely already happening here in the silicon valley. For example, when I wish to go to San Francisco, I’ll park 10 miles away and take the subway just to avoid the hassle of having to park.
Self-driving cars will increase congestion in these (and more) ways:
(1) So “safe,” “courteous,” and “law abiding” that they are just slower on the road and at stop signs. (Especially when bicycles, pedestrians, and non-self-driving cars are around.) Humans will make a sport of flummoxing the self-driving cars. For example, I saw how much fun people had around here by walking towards an intersection where a google car was trying to cross; even though the person hadn’t reached the road, the car waited patiently for the pedestrian still on the sidewalk.
(2) People who can’t legally drive now will be using cars. (Children, people who have lost their licenses, people too old to drive, people from other countries who never learned to drive, etc.)
(3) People will put a bed or living room in the back and just zone out as they are driven … long distance? Who cares! Right in the middle of the worst rush hour? Not a problem! If I can be as productive in the back of my self-driving car as I can be in my own livingroom/bedroom/office, I’ll take all sorts of new trips.
(4) Parking… With all these people now going all over the place for dining and shopping, there won’t be enough parking; so the self-driving cars will be just circling the streets waiting for me to finish and drive me back home. Or the car will drive home to park and then pick me up again.
People say that self-driving cars will optimize traffic so well that more cars will fit on the roads and higher speeds will be achieved. That could eventually happen for highways. But for places with mixed use (pedestrians, bicycles, etc), optimization isn’t really possible.
Only one of those complaints is truly legitimate, Sunny, and that’s the idiots who go out of their way to flummox a self-driving car in one way or another. They’ll be just like the “markers” who spray such elegant graffiti on to train cars, trucks and buildings. They’ll be just like the street racers who see it as a challenge to travel at full speed through moderately- and sometimes even heavily-congested traffic. The point is that when humans are taken OUT of the equation, traffic should travel more smoothly, especially when we reach the point that each vehicle can tell the ones around it where it’s going and why, at which point not only will the courtesy be obvious but the LACK of courtesy by so many will be non-existent.
Just yesterday I saw the results of a crash where someone signaled to change lanes and the vehicle behind moved up to block the lane change, resulting in one vehicle sitting on either side with cops in attendance at both and both losing much more time than they would have saved through their impatience and discourtesy and, as a result, slowing overall traffic through the crash scene. Worse, my wife witnessed a three-car accident this past weekend under the exact same kind of conditions. It’s not the courtesy that kills time, it’s the discourtesy by human drivers who tend to cause accidents, even if they don’t ever get involved themselves.
Anyone can push all the buttons on an elevator on their way out the door, if they want. I’ve never seen it happen.
Just because you haven’t seen it doesn’t mean it doesn’t happen. I, personally, HAVE seen it and many more than just one time (no, I was not the one doing it.)
So don’t say it doesn’t happen when it can. It does happen, even if you never see it yourself.
“Just because something is imaginary, doesn’t mean it isn’t real.”
“Yes, that’s exactly what it means!”
Care to clarify, Corey? What’s the “imaginary” part?
A quote from South Park, Imaginationland episode. That’s all.
Ah. Thank you. I especially loved the “World of War…craft” and “Casa Bonita” episodes.
More sopapillas please!
Airy and delicious, especially when you poured honey into them after biting off a corner.
Vulpine, I suspect you don’t live in Silicon Valley. Around here, people actually do commute 1.5 hours or more _each_way_ for work. “It’s OK, I just get on the Apple/Google/Facebook bus and arrive at my destination as I take care of my first hours’ worth of coding.” Once you buy an electric self-driving car (like Tesla is promising) the personal costs go down sharply. Since it’s self-driving the time-cost is no worse than the bus. Since you get free electricity from work (!!!) the propulsion isn’t much more than the “free” bus. A self-driving Tesla will be much like the “free” shuttle bus … convince people to travel much farther. However, with a self-driving electric car, people will not just be traveling much farther for work… they’ll be traveling much farther for a large number of reasons. Since the self driving electric is even more convenient than the bus, current bus riders will jettison the bus and take their cars (these high-tech companies _already_ have free valet parking because the sheer number of cars they need to accommodate).
>>> when humans are taken OUT of the equation <<<
Umm….. ever been in a busy city? Pedestrians, dogs, bicycles, zombies playing Pokemon Go, etc. None of these entities will be hooked up to your centralized control system; they will not contribute their situation or intent to the centralized control system and they will not obey commands from the centralized control system. I'll concede that limited-access roads could be greatly improved, though. But the busy mixed-use streets will be much much worse. Just watch a google car for a while in mountain view.
Well, we live in interesting times! It's (mostly) fun to watch the "progress" of these things. I just hope my shack increases a few hundred thousand dollars this year as a result! :-)
The circumstances of which we were talking have no connection to what you’re describing, Sunnyvale; Tesla has made it quite clear that their Autopilot is not ready for that kind of (city) driving. The article itself made it somewhat clear that the Autopilot only took the car from the point where he’d called his wife to the EXIT from the freeway near the hospital. It was not made clear and is quite unlikely that it drove itself from that exit to the hospital itself, where again it was stated that the driver manually parked the vehicle and was immediately processed into the emergency room.
Google is much closer to the in-town style of autonomy but its OTR capability is lacking; simply doesn’t have the scanning speeds or processing capacity to handle high-speed driving. That roof-mounted Lidar system can be effective, but its price and positioning making both prohibitively expensive and impractical for real-world driving due to the risk of damage both accidental and criminal as well as carrying too many blind spots around the vehicle itself. Scan speed could be accelerated through narrowing the sweep and using an electronic sweep rather than physically-rotating emitters. This would improve both the resolution of the scan and allow it to process data much more quickly–in the order of as much as 600% faster due to losing 60% of its sweep angle. Radar has already proven that phased-array radar has a faster sweep than a physical sweep and offers near-photographic-grade returns with combat mapping radar. Using Lidar to create a 16-bit “greyscale” image would become much easier and more accurate especially when combined with co-located radar and optical sensors.
Combine Google’s research with Tesla’s and the other systems currently in testing and you would have a nearly fully-aware sensor system for autonomous travel that would be many times more accurate and predictable than any one of those systems today.
@vulpine: the big mounted on the top LIDAR is on its way out. We now have the Velodyne VLP-16 & VLP32 pucks that can be mounted under the mirrors on each side. The 16 is only about $8k each.
As far as leaders in full autonomy are concerned, I’m not sure I’d include either Tesla of Google. Tesla’s current technology won’t cut it for full autonomy. I suspect that googles management has finally figured out how far away full autonomy is, so who knows what is going to happen with that program. I wouldn’t be surprised to see them bail and sell the program to another company given what’s happened over the last couple of weeks.
” the big mounted on the top LIDAR is on its way out. We now have the Velodyne VLP-16 pucks that can be mounted under the mirrors on each side. They’re only about $8k each.”
And when cars lose those mirrors, as they inevitably will when cameras prove better and more aerodynamic?
Yes, I’m glad to see the things are getting smaller, but why have a physically-rotating part at all? Why not use laser diodes in an array that can electronically scan the vicinity in a relatively narrow arc of, let’s say 120°each to cover the critical areas at nose and tail and let ultrasonic sensors cover the sides? You simply don’t need long-range sensing to either side because quite honestly on a T-bone style collision, there’s almost nothing the ‘target’ vehicle can do to avoid it. Inter-vehicular communications will help but every T-bone crash is human caused due to one of the vehicles ignoring traffic signs and signals or assuming they can complete the maneuver before the other vehicle arrives.
Here’s the thing: We can’t wait for total perfection. First off, perfection is impossible. Secondly, it will have to come in phases, as it already has. We’re already reading about how major OEMs are very actively trying to build and promote autonomy long before it’s ready as even Chevrolet is announcing a fully-autonomous Bolt for use by Uber in less than three years. Google’s system does work well enough in inner-city traffic where high speed is unimportant. Tesla’s system does work well enough on Interstate highways where incidents tend to take time to develop and every vehicle is supposedly traveling in the same direction with typically no more than a 25-35mph difference in speed. Yes, in both cases the systems need improving. But they do work well enough to start the movement towards fully autonomous driving and I see it very possible that significantly better systems can be on the road in less than 10 years. It will take at least 20 years before autonomous vehicles outnumber manually-driven vehicles on American roads and much, much longer as you move down from the “first world” countries.
The trucks problem has an easy solution. The autopilot drives, but you put a minimum wage armed security guard aboard to resist hijacking, for any loads with a value high enough for that to be a problem.
The driver can’t legally drive for more than 11 hours a day, I think, but the security guard just needs to wake up if the autopilot hijack alarm goes off, so he/she/they don’t limit range or endurance.
I think it will happen. IIRC GM is testing driverless Bolts for Lyft — you will have to opt-in to accept one, and for now there will be a human in the driver seat just in case. If conservative GM is down for it, the technology is there; Tesla is just thinking one step beyond, per usual.
And I think it should happen. I dabbled in Uber driving during a big tourist event in my city. If I could make that money without actually doing anything myself, I surely would. Drivers and passengers are rated by each other on a 5-star scale, which should help weed out the slobs on both ends.
And I should add that insurance is provided by Uber — but only while the rider is in the car and the trip is underway. That should be revised to cover the trip to go pick up the rider as well, as that’s obviously part of the same journey. Right now that would just be the ethical thing for Uber to do (so chanza gorda), but with autonomous cars it will be necessary.