How a billion dollar autonomous vehicle startup lost its way

https://www.bloomberg.com/news/features/2018-08-13/how-a-billion-dollar-autonomous-vehicle-startup-lost-its-way

by Ibethewalrus • 8 years ago

79 40 8 years ago

ejlangev• 8 years ago

I suspect this will be a fairly regular story in the next few months/years. Self-driving tech seems to be a lot further off than people think so a lot of companies that were formerly very exciting will not be able to deliver and will need to close up shop. I think there's some possibility that it will go the way of flying cars at least for now.

Fortunately a lot of the tech that does work can find a second life as progressive enhancement for human drivers. Blind spot detection, auto lane following, and smart cruise control are great features that ease some of the burden in driving and I think it's good to see them being more widely distributed in newer cars.

curo• 8 years ago

I thought Waymo has pretty much cracked that nut. They've covered several million miles of testing in AZ and CA. You might be right, but do you have any links where I could see where your skepticism is coming from.

Waymo vs Quanergy is apples and oranges I think

paganel• 8 years ago

Not the OP, but I think rain and especially snow and fog don’t happen that often in those two States and afaik driving autonomously under those circumstances is still an unsolved problem with no easy solution in sight.

Now, if you want to extend the self-driving thing to other parts of the world you’ll most probably need to solve the singularity problem, at least for driving in the Eastern European capital city where I’m currently living where you have to actually “guess” what the other traffic participants will do next: is that kid going against the flow of traffic on his bike? (answer: yes) Is that cab going to cut me off in the middle of the intersection and should I take my foot off the acceleration pedal so that we won’t crash? (answer: yes to both questions). Are those pedestrians going to jaywalk without checking for any incoming vehicles? (answer: yes). All these things happened to me during my short city drive half an hour ago.

leesec2• 8 years ago

Waymo has the best tech for sure and even they are far off from full autonomy. Full autonomy in Phoenix, sure, they're the closest, but not full global autonomy.

The other issue being those cars prolly cost somewhere in the $500,000 range at the moment. Not exactly a consumer product.

Check comma.ai if you're looking for cheap, accessible, self-driving capabilities (still a long ways off, but you can get hands free today on 15+ cars for under 1k in plenty of common situations).

nolemurs• 8 years ago

I've had the same thought. Where is the skepticism coming from?

I've noticed a lot of articles lately that talk about how far off self-driving cars are. That said the articles all seem to rest on one of two arguments. Some (like this article) point to the failure of some startup that most sane people didn't have any faith in. The others point to the small handful of accidents we've had: the one from Uber (which was the result of borderline criminal irresponsibility), and the couple from Tesla (which were the result of drivers not paying attention or and with their hands off the wheels of their emphatically not-self-driving cars). These sheer number of these articles creates the impression that there's a lot more going bad in the field than I think there's any evidence for.

When I try to dig deeper and figure out why these articles are all appearing I can't find anything.

Probably it's just that after all the stories about the amazing future of self-driving cars, this is the next natural story to tell since we don't actually have the self-driving cars yet, but there's definitely a paranoid part of me that thinks some industry group is sponsoring these stories for some reason.

ModernMech• 8 years ago

This is kind of what I thought when I was thinking about entering the space. Especially when I started looking around and interviewing at different places, they all seemed to be doing the same thing in the same way. They were just repeating the successes of the DARPA Urban challenge, and making the systems more polished and capable. They made impressive demos and then got bought up for impressive amounts. But like most things, this follows the 80/20 rule, so most of the work is still ahead of us. We're probably 80% of the way there, but that last 20% is going to take decades and probably some fundamental breakthroughs in sensors and AI. It's going to take a while.

It just seems like every automaker heard that autonomous is the next thing, so they went out shopping for autonomous startups without a clear sense of what they were after and how hard it was going to be to get there. I think a lot of them are going to be feeling buyer's remorse.

heurist2• 8 years ago

I've been enjoying Chevrolet's "adaptive cruise control" and "lane-keep assist". They both have a lot of room for improvement but are a good step into progressive enhancement, as you put it. I've been imagining different tools that we might start seeing in cars as this kind of product develops: adaptive high beams, rear view and side mirrors that auto-adjust to your gaze, automatic windshield wipers connected to a weather API and an internal camera system that measures visual noise from the windshield, a radar view in the dash so you know where the surrounding cars are... lots of little opportunities. Make the human driver redundant piece by piece.

briandear3• 8 years ago

Adaptive high beams are on Audi now, they are awesome. Weather API connected wipers? Bad idea, rain sensing already works, why use an API when you never turn on wipers before it rains anyway? Radar view for surrounding cars might be a nice gimmick but other than blind spot detection or parking assist, I’m not sure what benefit a radar display would have, unless it’s for parking, and an equivalent exists (Audi for example has 360 degree parking cameras and some cars have full autonomous parking already.)

But the fly-by-wire concept could be a good one — if reliability is ultra high.

Animats• 8 years ago

Oh, Quanergy, the vaporware LIDAR company. "Quanergy is the leading provider of solid state LiDAR sensors and smart sensing solutions." They announced a flash LIDAR in 2016, but don't seem to have shipped the product.

I expect that Continental, the German auto parts company, will be the winner here. They bought Advanced Scientific Concepts, which makes good but expensive flash LIDAR units for military and space applications. Now they have to make them cheaper and in volume, which is what auto parts companies do. They already have some shorter-range, lower resolution LIDAR products on the market for anti-collision systems, so they have experience with the automotive LIDAR environment.

Delco, the US auto parts maker, is also working on self-driving components, including LIDAR. They're not saying much, but they have a California autonomous vehicle testing license and some cars on the road. Bosch also has a self driving car in test. It's entirely possible that self-driving will just be an auto part, sourced from the usual big auto parts makers.

mikestew• 8 years ago

It's entirely possible that self-driving will just be an auto part, sourced from the usual big auto parts makers.

I've always assumed that this is the eventual outcome, even if there is a brief period where SelphDrivin, LLC, et. al., sell their own cars. For decades Bosch would sell a whole fuel injection and ignition system for your new model. Brembo and Bendix will sell you some brakes. I don't know who makes ABS sensors and computers, but I'd guess it's not GM (probably Delco, though) and Ford. MobileEye's lane assist was an option on the Mercedes Sprinter we bought. The next logical step in my mind is self-driving systems.

azernik2• 8 years ago

Speaking of MobileEye, from what I understand their goal is to slowly work up their parts offerings from driver assist to self-driving.

brenschluss2• 8 years ago

This makes sense from a branding / insurance / shiriking responsibility perspective also.

“Our newest Toyota uses a state of the art NeverHandsTM system!”

After a few bad car crashes: “We at Toyota have decided to discontinue our relationship with NeverHands following the NTSB review of their product.”

rossdavidh• 8 years ago

I'm not saying this is as extreme a case, but I got a whiff of similarity to Theranos, in the aspect of an emphasis on loyalty making it difficult for the CEO to listen to debate. Of course, that could also be because the journalists covering tech businesses nowadays are looking for the next Theranos story.

kovrik• 8 years ago

Yes. Looks like it is becoming a common thing with all that startup hype.

rhizome2• 8 years ago

And Theranos was kept alive for too long by a lot of symbolic value.

xvector• 8 years ago

In Eldada’s vision, lidar sensors could be deployed along the border, monitoring remote regions for movement, then determining whether there were people attempting unauthorized crossings [...] Eldada acknowledged that a contract for a virtual wall would be controversial, but he described it as a kind of protest against Trump’s version of border security. “I hate the word ‘wall’. We are anti-wall. We have a technology solution,” he said.

This physically nauseates me. Where are our ethics? Just because we can engineer something does not mean we should.

It makes me glad that this company seems to be nothing but vaporware.

notfromhere• 8 years ago

i dont think the arguments about a border wall were ever about morality, but more about it being a huge waste of money

zaroth2• 8 years ago

Would the border wall be a huge waste of money?

It’s a weird game where the stricter you enforce the less people try to come in the first place. So it’s all about signaling.

And if a wall is to be built, should it not be a virtual wall to reduce cost and improve the likelihood that one day it can be merely switched off?

Wouldn’t broad sensors of this type have massive secondary applications?

nafizh• 8 years ago

“I hate the word ‘wall’. We are anti-wall. We have a technology solution,” - Oh, the amount of self-unconsciousness there.

mgaffneyny• 8 years ago

pony.ai

How a billion dollar autonomous vehicle startup lost its way

Related Stories