Waymo's car safety systems engineered in from the ground up bode well for their reputation and also distinguish their system from other vehicles with self-driving "tacked-on".
The latest statistics suggest that Waymo's car requires about an order of magnitude fewer interventions from safety drivers. If these systems are a key reason for that or will further boost their safety records, it will be a long while before competitors can catch up.
"Waymo says it has done extensive work to make sure that computer crashes don't lead to car crashes. All of the key systems on its cars—the computer, brakes, steering systems, and batteries—have backups ready to take over if the main system fails."
"Safety-critical aspects of Waymo's vehicles—e.g. steering, braking, controllers—are isolated from outside communication," Waymo writes. "For example, both the safety-critical computing that determines vehicle movements and the onboard 3D maps are shielded from, and inaccessible from, the vehicle's wireless connections and systems."
"Waymo's car requires about an order of magnitude fewer interventions from safety drivers."
I'm a little skeptical about using interventions per x-distance as a metric for comparison.
There's just too many different things that could skew that. Like driving the same routes over and over. Or more/less stringent rules about when to trigger intervention. Or choosing when not to drive, like poor weather. Or purposely choosing difficult scenarios and edge cases for your routes.
Wow, finally someone taking vehicle system security seriously! I always find it terrifying that most car manufacturers seem to have no problem with tying safety-critical systems in with the entertainment system and pushing OTA updates to them.
> are shielded from, and inaccessible from, the vehicle's wireless connections and systems."
Does this mean totally isolated with no connection or generally inaccessible but only really at the software level and if you overflow the right buffer with the right bytes you might be able to connect.
But what do they look at? If the failure mode for example is that the steering signal always signals half a degree more to the left in left turns than necessary, would the watchdog see that and takeover? How does it know that IT'S sensing mechanism isn't broken?
What does it mean to "tack-on" self driving? Did Waymo engineer their physical vehicle differently from other companies in some way that helps self driving? I don't think they ever claimed this.
Link to actual report.[1] (When you're posting a link to something important, post the link to the real document, not some pundit blithering about it. Yes, you may have to do some digging to find it when the pundit outlet doesn't provide a link.)
There's not much new here if you've been following what Waymo is doing. Chris Urmson's SXSW talk from 2016 [2] covered most of this material in more detail. This is more of a PR-level version of that content. The new material covers mainly the limitations Waymo is imposing on their system. It works only where they have very detailed maps of the road system, with more information than even StreetView. This is conservative, but not a bad decision for initial release.
They have a button to call customer service. (That's so unlike Google.)
OP linked to a news article that presented a pretty good summary of the report along with some additional context. It was not "some pundit blithering about it."
I don't think there is anything wrong with linking to a news article like this that is published by a reputable news source such as Ars Technica. Not everyone is going to have the time to read a 43 page report, especially if it's not clear why they should be reading it.
To make web sites "sticky", so people don't go elsewhere.
That doesn't apply to YC, so when posting here, please give the link to the original source. Sometimes it's necessary to dig through three levels of blog to get to the source. (This seems to be especially true of articles about battery chemistry and surface chemistry ("nanotechnology").)
It's certainly smart to start this as a ride-sharing service rather than DTC ownership model. By being able to choose which trips it is confident in taking, it removes a lot of the remaining question marks that would come up by a car owner trying to take it anywhere.
The question is: can this be a viable commercial product if it doesn’t actually do what people expect from a taxi service? And how extensive is the deployment going to be that people will choose to use Waymo over just driving? If I have to plan ahead or wait 15 minutes for a car to arrive, I’m a lot less likely to even try the service.
If they do it right, the user doesn't have to care. You call a car, and either a self-driving or a non-self-driving one will show up, depending on the trip you selected.
But that's for the future; for now I'm pretty sure every car will have a driver, even if just as a backup/supervisor.
They've had cars driving all over the place for years. If they've got the driving part "done", the other parts of it are about product stuff; how to request rides, in-ride experience, etc. They definitely would've had that all happening in parallel with refining the autonomy tech. I don't have any information, but it seems plausible to me that it could be 2018.
They have mostly been driving in sunny areas with very nice weather. Driving with strong weather hasn’t really been tackled by anybody yet. Additionally the legal side is still not (officially) resolved.
Waymo says its cars can handle nighttime driving and light rain, but its cars will be geofenced
Nice weather driving is predicable out ~6 hours and easily 98+% of driving conditions nationwide. If the system simply checks a weather service before driving and says no if there is problems with the route that's fine for version 1.0
There are recent rumors they'll be launching in Phoenix soon. That they're releasing details like this at about the same time as the rumors is a promising indicator those rumors may be right.
There’s a big difference between running a project like this in “R&D mode” vs “full on production mode” and I mean that from every angle: hardware, software, operations, legal, etc. So I’m skeptical on a 2018 delivery.
Call me cynical but this looks more like Google gave Waymo a deadline and now they’re trying to ship a MVP to keep the funds coming.
Google doesn't do deadlines. They also don't announce until it's ready. It's deeply ingrained in their culture. They dabbled with a deadline a couple times and the results were a disaster. I know technically waymo is a different company but it incubated in Google itself. My experience with how Google operates runs counter to your experience.
This is all anecdotal on both of our parts of course but I wouldn't bet against Waymo on this. They have what looks like a massive head start against the competition here.
Sadly we do deadlines now, depending on the product. Before Google got into the world of hardware, that may have been true. But consumer hardware requires deadlines to be successful.
> "Passengers cannot select a destination outside of our approved geography, and our software will not create a route that travels outside of a geo-fenced area, which has been mapped in detail," Waymo says. The Information reported earlier this month that Waymo's cars will initially be made available in the area of Chandler, a Phoenix suburb.
I saw one of their minivans a week or two ago. Chandler is mostly "new" construction - New strip malls, newish neighborhoods with nice sidewalks, etc. It's also relatively well-to-do. Average incomes are much higher than in Mesa, which is just to the north.
Phoenix (other side of the Salt River) has nice areas, and it has ghettos. I wonder where the geofence's lines are drawn. Certainly they'd include ASU/Tempe...
I wonder how these would work for passengers who need to get their illegal plant products, dealers of which don't always live in neighborhoods with nice sidewalks. Mostly those passengers didn't tell me what they were actually doing, but sometimes I figured it out.
From this report I get the sense that Tesla, Cruise, et. al. are years behind compared to Waymo. Can anyone that has more knowledge comment on the state of self-driving companies today? Ex. The level of failure redundancy they highlight I haven't seen by Tesla which promises the current Model S's etc they sell have full autonomy capability. I may be missing something though...
This article links to another one of their articles that talks about another marketing step waymo took in the last week, the partnership with a number of organizations to help shoe where self driving cars can help various people in our society.
+ Jealous cap drivers, truck drivers, neighbors can easily put some chewing gum or paint on one or a few sensors. The system will be too brittle to cope with that, so all the time your/the sweet self-driving car doesn't even want to start.
+ Joyriding and pranks. Put your silly little brother or the dog who you've to babysit in the car. Bye bye!
+ Criminality. Just put your drugs on board or a corpse. Or do the companies that own these cars use them as remote cameras all the time?
+ Insurance. Just set up a situation perhaps with multiple cars where the self-driving car slams in your insured car. Use some doll etc.
+ Destruction. Do the same to cause the car falling into the river or crashing into something you dislike.
+ Vandalism. Graffiti etc. will be recognized much later than with a human driver.
+ Theft of sensors or other car parts can be orchestrated on a location the car first drives to.
Most of these things with self driving cars already apply to normal cars. Also with self driving cars having tons if sensors and Internet connectivity you should have a good scan of whoever messes with one and quick police response if anything goes wrong. For stuff like bogus insurance claims they could just add a few internal sensors in the car or something to make sure a real person is in there. None of these problems really seem all that insurmountable
To argue that the legal context is exactly the same for self-driving cars as for normal cars seems unjustified.
I'm not arguing that it is insurmountable. I'm just saying that it's too early. Most of these things have not been worked out, which seems unnecessary.
I think Tesla idea of tunnels with self driving cars is more competent then those autonomous vehicles driving on roads with other cars. At least they got some data from their cars with real world usage not predictions.
Waymo has hundreds of vehicles logging millions of self driving miles on the road. They have plenty of data.
Tesla's "tunnels": I assume you are referring to elon's Boring Company? How many decades of tunneling do you think that will require before it even covers a fraction of routes people want to take?
> Passengers cannot select a destination outside of our approved geography, and our software will not create a route that travels outside of a geo-fenced area
If all our movements will be subject to continuous monitoring and approval, then SDC's won't be anything better than taxis or planes.
monitoring is one thing, but this particular line seems to me like a safety measure. If you haven't trained in a particular geography, initially, you might want to be a bit more cautious.
disclaimer: google employee, but no knowledge of waymo things.
Part of me admires their sober approach to safety, but another part is extremely skeptical of the implied reliance on mapping. Reality changes all the time, so your technology should better be able to deal with outdated models. But if it can, shouldn't it be able to cope with an absent model just as well? If it does need a pre-made model of reality, I am not convinced of its ability to deal with unexpected change. In sports climbing terms (hope I use them correctly, not being one): how bad is the car at onsighting a road that it needs beta so desperately to flash it?
Maybe their "onsight" capability is just very slow and the first Waymo to meet a change has to onsight it at hardly more than roomba speed (or worse: mturk?), sharing the knowledge ("beta") with the Waymos that come later, but then you still have to reliably detect that your model is outdated while driving regular speed while all false positives would considerably harm the relationship with other road users.
I think it is a cautious approach. Mapping is one aspect, but there will always vehicles and pedestrians that the vehicle has to plan for.
This reminds me darpa grand challanhe and urban challenge difference. The cars that were built for the desert wouldnt work great in an urban environment.
Come back to me when it can handle fog, snow, heavy rain, and areas it hasn’t seen before. Until then it’s just overgrown version of driver assistance.
The latest statistics suggest that Waymo's car requires about an order of magnitude fewer interventions from safety drivers. If these systems are a key reason for that or will further boost their safety records, it will be a long while before competitors can catch up.
"Waymo says it has done extensive work to make sure that computer crashes don't lead to car crashes. All of the key systems on its cars—the computer, brakes, steering systems, and batteries—have backups ready to take over if the main system fails."
"Safety-critical aspects of Waymo's vehicles—e.g. steering, braking, controllers—are isolated from outside communication," Waymo writes. "For example, both the safety-critical computing that determines vehicle movements and the onboard 3D maps are shielded from, and inaccessible from, the vehicle's wireless connections and systems."