CA DMV suspends Cruise deployment and driverless testing permits
email.dmvonline.ca.gov
email.dmvonline.ca.gov
I'd be very curious to see the terms of reinstatement; we've all seen the viral clickbait taxis stopping traffic in SF, but I have no idea of the actual stats as to safety.
The California PUC now requires quarterly reports that are published at https://www.cpuc.ca.gov/regulatory-services/licensing/transp...
San Francisco's city/county transportation agencies, police, and fire departments have requested that additional information be tracked about crashes involving autonomous vehicles. But my understanding is that CA DMV and CPUC have denied those requests. There are currently some academic researchers filling this void with their own recordkeeping and analysis. Each time academics circulate a pre-print for peer review, Cruise appears to issue a press release based on its own internal statistics and its CEO goes on the offensive.
So, to a certain degree, it's unsteady and loose regular by CA DMV and CPUC that have led to the former needing to suspend Cruise's permit today (rather than a much more gradual and controlled process of granting industry more leeway to test without safety drivers on public right-of-way).
Also, OpenAI says hi!
A youth side by side I guess.
The pervasive "we're just doing it in software" and "city legislature dark arts" narratives aren't confidence inspiring.
And there should be 100% transparency regarding accidents where they are involved.
Nobody should be able to sell any cars at all unless they can demonstrate decent headway in self-driving research.
Car manufacturers should be put on notice that the government expects 10% annual reduction in deaths on the roads every year for the next 20 years. If their car brand doesn't keep up with that, then they're gonna be fined $10M per excess death (for selling a dangerous product). They can achieve that any way they like, but a decently working self driving system is probably the only thing thats gonna let them continue to sell cars a decade from now.
Are you saying Tesla should stop calling it FSD? Or that regulators should prohibit certain driver assist features without more strict regulatory approval?
If Tesla were putting actually driverless (or cars without pedals and wheels) on the road, I'd be with you.
The FAA does a pretty good job for aircraft some errors notwithstanding, I don't see why there couldn't be a similar institution for road vehicles. Oh, wait: https://www.ntsb.gov/Pages/home.aspx
Kudos to Cruise for getting anything done in CA ever, look forward to riding these autotaxis in Dallas.
In one instance, as a Muni bus was making a left-hand turn onto a busy street, a Cruise vehicle pulled forward into the intersection (from the direction the bus was turning toward) blocking the bus from completing its turn. As bus drivers cannot—in general operation—go in reverse, the only solution to this was for the Cruise vehicle to back up or to exit the intersection to its right and allowing the intersection to become unwedged.
Unfortunately, without any real motivation to do otherwise, the Cruise vehicle was seemingly willing to wait patiently until the heat death of the universe for the problem to remedy itself. I watched for fifteen minutes as the intersection was completely wedged due to this situation, while the bus driver and others attempted unsuccessfully to convince the Cruise to do anything to fix the problem.
I eventually left. I have no idea how long it was before someone at Cruise managed to get the car to move, but it highlighted for me how many situations in traffic require impromptu and improvised coordination between drivers in order to keep things moving.
Just because you make your large industrial robots look like cars does not mean you should be able to test them in the street where the kids and old people (and everybody else) exist and do their thing. It's fucking insane (and I almost never swear online.)
We know full well that those drivers are one of the leading causes of death in the country, but we shrug it off.
With the press watching self-driving companies like hawks waiting to freak out over any mistake (OMG, self-driving cars caused a traffic jam in SF! Better tell everybody in the world!) I feel okay about having them on the road.
Item: humans driving huge machines on every street is and always has been totally insane.
The reason we do it is because of a deliberate domestic propaganda campaign. No really. See: "The Real Reason Jaywalking Is A Crime (Adam Ruins Everything)" https://www.youtube.com/watch?v=vxopfjXkArM
You can see the origin of the situation from this extraordinary film shot just before the 1906 Earthquake: "San Francisco, a Trip down Market Street, April 14, 1906" https://www.youtube.com/watch?v=VO_1AdYRGW8
You can see that horses, horse-drawn wagon and carts, and pedestrians could and did mix together in the public way. There are also early automobiles, and the problem is obvious: cars accelerate faster than horses. Already we can see reckless driving by cars that is only possible due to their ability to accelerate quickly!
BTW, just for the record, cars themselves are generally awesome. It must be said. Cars are cool.
It's only when we use them as mass-transportation and/or mix them with non-car-driving people (bikes, pedestrians, etc.) that it becomes mad.
- - - -
Ergo, safe automatic automobiles can't come soon enough!
- - - -
However, none of the above justifies using random people without their consent as test dummies in an open-air testing facility for massive, fast robots.
That's also insane.
- - - -
To me it seems obvious that you do two things:
1) Start with slow light robots! Build a nerf foam golf cart that is limited to 5mph and get that working reliably. Going straight for Knight Industries Two Thousand right out of the gate is a fetish.
2) Build a demo city and populate it with people who have signed a release. In fact there's already a test city out in the desert somewhere for to test IoT and smart city stuff. Pay some people to live there and walk around and risk getting creamed by your robots there.
For you the rider. Everyone else in this thread who has to deal with these vehicles in traffic is saying otherwise.
If these things are perfect, they’d save about thirty lives per year in SF. That’s not nothing, but it’s not some unparalleled achievement either. Particularly when the cost is putting a large and vulnerable set of low-income people entirely out of work with nothing on the horizon to replace those jobs.
Is it the future? Sure. Is it the future we want? I’m unconvinced.
Like this:
https://www.sfchronicle.com/sf/article/cruise-muni-sf-storms...
Cruise likes to bloodlessly recite stats about human driver fatalities in response. Which is the sort of bullshit non sequitur you get from a PR consultant, who gets away with it because we have stenographers in place of journalists.