Live data from Hacker News

Driverless robotaxi fleet paralyzed for hours in San Francisco

thelastdriverlicenseholder.com

361–370 of 514 posts

Re: Driverless robotaxi fleet paralyzed for hours in San Francisco

#361

Earlier quoted context omitted.

You've nailed the dirty little secret of the AV industry. Everyone's safety model is underpinned by the ridiculous assumption that a human will be (a) reachable with very high reliability over shitty cellular networks; (b) able to correctly interpret complex situations with extremely limited information; (c) able to resolve safety-critical situations by making planner adjustments, without direct control over steering…

Wow. I had no idea that these things had such fundamental limitations. I'd feel a lot less safer using one now, to be honest. If you had to estimate, how long would it be before AVs would be able to do all their calculations locally? Is that something that's even under consideration at the time?

It's a well-known thing in AV circles that some companies were actually faking their route by hardcoding them for demos etc. Also one shortcut is to use HD map data, and relying on 100k+ LIDAR sensor data that will never will be cost-effective. I know of only a handful of companies that rely on 'cheap' sensor data (cameras/radar) and use real-time calculations.

Re: Driverless robotaxi fleet paralyzed for hours in San Francisco

#362
post #233

Earlier quoted context omitted.

Who says the self driving vehicles won’t be articulated buses?

Every company currently trying to sell a self driving car

Just because that’s the current market strategy doesn’t mean governments won’t hire those companies on contracts to run buses. It would be a lot cheaper than paying human drivers most likely, but even more than that it’s probably easier to build and maintain a bus and dedicated tunnel/lane than it is to build and maintain a railroad track.

Re: Driverless robotaxi fleet paralyzed for hours in San Francisco

#363
post #37

Yup, it's a mess. What it's not is, a huge multi-car multi-injury accident. So I don't mind it much more than if all the traffic lights in the area were out.

But it's not great if now we've moved denial of service attacks into our real-world transportation infrastructure. It's not too different from truckers in Ottawa blocking all the streets and causing chaos. If this can happen by accident, it seems likely that a future version could result from being hacked.

Re: Driverless robotaxi fleet paralyzed for hours in San Francisco

#364

Earlier quoted context omitted.

Disagree. It’s easy to see how far FSD has progressed, there’s lots of videos to document how far it’s gone from then and today. It’s also not easy to say they’ve reached the point of diminishing returns (which generally takes a massive amount of data to hit, even GPT3 has not hit it) as it’s not clear how much of FSD is powered by heuristic based code and what is NNs. Judging by some of the discussions there’s still…

> Disagree. It’s easy to see how far FSD has progressed, there’s lots of videos to document how far it’s gone from then and today. Yeah, but videos aren't objective metrics. For me, an objective metric is "went $FOO miles in $BAR conditions without needing any intervention". What would you consider an objective metric?

Actually, intervention is also not the best metric. The best metric is driverless miles. Removing the driver is the true test of autonomy.

Re: Driverless robotaxi fleet paralyzed for hours in San Francisco

#365
post #45

There is currently some kind of geomagnetic storm which is causing a lot of GPS tracks to be offset over the past few days. It's all over the running/cycling forums. Could be related, here's an example: https://old.reddit.com/r/Garmin/comments/vo17e2/the_magnetic...

It couldn't possibly be GPS, right? Networks are unreliable, the cars must be able to reach a safe state if the network drops out. They appear to have stopped in the road. If these cars can't safely navigate their way to the side of the road using local sensors, they are wildly, hilariously unsuited for autonomous operation. > at least a dozen autonomous Chevrolet Bolts from GM Cruise Automation were spotted blocking…

This is the difference between Tesla and competitors.

Tesla cars „can” drive you „safely” without connection to the network.

Some say that discard of lidar was a huge mistake by Elon, but he knows very well that this is not the way to go forward.

Re: Driverless robotaxi fleet paralyzed for hours in San Francisco

#366
post #365

Earlier quoted context omitted.

It couldn't possibly be GPS, right? Networks are unreliable, the cars must be able to reach a safe state if the network drops out. They appear to have stopped in the road. If these cars can't safely navigate their way to the side of the road using local sensors, they are wildly, hilariously unsuited for autonomous operation. > at least a dozen autonomous Chevrolet Bolts from GM Cruise Automation were spotted blocking…

This is the difference between Tesla and competitors. Tesla cars „can” drive you „safely” without connection to the network. Some say that discard of lidar was a huge mistake by Elon, but he knows very well that this is not the way to go forward.

What does lidar have to do with connection to the network? If anything, once GPS goes down, more sensing modalities can only be helpful.

Re: Driverless robotaxi fleet paralyzed for hours in San Francisco

#367

Earlier quoted context omitted.

Disagree. It’s easy to see how far FSD has progressed, there’s lots of videos to document how far it’s gone from then and today. It’s also not easy to say they’ve reached the point of diminishing returns (which generally takes a massive amount of data to hit, even GPT3 has not hit it) as it’s not clear how much of FSD is powered by heuristic based code and what is NNs. Judging by some of the discussions there’s still…

> Disagree. It’s easy to see how far FSD has progressed, there’s lots of videos to document how far it’s gone from then and today. Yeah, but videos aren't objective metrics. For me, an objective metric is "went $FOO miles in $BAR conditions without needing any intervention". What would you consider an objective metric?

> "went $FOO miles in $BAR conditions without needing any intervention"

But who decides whether intervention was necessary? This is still too vague.

Re: Driverless robotaxi fleet paralyzed for hours in San Francisco

#368
post #316

The promises of self driving cars have always seemed weird to me. Sure, I would love to have a world where all cars are self driving. The security benefits seem large. But decreasing congestion? We already know how to do that, and in many ways it seems like a safer bet than some kind of technogy with large problems and even larger unknowns; public transport. Cheap, well developed, reliable public transport will beat…

But you have to give up point to point, of course the overall throughput is better but the individual experience is (perceived) worse. This is why I love the idea of "public luxury" make sitting on a train or a bus nicer than sitting in a brand new mercedes.

That depends how you look at it, though. To my mind, the extra walking distance with public transport is generally a plus, since it serves as a decent baseline of physical activity. Even a small amount of added walking can have a significant health benefit. If you add carrying groceries and walking uphill to that the effect quickly adds up. If I had a car or (god forbid) a personal driver I'm willing to bet I'd be significantly heavier and in worse shape, and consequently all the more unhappy in general.

Re: Driverless robotaxi fleet paralyzed for hours in San Francisco

#369

Earlier quoted context omitted.

Wow. I had no idea that these things had such fundamental limitations. I'd feel a lot less safer using one now, to be honest. If you had to estimate, how long would it be before AVs would be able to do all their calculations locally? Is that something that's even under consideration at the time?

It's a well-known thing in AV circles that some companies were actually faking their route by hardcoding them for demos etc. Also one shortcut is to use HD map data, and relying on 100k+ LIDAR sensor data that will never will be cost-effective. I know of only a handful of companies that rely on 'cheap' sensor data (cameras/radar) and use real-time calculations.

If a full time driver earns 25k-50k that's sort of equivalent to every regular taxi and uber having a $500k / $1M driving package onboard already (at 4%), right?

I don't know the industry but very naively it doesn't seem to me like $100k really breaks the unit economics? Particularly if volume / learning curve can reduce the price?

Re: Driverless robotaxi fleet paralyzed for hours in San Francisco

#370
post #365

Earlier quoted context omitted.

This is the difference between Tesla and competitors. Tesla cars „can” drive you „safely” without connection to the network. Some say that discard of lidar was a huge mistake by Elon, but he knows very well that this is not the way to go forward.

What does lidar have to do with connection to the network? If anything, once GPS goes down, more sensing modalities can only be helpful.

Car should be able to drive without Lidar and connection to the network.

There is currently only a single car company that accomplished that.

Post reply on HN