- cross-posted to:
- technology@lemmy.world
- cross-posted to:
- technology@lemmy.world
Two Cruise driverless taxis blocked an ambulance carrying a critically injured patient who later died at a hospital, a San Francisco Fire Department report said, in another incident involving self-driving cars in the city.
On Aug. 14, two Cruise autonomous vehicles were stopped in the right two lanes of a four-lane, one-way street in the SoMa neighborhood, where the victim was found, according to the department report. It said that a police vehicle in another lane had to be moved in order for the ambulance to leave.
Question is, are they a net positive?
They’re getting in less collisions. Autonomous vehicles in SF have only been at fault in one death. And it was a dog, and a safety driver was behind the wheel in the AV.
AVs are going have problems, but are those problems worse than the ones human drivers cause?
In order to save lives we need to study the bigger picture and not get hyper focused on individual tragedies.
In order to save lives, USA needs to get off car centric transportation. More cars is not the solution. Neither is automating them in urban and dense environments. AVs belong on the highways only.
We are trying to solve a problem with “tech” that has been a solved problem by other countries for decades. Netherlands is a great example of how to move people around efficiently without using cars as the primary mode of transportation. Amazing public transportation. Towns and cities designed around alternative forms of transportation such as walking, or biking. Infrastructure is cheaper to maintain since it lasts longer and is not constantly pounded on by multi ton vehicles.
I agree that the US actually needs more public transport. Fatalities aside, that’s often going to be the best solution for congestion and climate change. Congestion alone in SF is still a fucking nightmare. SF is small as fuck, but driving across that town between 3 and 7pm can take 1-2 hours.
As a local, I feel like the current state of MUNI, BART, CalTrans, AC Transit, and cycling are not going to be a good fit for EVERY single use case. If I’m injured, am carrying bulky stuff, or am trying to hit up a part of town that would take too long with public transport, an AV EV could be a good solution.
I usually try to avoid cars in SF. They’re often more trouble than they’re worth. But, there are times, IMHO, when cars solve a current route and use case better than alternative solutions. And it if they’re still going to be used for certain use cases, it would be nice if they killed fewer people.
Why not a car share instead? Or just an Uber?
Because the goal is to have an Uber that removes human error.
I’d say this article shows that error is still a problem.
It very much is, but the errors are different. An AV isn’t going to get distracted by their phone, by an argument, by rubbernecking, etc. But an AV might encounter something that the sensor AI is confused by, and the cars might Mitch McConnell themselves in the middle of the road. So far at-fault accidents are way down with the AVs, but stalls are way up.
Thank you, that at least is a good argument.
Combined with e-bikes to “flatten” hills and make distance traveling easier, we could really make some amazing improvements to city design.
Driverless cars could really help solve the “last mile” issue in many transit systems.
I dislike taking transit because I have to take one unpredictable bus from my house to the train, take the train the majority of the distance, then take another unpredictable bus to my destination.
The issue of infrequent buses through neighborhoods isn’t going to be solved anytime soon. But if I could take an electric driverless car from my house to the train I would be a lote more likely to take public transit over just taking my existing car.
You are not saving lives by blocking emergency services.
Thats not what hes asking. Humans are notoriously awful drivers. Does gradually replacing humans with AI drivers save more lives than unintentionally blocking an ambulance?
How many drivers cause this exact thing every week? You’re only hearing about this because it’s novel.
Right. Where are the corresponding articles about human drivers that have blocked ambulances due to ineptitude? I’m sure we won’t find them, but I’m sure it happens more often than we think.
This is just like reporting of Tesla crashes and fires by the media. It makes it seem like a big deal, but only because regular vehicle crashes and explosions are so frequent they just don’t get reported. I’ve personally seen several cars on fire on the side of the road, and just a few weeks ago saw a car rollover crash on the highway. None of these ever made the local news.
Removed by mod
That’s correct. But we don’t have the data. Musk, for example, won’t release it for Tesla and forced the NHSTA to redact it.
And the raw data is no good anyway. You have to compare autopilot systems with similar road situations (eg mostly highway, or established taxi zones) and similar drivers/cars (they’re not a random selection of all demographics and models).
It’s absolutely correct to say that we need to compare the new with the old, not simply present statistics in isolation. But we don’t have the data and it needs an established independent body to analyse it because the analysis is too easy to manipulate to leave in the hands of the companies that stand to profit.
Telsa isn’t part of this AV taxi pilot in SF. It’s Waymo (Google) annd Cruise (GM).
And one of the conditions for allowing robo taxis in SF is that the robo taxi companies have to share their accident / fatality data.