Live data from Hacker News

Cruise is opening driverless cars to the public in San Francisco

getcruise.com

121–130 of 652 posts

Re: Cruise is opening driverless cars to the public in San Francisco

#121
post #65
post #49

Earlier quoted context omitted.

They are wholly owned by GM are they not? Starting this way doesn’t preclude them taking Tesla’s vision only approach in the future. Even Tesla initially had a radar vision combination approach before moving to pure vision. The real question with whatever technique being used is can they drive the hardware cost low enough that it can be widely deployed.

They moved to pure vision because they were constrained by radar suppliers because of supply chain issues and the chip shortage, not because of any ML progress - they were actually investigating using a higher resolution imaging radar before the pandemic.

They also have Lidar cars in Fremont that drive around every so often, that doesn't mean they plan to go into every car anytime soon. It'd be short-sighted to not continuously evaluate solutions that previously had constraints (despite what Elon says, they'd add Lidar if it made economic sense and it showed an improvement over camera-only, as their camera detection is pretty accurate these days in FSD).

Re: Cruise is opening driverless cars to the public in San Francisco

#123

Earlier quoted context omitted.

At least in the UK, one of the most prolific serial rapists was a licensed taxi driver (not even an Uber driver (who get a lot of stick about not being as safe as "real" taxis) - but a licensed and supposedly-vetted London black taxi driver). https://en.m.wikipedia.org/wiki/John_Worboys So I guess safety is a major plus based on that.

Totally illogical supposition. I'll bet a lot more taxi drivers have rescued people in trouble than assaulted them, a young female relative of mine was rescued by a passing taxi driver from a dicey situation outside Sheperds Bush tube station. The chances of thugs standing in front of future driverless vehicles like highway robbers to stop them is very high, especially if the predators can see their prey has somethin…

The scenario you described already happens in San Francisco and the Bay Area on a regular basis to regular human drivers. There are plenty of videos of peoples' back window being smashed while they're sitting in traffic and their expensive goods are taken. The solution to this is not having human drivers, it's cracking down on property crime and treating it as equally as important to violent crime.

Re: Cruise is opening driverless cars to the public in San Francisco

#124
post #66

Earlier quoted context omitted.

18+ is reasonable since it's still a dangerous experiment. I actually would love an automated version of this that goes at 200 or 300 miles an hour, the ultimate thrill ride. Have it run on something like the Autobahn. While it's impossible for a human being to drive at that speed, computers definitely could . It would be like sky diving but on land

Why not go on a train?

I love it when the self driving threads always circle back to reinventing the concept of public transportation as if it doesn't already exist. Want to know what the best hyperloop system in the world is called? The subway, invented 200 years ago.

Re: Cruise is opening driverless cars to the public in San Francisco

#125
post #49

Earlier quoted context omitted.

They are wholly owned by GM are they not? Starting this way doesn’t preclude them taking Tesla’s vision only approach in the future. Even Tesla initially had a radar vision combination approach before moving to pure vision. The real question with whatever technique being used is can they drive the hardware cost low enough that it can be widely deployed.

The reason camera only won’t happen is because it won’t work. Elon has been fighting reality for years.

How does a human identify obstacles, VRUs [0], and other cars?

0: vulnerable road users, eg pedestrians and bicyclists

Re: Cruise is opening driverless cars to the public in San Francisco

#126
post #103

Earlier quoted context omitted.

They will love it until they are leaving from the bar and the last bar attendee left them a big wet bile-smelling present all over the seat. At least when this sort of thing happens on a bus there are other seats you could use.

Reject the ride due to cleanliness and request another one.

And who knows how long that will take? In some places it takes long enough just to request a single uber trip much less two in a row. Imagine how livid you'd get when a half hour after you first intended to leave the bar, the second self driving car arrives, and it to is soiled.

Re: Cruise is opening driverless cars to the public in San Francisco

#127
post #93
post #8

Based on this image: https://images.ctfassets.net/95kuvdv8zn1v/6h1C7lPC79OLOlddEE... They and their VC backers are clearly betting on the concept that radars + lidar + imaging will be the ultimate successful solution in full self driving cars, as a completely opposite design and engineering philosophy from Tesla attempting to do "full self driving" with camera sensors and categorical rejection of lidar. It is interes…

Cruise CEO here. Our strategy has been to solve the challenges needed to operate driverless robotaxis on a well-equipped vehicle, then aggressively drive the cost down. Many OEMs are doing this in reverse order. They're trying to squeeze orders of magnitude of performance gains out of really low-cost hardware. Today it's unclear what strategy will win. In a few years our next generation of low-cost compute and sensin…

My conclusion from years of self-driving, LIDAR, etc. research is that managing medium to heavy precipitation reliably might be impossible.

Visual algorithms run into the same problem as human brains, and the size of e.g. rain drops interferes with the frequencies employed by radio techniques.

Is anyone aware of any strategies that give us hope in solving this problem?

Re: Cruise is opening driverless cars to the public in San Francisco

#128
The next thing is a driverless car and a driverless bus “docking” at speed, so you can change vehicle without affecting your trip time.

Economics will outcompete 50 self-driving cars going in the same direction for 30min when you could have one driverless-bus doing the same.

If this transition is smooth and fast, think what this will do to traveling time, car prices and property prices. How far can you travel without stops and no traffic jam?

Re: Cruise is opening driverless cars to the public in San Francisco

#129

When I was in Vegas last I called for a "driverless car" via Lyft but instead of an empty car a male driver with a second man riding shotgun showed up to pick me up. The entire scenario made me feel sketched out so I refused to take the ride because I've seen so many news reports of people getting assaulted by drivers. I and many of my friends would feel more comfortable with the option of a driverless ride so I defi…

No post body was provided.

Re: Cruise is opening driverless cars to the public in San Francisco

#130
post #65

Earlier quoted context omitted.

They moved to pure vision because they were constrained by radar suppliers because of supply chain issues and the chip shortage, not because of any ML progress - they were actually investigating using a higher resolution imaging radar before the pandemic.

They also have Lidar cars in Fremont that drive around every so often, that doesn't mean they plan to go into every car anytime soon. It'd be short-sighted to not continuously evaluate solutions that previously had constraints (despite what Elon says, they'd add Lidar if it made economic sense and it showed an improvement over camera-only, as their camera detection is pretty accurate these days in FSD).

I’ve seen their LIDAR cars in SF too - if I had to guess they’re gathering ground truth data to train monocular depth models on.

And even really naive integrations of LIDAR will show big improvements over camera only. You can do something as simple as overlay the returns from the most recent LIDAR spin over a camera image as a fourth channel and feed it into your models and most of your depth prediction/spatial predictions will improve dramatically.

Post reply on HN