They announced 1 million trips in August 2022, so 6 to 7 months for 1 million paid trips. At ~5300 trips per day they're well beyond the prototype stage (from a commercialization standpoint).
Last edited:
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Autonomous Car Progress
- Thread starter stopcrazypp
- Start date
-
- Tags
- Autonomous Vehicles
They announced 1 million trips in August 2022, so 6 to 7 months for 1 million paid trips. At ~5300 trips per day they're well beyond the prototype stage (from a commercialization standpoint).
EVNow
Well-Known Member
Why do you say that ? Looks good given the space. Most humans can't park there without scraping the car.
We are about 20 days before Waymo goes fully public in SF as they are waiting on approval that's due on May 11th.
Waymo had their own incident recently, although not anything as severe as Cruise:
Waymo says dense S.F. fog brought 5 vehicles to a halt on Balboa Terrace street
Of course disabling the vehicle for weather is totally acceptable for L4, but it does show that even if you have the most advanced sensors available that can detect objects through the fog, the driving task still requires the vehicle to "see" visually to safely handle the driving task.
Waymo says dense S.F. fog brought 5 vehicles to a halt on Balboa Terrace street
Of course disabling the vehicle for weather is totally acceptable for L4, but it does show that even if you have the most advanced sensors available that can detect objects through the fog, the driving task still requires the vehicle to "see" visually to safely handle the driving task.
Last edited:
Let’s keep in mind it’s early days for driverless operation. Which means the goal isn’t to make your system operate as close to its tolerance. You’re are not trying to get as close to the sun as possible before getting burned.
I’m sure the Waymo Driver can operate in these conditions and even worse. It’s similar to when Waymo was pulling over at the sight of very light rain. Which made rounds in the Tesla community as they gain-said.
Yet now they handle moderate and heavy rain flawlessly.
diplomat33
Average guy who loves autonomous vehicles
Interesting video. We see Cruise AV apparently struggling at an intersection. But we see it from the perspective of the Waymo behind them so we get a pretty good view. It looks like traffic was annoyed by the 2 Cruise AVs.
Looks to me like the one Cruise wanted to turn right but thought the path was blocked by the postal truck and so it just stopped in the intersection waiting fo things to change. This caused the other Cruise to be temporarily stuck but it eventually goes around. If so, it could be a hint to why Cruise has had more "stalls". It would seem the Cruise planning is just not advanced enough if a common case like a double parked postal truck causes it to just stop in the intersection.
Looks to me like the one Cruise wanted to turn right but thought the path was blocked by the postal truck and so it just stopped in the intersection waiting fo things to change. This caused the other Cruise to be temporarily stuck but it eventually goes around. If so, it could be a hint to why Cruise has had more "stalls". It would seem the Cruise planning is just not advanced enough if a common case like a double parked postal truck causes it to just stop in the intersection.
Last edited:
Yes, it was very interesting.
I also noted the Waymo's wipers going all the time. Seems that Tesla isn't the only AV with this problem.
diplomat33
Average guy who loves autonomous vehicles
Actually, I think that is on purpose. Waymo uses the wipers to "wave" at other AVs. Maya actually mentions that the Waymo will wave at the Cruise.
diplomat33
Average guy who loves autonomous vehicles
Well, the "no accidents on FSD beta" argument is now gone.
To hit Elon's stated goal of 10x safer than a human, they need to get down to 0.15 accidents per 1m miles without supervision. FSD beta drivers are supposed to supervise at all times.
So, if the 0.31/1m is correct, they need to cut that in half while taking away the need for driver supervision.
Who has been arguing that there’s no accidents on FSD beta? That would be insane since you can be victim to an accident caused by someone else. No system is ever going to deliver zero accidents.
I haven’t seen the small print on these stats, but if they’re like their previous stats then these figures include any accident within x (5?) seconds of the assistance system being active, so removing the need for driver supervision shouldn’t have any direct negative effect on the accident rate. Expanding the scope of the situations FSD is happy to attempt to handle may do, if (eg) U turns are particularly accident prone.
diplomat33
Average guy who loves autonomous vehicles
That is going to be the tricky part. We have no idea what the accident rate without supervision is. The accident rate with driver supervision is really meaningless. It's the accident rate without supervision that matters if you want to remove driver supervision.
The other issue is that Tesla owners might only be using FSD beta in certain situations where it is safer and not be using FSD beta in situations that are less safe. I know I often disengage FSD beta in certain situations like an unprotected left at a busy intersection because I don't trust it and I don't want to take the chance of a mishap. If users are only using FSD beta in low risk situations, that will bias the accident rate and not tell you the true accident rate without supervision.
I don't think it's completely meaningless, as the government and industry certainly still cares a lot about L2 accidents (which is why they introduced the reporting requirements). A long running criticism of AP stats is that it's used mostly on the highway, so people say that skews the per mile accident stats. FSD Beta in its current implementation is used mostly in the city, so that provides a more diverse road type.
It could be that it makes people pay more attention or be extra careful, but that is a good thing for safety. I've seen it been argued AP makes people more likely to be distracted or stop paying attention and thus is less safe than driving with it off, but without stats broken down by road type, people keep arguing the AP stats don't show it to be safer overall, just that it's a side effect of the road type it's typically used in.
diplomat33
Average guy who loves autonomous vehicles
In context, I meant it is meaningless for removing driver supervision (ie L4 autonomy). It is obviously not meaningless as a stat for the safety of a L2 driver assist system.
Having said that I still think the stat is very misleading. Tesla only measures accidents with airbag deployment or other active restraint employed. So Tesla is only measuring serious crashes. But they measure that accident rate against the overall industry average which includes all police reported accidents and vehicles of different years, some with ADAS and some without. So it is not an apples to apples comparison. Additionally, city driving is at lower speeds so we would expect less of a chance of a crash involving airbag or active restraint deployment, especially if the driver is attentive and taking over before FSD gets into trouble or not using FSD Beta in risky situations in the first place. So the stat is very stacked IMO.
Really, the only thing that the stat says is that an advanced driver assist with an attentive driver will have fewer serious crashes than the average of all accidents. Well, duh, that is not surprising.
diplomat33
Average guy who loves autonomous vehicles
"Cruise continues to operate in the red, losing $561 million in the first quarter of 2023."
"The company earned just $30 million in revenue for GM, though most of that money comes from interest and other non-operating sources. Cruise, which operates robotaxis in San Francisco, Phoenix, and Austin, says it remains on track to hit $1 billion in revenue by 2025 and $50 billion by 2030."
"Cruise CEO Kyle Vogt said the company expanded its driverless fleet size by 86 percent, from 130 to 242 “concurrently operating AVs.” The vehicles have since passed the 1.5 million miles mark and regularly conduct 1,000 driverless trips with passengers each day, Vogt added."
Source: Cruise continues to burn GM’s cash as robotaxis expand to daylight hours
"The company earned just $30 million in revenue for GM, though most of that money comes from interest and other non-operating sources. Cruise, which operates robotaxis in San Francisco, Phoenix, and Austin, says it remains on track to hit $1 billion in revenue by 2025 and $50 billion by 2030."
"Cruise CEO Kyle Vogt said the company expanded its driverless fleet size by 86 percent, from 130 to 242 “concurrently operating AVs.” The vehicles have since passed the 1.5 million miles mark and regularly conduct 1,000 driverless trips with passengers each day, Vogt added."
Source: Cruise continues to burn GM’s cash as robotaxis expand to daylight hours
Similar threads
- Article
