A 2030 Self-Driving Car Bet

Re:A 2030 Self-Driving Car Bet by Animats on Friday March 04, 2022 @03:23PM (#30559990)

We need solid Level 3 self driving, at least on freeways. The commercial products are stuck at level 2. None of those systems have achieved "does not run into stuff". That's not good enough. People just don't pay close enough attention when the system has control.

Waymo and Cruise have level 4, since they already have driverless taxis in use. Only in some conditions, true, but in actual use. Waymo just got approval to start charging fares, so they're out of experimental mode.

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by nradov on Friday March 04, 2022 @03:48PM (#30560323)

GM Ultra Cruise will supposedly allow real self driving on some freeways, at least in certain conditions. It's not available for consumers yet, due to launch next year.

https://media.gm.com/media/us/en/gm/news.detail.html/content...

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by ilaksh on Friday March 04, 2022 @03:26PM (#30560028)

Are Waymo and Cruise not commercial products?

I think they work in most conditions. How someone can look at Cruise and Waymo already in operation today and think they won't be generally available in 7 years is amazing to me.

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by msandford on Friday March 04, 2022 @03:30PM (#30560105)

There's a huge difference between "works fine 95% of the time in a place with very cooperative weather" and "works 99.99% of the time no matter the weather"

I mean there are times when human drivers are insufficient for the conditions.

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by fdgsdfogijq on Friday March 04, 2022 @03:44PM (#30560280)

Honestly I dont think there is that much difference. California is massive. Having self driving cars here will completely change society.

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by ClumsyPilot on Friday March 04, 2022 @04:32PM (#30560882)

I disagree with all three assertions above - the dififculty increases exponentially for the last few percentage points, just like energy required increases towards infinity as you approach the speed of light,

California is a small part of one country.

All we will get out of self-driving cars is more traffic jams, more consumerism, and less security in our lives because you might be locked out of essential service in case you didn't sign a lisence agreement or made an offensive tweet, or replied to a spammer

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by lhorie on Friday March 04, 2022 @04:08PM (#30560605)

California weather is very forgiving even if you account for places like Tahoe. AFAIK, none of the self driving companies have much experience with scenarios like New York in winter.

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by thetallstick on Friday March 04, 2022 @03:30PM (#30560096)

To be commercially available for passenger use in major cities at level 5, regulators have to also agree it's level 5. It's not just a technical problem.

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by ska on Friday March 04, 2022 @04:11PM (#30560644)

> Are Waymo and Cruise not commercial products?

Not really, and they aren't level 5. Essentially the bet is that getting to 5 is harder than people think, and getting to commercially available SAE5 is going to take (at least) longer than 7 years. Bear in mind all commercial systems now are level 2 and testing features for 3. Waymo/Cruise are testing 4.

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by carstenhag on Friday March 04, 2022 @04:05PM (#30560571)

Mercedes will release the EQS in the first half of the current year with the Level 3 "Drive Pilot". Works (afaik) on the entire german Autobahn.

German source, you will probably find an English one as well - even contains a video: https://www.golem.de/news/s-klasse-mit-level-3-mercedes-darf...

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by pvorb on Friday March 04, 2022 @05:25PM (#30561431)

Mercedes had a self-driving car prototype in the 80s and got it to drive at high speeds on the Autobahn in the 90s. [1]

So it took them only 30 years from protoype to product.

[1]: https://www.politico.eu/article/delf-driving-car-born-1986-e...

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by layer8 on Friday March 04, 2022 @04:34PM (#30560894)

Autobahn is a much, much lower bar than urban traffic though.

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by bhauer on Friday March 04, 2022 @04:51PM (#30561087)

Furthermore, although it's for use on the autobahn, it's limited to 37 MPH, further narrowing its use case.

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by carstenhag on Friday March 04, 2022 @05:46PM (#30561646)

People buying EQS' will very likely spend most of their time on the Autobahn, so it's a perfect fit.

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by Dylan16807 on Friday March 04, 2022 @04:23PM (#30560771)

"Does not run into stuff" can also be achieved at level 2, and we should demand that of all level 2 systems.

Level 3 would be nice, but it very much depends on how many seconds of warning it gives when it wants to disengage.

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by Tempest1981 on Friday March 04, 2022 @05:45PM (#30561637)

"Does not run into stuff" + "Does not phantom brake unexpectedly" -- that's the challenge.

[ Parent | on HN ]

Re:A 2030 Self-Driving Car Bet by Dylan16807 on Friday March 04, 2022 @06:17PM (#30561938)

Honestly I'll accept a good bit of phantom braking if it means I can drive less.

[ Parent | on HN ]