Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Autonomous Car Progress

This site may earn commission on affiliate links.

This just confirms what I have always said that the Apple self-driving car was always vaporware. It was never serious. Project Titan was nothing more than a side project. Now it is confirmed with the fact that it has been delayed to 2028 and will only be L2 highway. So basically, Apple is planning to do basic autopilot in 2028 at the earliest and might get delayed again. By 2028, who knows how far Tesla FSD will be, Waymo will continue to scale driverless robotaxis to more places, Mobileye will likely have some eyes-off highway deployed on some consumer cars etc... Apple will be hopelessly behind.
 
This just confirms what I have always said that the Apple self-driving car was always vaporware. It was never serious. Project Titan was nothing more than a side project. Now it is confirmed with the fact that it has been delayed to 2028 and will only be L2 highway. So basically, Apple is planning to do basic autopilot in 2028 at the earliest and might get delayed again. By 2028, who knows how far Tesla FSD will be, Waymo will continue to scale driverless robotaxis to more places, Mobileye will likely have some eyes-off highway deployed on some consumer cars etc... Apple will be hopelessly behind.
I don't see Apple touting any compelling capability.
 
  • Like
Reactions: diplomat33
A November 2023 test so current FSD at that time.

Title: We Put Tesla FSD Beta Up Against Its Chinese Competition

I really appreciate that he gives a pretty in-depth comparison of the two systems. I don't think we've really had a good direct comparison of FSD beta with XNGP. It seems he ranks FSD beta better for highway driving but XNGP better for city driving and for driver monitoring. I do like that Xpeng requires drivers to watch their tutorial videos and take a test. I think many Tesla owners could benefit from that with FSD beta. This video also confirms for me that Xpeng could be a competitor to Tesla in terms of "self-driving" driver assist if Xpeng ever came to the US.
 
What happens when these driving systems age? Will the Government require recalls on these systems when they are 10 to 15 years old due to not working correctly and being dangerous on the Road? If a 10 year old car on level 3,4 or 5 crashes due to Sensor or Computer failure who is held responsible?
Presumably software itself will continue to be updated until the hardware platform becomes obsolete and is no longer supported, just like so many other devices today. Regarding "crashes" due to failure, the system should have various levels of diagnostics, including both startup pre-check and drive-time calibration and sanity checks.

There has been a nearly constant sub-plot of redundancy considerations for AV hardware. My personal take is it some of these are overblown, but the point is that Hardware reliability and failure mitigations are hardly a new topic for AVs.
 
What happens when these driving systems age? Will the Government require recalls on these systems when they are 10 to 15 years old due to not working correctly and being dangerous on the Road? If a 10 year old car on level 3,4 or 5 crashes due to Sensor or Computer failure who is held responsible?
My guess is that L4 vehicles will be leased.
With L3 vehicles the driver is responsible for recognizing obvious mechanical issues like bad wheel bearings and loose lugnuts. I think the system has to be able to recognize sensor and computer issues itself.
 
  • Like
Reactions: scottf200
This just confirms what I have always said that the Apple self-driving car was always vaporware. It was never serious. Project Titan was nothing more than a side project. Now it is confirmed with the fact that it has been delayed to 2028 and will only be L2 highway. So basically, Apple is planning to do basic autopilot in 2028 at the earliest and might get delayed again. By 2028, who knows how far Tesla FSD will be, Waymo will continue to scale driverless robotaxis to more places, Mobileye will likely have some eyes-off highway deployed on some consumer cars etc... Apple will be hopelessly behind.
Why do you think that this is likely?
 
Why do you think that this is likely?

Mobileye has Chauffeur which takes everything SuperVision has and adds radar and lidar for extra perception reliability. Hardware wise, it looks to be capable of eyes off on highways. The extra radar and lidar should help improve the perception side of the MTBF. Of course, the real question is the planner software and the overall MTBF. I know you say that the current MTBF of SuperVision on highways is super bad but 2028 gives Mobileye 4 years to improve the MTBF of their system. I think it is likely that in 4 years, Mobileye would be able to improve the planner and improve the overall MTBF to the point of doing eyes-off on highways.

And note that I just said "some" eyes-off on highways on "some" consumer cars. I am not suggesting that Mobileye will solve L4 or have eyes-off that works everywhere or eyes off on millions of consumer cars, just that they will be able to offer some sort of eyes-off on highways product by 2028 which is 4 years away.
 
Last edited:
Here is a 45 mn video with XNGP.


Hard to know how good it is. The video was poorly instrumented and seemed to be more chat than honest assessment. Funny they needed the two extra occupants in the back with one holding a camera pointed at the dash. There was one sketchy moment where it began accelerating into a vehicle cut-in and the ego then tried making a left lane change into a clearly occupied lane. Funny they didn't comment on the faux pas.

 
  • Like
Reactions: diplomat33
Beijing (Gasgoo)- On January 31, Chinese autonomous driving solution provider Pony.ai received the unmanned intelligent connected vehicle commercial pilot license from Shenzhen's Bao'an District. This marks the commencement of the country's first commercial operation of unmanned Robotaxis in the central urban areas of a first-tier city.
Since introducing autonomous ride-hailing services in Shenzhen in 2022, Pony.ai has garnered widespread market recognition and positive user reviews, with an average daily order volume reaching 200. Operating in several core areas, Pony.ai's service spans nearly 500 pick-up and drop-off points, operating from 8:30 AM to 10:30 PM, including rush hours. Users can seamlessly summon unmanned Robotaxis through the PonyPilot+ app, enjoying a safe, comfortable, and efficient travel experience.
Pony.ai has so far deployed unmanned Robotaxi fleets in Beijing, Shanghai, Guangzhou, and Shenzhen, initiating commercial fee-based services in these cities. The company has accumulated over 29 million kilometers in autonomous road-testing distance, including 2.5 million kilometers in autonomous testing with safety operators onboard and 600,000 kilometers in fully unmanned driving tests.

 
  • Like
Reactions: WyoDude and mspohr
The video mentions nothing about Tesla, so not sure what you are going on about. Skipping through it, the end mentions GrowSF, which is a group of Tech workers trying to influence politics in SF and having a grudge with certain supervisors (and unsurprisingly they are using this as a chance to trash certain supervisors even though they have little to do with the concerns brought up by SFMTA). Not that hard to research and it's obvious they are heavily politically biased (just as they claim the opposing side to be).


Looking through it more, the only stat they analyze on the actual SFMTA letter is the misleading one on the collisions, but fail to acknowledge the primary problem brought up (namely halting incidents increasing drastically after the switch to driverless).

They mention a tweet by a supervisor about incident count being 18 and focus on one incident counted that was not applicable (forgetting of course other incident stats also include incidents that are not the fault of the subject vehicle!), but that neglects that supervisor tweets are irrelevant to the permit approval (CPUC does not consider them). Also does not acknowledge there are plenty of other incidents that are applicable. Part of the request is more data so that it is it is easier to filter out irrelevant incidents.

The later part of the video is all political stuff and promotion of their group.
As a follow up post to this, just saw this in the news. The guy in the video posted a profanity laden tweet against a bunch of San Francisco supervisors, ending in calling them to "die slow" and some activist took his tweet and mailed death threat letters to the supervisors. I remembered he seemed familiar and it made clear exactly how politically biased he is, so I was clearly right his comments needed to be taken with a huge grain of salt (especially now with hindsight about what happened with Cruise).
https://www.sfchronicle.com/sf/article/sf-garry-tan-x-threatening-peskin-police-report-18637813.php
 
  • Informative
Reactions: JB47394