-
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
powertoold
Active Member
Who to believe and what approach is correct
To me, Tesla is clearly leading now (all things taken into account), but whether or not they can widely release fsd beta is 80-90% certain.
The issue isn't whether Tesla *can* widely release fsd beta, but whether they can widely release fsd beta and not get into too much trouble for it, because people will obviously abuse it. Therefore, wide release of fsd beta requires a very high performance threshold.
It's kind of a joke that Elon thought it'd be widely released last December, lol. We all know Elon's fsd timelines are horrendous.
linux-works
Active Member
Daniel in SD
(supervised)
They released Smart Summon. I’m not sure it requires a high performance threshold to release.
Can you quantify the performance threshold that should convince us that Tesla’s approach is correct?
powertoold
Active Member
Nope, I do software development (among other fields), with little bits and pieces in machine learning, but just from a high level (like using APIs). I also work on barcode readers from cameras and whatnot (image processing and deciphering).
I think I have good intuition about technology, software, and related industries. Everything I post is mostly my opinion.
powertoold
Active Member
Smart summon won't cause a serious accident, whereas fsd beta has done things that would have caused serious accidents.
Daniel in SD
(supervised)
Right, but it really seems to keep people on their toes! If it never did things that would cause serious accidents it would be done. Honestly I have no idea if wide release in its current state would be safe or not. I think few people would use it for very long (like summon) which would definitely help with safety.
What performance threshold will show that Tesla’s approach is correct?
Hopefully that is true about Smart Summon. We've seen Smart Summon wedge itself against the garage door, mount curbs, hit cars and objects beside it, come close to hitting pedestrians in front, cause other drivers to make avoidance maneuvers in driving lanes, and cause some mild panic from pedestrians who think it's driving uncontrolled.
I would say the risk profile is mainly to nearby people in crossing paths and standing beside the car. It's quite capable of running over or crushing them. Also you can't say what reaction could happen when it crosses driving lanes in the parking lot. I'd also like to see some kind of visual and audible signal given when a vehicle is moving without a driver, more than just the hazard lights. To some degree Waymo and other marked cars or those with Lidar are more obviously different, so they probably cause more curiosity than panic when they move around with no driver. The Tesla just looks like a normal car, but mostly silent, just driving with hazard lights on.
I don't think Smart Summon activates the hazard lights. (It does use turn signals sometimes when turning, I think.)
It will flash the hazard lights at the start of each Smart Summon activation. There may also be local laws restricting driving with hazard lights on.
Immediately after initiating Smart Summon in either mode, hazard lights briefly flash,
mirrors fold, and Model S shifts into the appropriate driving gear
So is this logic called "all things taken into account" being applied to others or is it only reserved for Tesla which would be par for course.
I remember back in 2015 when Tesla fans used to say that AP1 worked everywhere which meant that Tesla was years ahead of Waymo who is geofenced.
Then when i asked them if that meant that other L2 systems are also years ahead of Waymo. They would go uhmmmm....
Tesla fans would say idiotic things like this about Waymo "They don't have Adrej", "They don't have FSD Computer", "They don't have Dojo"
Its like you haven't heard about the TPU...that makes Dojo look like a toy? Which waymo has been using for 6 years...
At the end of the day its basically a litmus test of "are you Tesla?". Anything that isn't Tesla doesn't matter and is worthless.
Its actually nothing about what Tesla is doing but the fact that they are Tesla. As people used to say, you could get a pile of *sugar* and label it Apple and the apple fans would go crazy and call it the best thing ever. Well Tesla lovers are 1000x worst than Apple fans at their peak.
If you actually cared about determining who is leading, your "all things taken into account" won't be based on Tesla.
It would be based on objective points that could be used to evaluate other companies and objective points that other companies can actually pass.
But your "all things" is whatever Tesla is doing. Just like people who were harping on Lucid because Elon said the plaid+ would have 520 miles, 3 miles over lucid. When he later said its not happening cause you don't need it. The same people who were going goo goo gaga over range and using 520 miles and efficiency as the test of whose ahead were now saying lol 500 is useless you only need 400.
The funny part is that your "all things taken into account" is literally just Tesla giving a beta software of a door to door system that works anywhere in the country to 77 people and telling them to make videos but not livestream and having them remove videos they don't like. A system that goes acouple mins without a safety disengagement. That's it.
So that is the barometer. Yet you won't say it because there's a bunch of companies who will do this in 2022 in a much harder environment at better safety disengagement rates. Especially Chinese companies who are more like tech companies and not auto companies. Compared to traditional US and EU automakers. Huawei, Mobileye, Nio, Xpeng, Baidu...etc
Since you declare Tesla as the leader that is way way ahead and that its "game over'. With a 5-10 years lead according to most Tesla fans.
These other companies should also be 5-10 years ahead of waymo.
Ofcourse that would be logical thinking of which you tesla fans forbid. Circular logic is what's preferred.
and Yet you listened to him and continue to listen to him after he's been wrong for 6 straight years about AV and Level 5.
While the other SDC companies have been right for 6 years yet you call them a joke.
Last edited:
powertoold
Active Member
Having worked on large software products from ground up, there's a huge difference between something that you can demo vs something you can deploy to laypeople / users. We always have to take that into account when dealing with fsd software.
linux-works
Active Member
if you had a hardline wired connection to the car, THEN I'd feel a lot better about the whole idea; but knowing how bad wifi and other consumer wireless protocols are for realtime control, it makes me shudder to think they use this for controlling thousands of pounds of car, like some RC quadcopter toy.
hell, with the quadcopter, you are not running IP and wifi and the 2.4ghz control channel is a TRUE control channel with realtime characteristics.
it should be outlawed to use things that are so jittery for industrial or heavy-machinery control. at the very least, its highly irresponsible.
this includes all rf protocols that phones can speak. I dont know of any that have a true dedicated control rf channel like a 2.4ghz nrf24l01 (etc) chip would be able to do. just the fact that it would run IP makes me nervous as hell (QoS be damned, still not good enough for life/death control of heavy machinery).
hell, with the quadcopter, you are not running IP and wifi and the 2.4ghz control channel is a TRUE control channel with realtime characteristics.
it should be outlawed to use things that are so jittery for industrial or heavy-machinery control. at the very least, its highly irresponsible.
this includes all rf protocols that phones can speak. I dont know of any that have a true dedicated control rf channel like a 2.4ghz nrf24l01 (etc) chip would be able to do. just the fact that it would run IP makes me nervous as hell (QoS be damned, still not good enough for life/death control of heavy machinery).
FloridaJohn
Member
But it tops out at 6mph, which reduces the risk of serious accident drastically and greatly increases the reaction time available to both the person monitoring it and also any people interacting with it. It also is confined mostly to parking lot usage, which while there may be a lot of traffic, it's mainly low speed traffic.
As I pointed out, the risk of serious injury for pedestrians go up drastically as impact speeds go past 15mph. 6mph is well below that (and risk to occupants of the vehicle obvious is close non-existent, even if someone remained in vehicle during the Summon).
Impact Speed and a Pedestrian's Risk of Severe Injury or Death - AAA Foundation
City Streets obviously will be at much higher speeds and interacting with much higher speed traffic. But the other bit is it'll be interacting with a lot more different road/lane types and signals, so would need a lot more development just for basic competence. Smart Summon basically only needs to deal with stop signs.
powertoold
Active Member
That's exactly what I mean, also the case with the latest Tesla fsd promo video from Autonomy Day (which I suspect used HD maps).
Last edited:
Depends on what you define for "right". which is the crux of the argument (and why as I mentioned trying to pick "winners" in a comparison chart doesn't work given people have different criteria).
If "right" is releasing L5, then none of the players are "right" given none of them have reached L5 (or even just wide release L4). The basic paths being taken I see is:
1) Tesla's approach: work to release end-to-end L2, then improve reliability of that until it is to L4 then eventually L5 level. The idea is that they can convert their fleet at a flip of a switch as soon as the software is ready.
2) Others: release L4 in geofenced area relying heavily on HD maps, and achieve L5 by mapping the entire world or at least specific countries. Note that many companies don't have a explicit goal to achieve L5, they would be perfectly happy with L4 (robotaxies can operate just fine in geo fences), in which case the goal would instead be wide release L4 (releasing L4 in many more places than just experimental fleets, like Uber/Lyft's current coverage area vs when they originally started a decade ago).
Propilot is probably the biggest disappointment. I remember all the hype about it being L4 (from your post in 2017 "Complete autonomous driving for all driving situation on the highway.") and then the final release isn't even L3 and it hasn't even reached US shores yet as of now. From what I can find, it was released in a limited manner in Japan in late 2019 for some Skylines. That's the biggest problem with all of Mobileye's solutions, it takes forever to reach the consumer.
Four Upcoming Self Driving Level 3 Cars by 2019
That system is based on EyeQ4, and from a quick google, Mobileye had engineering samples already back in 2015 (they may have demoed an early release earlier than that even, given you follow Mobileye closely, perhaps you know):
Moving Closer to Automated Driving, Mobileye Unveils EyeQ4® System-on-Chip with its First Design Win for 2018
And here 6 years later we are still waiting for it to hit US shores in volume. Does that mean we won't see any significant release of the things demoed by Mobileye recently until 2027? A lot can change by then.
Last edited:
This is Ford's Co-pilot/Mobileye in 2021... Utterly ridiculous, but I'm repeating myself what I said that Mobileye hasn't progressed in years. They don't even have the basics to match Tesla base AP.
Yup. All of these auto OEMs and robotaxi services have marketing spiel that consists of "future plans". Waymo/Google has immensely deep pockets and they've been focusing one ONE thing. Their plans years ago to buy thousands of vehicles to expand their service didn't happen.
My hope is Tesla getting to FSDbeta 9.x (moving to FSDbeta 10.x) is going to make the ADAS industry finally put-up, get serious, provide competition, and pressure Tesla.
Yup. All of these auto OEMs and robotaxi services have marketing spiel that consists of "future plans". Waymo/Google has immensely deep pockets and they've been focusing one ONE thing. Their plans years ago to buy thousands of vehicles to expand their service didn't happen.
My hope is Tesla getting to FSDbeta 9.x (moving to FSDbeta 10.x) is going to make the ADAS industry finally put-up, get serious, provide competition, and pressure Tesla.
rxlawdude
Active Member
If only others were as honest as you were in the quote above. Because most are sopranos in the torch and pitchfork choir.
Similar threads
- Article
