Potholes are due to lack of road funding. The evils of socialism
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
FSD rewrite will go out on Oct 20 to limited beta
- Thread starter diplomat33
- Start date
Gotta laugh at the amount of people who can’t grasp what Level 5 is
I think we need to introduce Level 6, just so you can delineate that it means ANYTHING A HUMAN CAN DO. Now, you’re not expected to have a ‘beam of light’ Einstein level thought experiment, but please just try to have some basic level imagination of where cars have gone & what they have traversed.
I’ll help - watch some off road 4WD videos to get the creative juices flowing
I think we need to introduce Level 6, just so you can delineate that it means ANYTHING A HUMAN CAN DO. Now, you’re not expected to have a ‘beam of light’ Einstein level thought experiment, but please just try to have some basic level imagination of where cars have gone & what they have traversed.
I’ll help - watch some off road 4WD videos to get the creative juices flowing
And potholes are caused by people. People driving. We need less people... where is the final solution when you need it, right?
wonder what bugs fixes were released in 2020.40.8.11
Tesla 2020.40.8.11 Software Update Rolls Out to FSD Beta Testers - TeslaNorth.com
Tesla 2020.40.8.11 Software Update Rolls Out to FSD Beta Testers - TeslaNorth.com
powertoold
Active Member
2020.40.8.11 below. Better in some ways worse in others, as Elon tweeted. I honestly don’t even know what reaction I should have about this FSD beta anymore. I think in a lot of ways we’ve become numb to the level of achievement this is. We’ve gone from questioning whether or not Tesla will ever achieve FSD to nitpicking about minor aspects of an essentially feature complete FSD software. Overnight.
https://twitter.com/kimpaquette/status/1320083244256268289?s=21
https://twitter.com/kimpaquette/status/1320083244256268289?s=21
I'm sure it's been joked about before, but the neural nets being trained for we AP drivers
include the ones between our ears, with each major feature released exposing the
limitations to accept. It's very Pavlovian.
include the ones between our ears, with each major feature released exposing the
limitations to accept. It's very Pavlovian.
DDotJ
Member
Almost 1.5 hours of FSD.
(May take a while for YouTube to process the full resolution though)
(May take a while for YouTube to process the full resolution though)
Oliver Cameron (CEO of Voyages, a robotaxi ride-hailing company in Florida) breaks down his thoughts on Tesla FSD Beta on Twitter:
I copied the tweets for your convenience below:
1.@Tesla FSD is a polarizing topic in AV for a few reasons:
2. Whatt@Tesla FSD is thus able to accomplish:
Remember,@Waymo only just unlocked commercial driverless (i.e. no Safety Driver) utilizing both a pre-recorded map and extra sensor modalities (lidar, and higher-resolution cameras and radars). Let’s talk about those challenges.
4. Map challenges
FSD appears to not detect this median, and thus tries to drive down the wrong side of the road. Is this an “edge case” to iron out, or is it a monstrously large technical challenge to infer road rules in real-time?
5. Map challenges
FSD appears to not understand that this is a one-way street, preventing the lane change to the left. Humans intuitively recognize this based on the directions of parked cars (and signs). Machine intelligence is not quite at that level.
6. Map challenges
I’m not quite sure what’s going on here, honestly. The route for the vehicle keeps switching from going left to right, causing a need for driver intervention as the vehicle dives for the curb. This is why drivers need to be attentive at all times.
7. My bias: no company utilizes a pre-recorded HD map because they love adding cost. They do so because inferring road features in real-time is an exceptionally hard challenge. Perhaps you can do so to 99.9% accuracy in the short-term, but is that good enough?
8. Vision challenges
In this instance, FSD appears to be about to hit a sign, requiring intervention. There are no detected object in the visualization. Is this an “edge case” that more data will iron out? Or is it that depth-estimation with only cameras is fallible?
9. Vision challenges
FSD decides to proceed at an unprotected junction even while a vehicle in cross-traffic is oncoming. This requires driver intervention. Perhaps the dark limited the range of the cameras and vision algorithms?
10. My bias: no company adds lidars to their robotaxis because they love the added cost. They do so because lidar complements the weaknesses of cameras (like seeing in darkness) and radars incredibly well.
11. Given FSD’s “beta” designation, these sorts of issues are to be expected. However, the clips above were taken from only 7 minutes of driving. Seeing these types of issue, with that frequency, gives me pause that this system is ready for fully self-driving anytime soon.
12. Now, we should also spend some time acknowledging that FSD is a damned fine accomplishment. It has been built with a relatively small team, and there are many impressive interactions. For instance…
13. When you don’t have a pre-recorded HD map to localize to, or a lidar, it can be tough to accurately perceive the exact proximity of objects with the required granularity. As such, this slight deviation for a parked vehicle was very nice!
14. Traffic light detection, without encoding positions in a pre-recorded HD map, is inherently a data-driven problem. From the small amount of clips I’ve seen, FSD is able to accurately detect not just traffic light state, but the relevance of traffic lights to each lane. Nice!
15. Even though I pointed out a few failure modes of FSD’s attempt at inferring road rules in real-time, it is still super impressive to see their progress here.
16. After balancing the current weaknesses and strengths of the system (albeit with limited data), it is clear that FSD is an impressive technological accomplishment. However, is a fully self-driving @Tesla imminent?
17. According to the little data I have, the answer is no. FSD has taken a complex problem and made it more complex, with no pre-recorded HD map and reduced sensor modalities. Their data advantage helps, but given this starting point, it is unclear if it is meaningful.
18. The fact we won’t have fully self-driving @Tesla ’s soon does not mean we cannot be excited about FSD. It’s healthy to see diversity in approach. It drives our industry to deliver a better product for customers. Congrats to @Tesla on shipping. Now, add driver monitoring!
19. Caveat: All of the above is speculation based on only what I can see. I am sure I am wrong in many places, so please don’t take the above too seriously. Thank you for reading.
https://twitter.com/olivercameron/status/1319835514887831552
I copied the tweets for your convenience below:
1.@Tesla FSD is a polarizing topic in AV for a few reasons:
- No use of pre-recorded HD maps
- Perceiving the world with cameras (no lidar!)
2. Whatt@Tesla FSD is thus able to accomplish:
- By generating a map on the fly, instead of pre-loading one recorded earlier, FSD can theoretically drive anywhere.
- Realizing cost-savings because of fewer sensor modalities.
Remember,@Waymo only just unlocked commercial driverless (i.e. no Safety Driver) utilizing both a pre-recorded map and extra sensor modalities (lidar, and higher-resolution cameras and radars). Let’s talk about those challenges.
4. Map challenges
FSD appears to not detect this median, and thus tries to drive down the wrong side of the road. Is this an “edge case” to iron out, or is it a monstrously large technical challenge to infer road rules in real-time?
5. Map challenges
FSD appears to not understand that this is a one-way street, preventing the lane change to the left. Humans intuitively recognize this based on the directions of parked cars (and signs). Machine intelligence is not quite at that level.
6. Map challenges
I’m not quite sure what’s going on here, honestly. The route for the vehicle keeps switching from going left to right, causing a need for driver intervention as the vehicle dives for the curb. This is why drivers need to be attentive at all times.
7. My bias: no company utilizes a pre-recorded HD map because they love adding cost. They do so because inferring road features in real-time is an exceptionally hard challenge. Perhaps you can do so to 99.9% accuracy in the short-term, but is that good enough?
8. Vision challenges
In this instance, FSD appears to be about to hit a sign, requiring intervention. There are no detected object in the visualization. Is this an “edge case” that more data will iron out? Or is it that depth-estimation with only cameras is fallible?
9. Vision challenges
FSD decides to proceed at an unprotected junction even while a vehicle in cross-traffic is oncoming. This requires driver intervention. Perhaps the dark limited the range of the cameras and vision algorithms?
10. My bias: no company adds lidars to their robotaxis because they love the added cost. They do so because lidar complements the weaknesses of cameras (like seeing in darkness) and radars incredibly well.
11. Given FSD’s “beta” designation, these sorts of issues are to be expected. However, the clips above were taken from only 7 minutes of driving. Seeing these types of issue, with that frequency, gives me pause that this system is ready for fully self-driving anytime soon.
12. Now, we should also spend some time acknowledging that FSD is a damned fine accomplishment. It has been built with a relatively small team, and there are many impressive interactions. For instance…
13. When you don’t have a pre-recorded HD map to localize to, or a lidar, it can be tough to accurately perceive the exact proximity of objects with the required granularity. As such, this slight deviation for a parked vehicle was very nice!
14. Traffic light detection, without encoding positions in a pre-recorded HD map, is inherently a data-driven problem. From the small amount of clips I’ve seen, FSD is able to accurately detect not just traffic light state, but the relevance of traffic lights to each lane. Nice!
15. Even though I pointed out a few failure modes of FSD’s attempt at inferring road rules in real-time, it is still super impressive to see their progress here.
16. After balancing the current weaknesses and strengths of the system (albeit with limited data), it is clear that FSD is an impressive technological accomplishment. However, is a fully self-driving @Tesla imminent?
17. According to the little data I have, the answer is no. FSD has taken a complex problem and made it more complex, with no pre-recorded HD map and reduced sensor modalities. Their data advantage helps, but given this starting point, it is unclear if it is meaningful.
18. The fact we won’t have fully self-driving @Tesla ’s soon does not mean we cannot be excited about FSD. It’s healthy to see diversity in approach. It drives our industry to deliver a better product for customers. Congrats to @Tesla on shipping. Now, add driver monitoring!
19. Caveat: All of the above is speculation based on only what I can see. I am sure I am wrong in many places, so please don’t take the above too seriously. Thank you for reading.
https://twitter.com/olivercameron/status/1319835514887831552
First couple of minutes were pretty scary. Gonna watch the rest when I get a chance. Brandon’s vids remain consistently the scariest representation of the beta, and I’m glad he’s alert. Hopefully the new release helps him at least a little.
‘That was waaaay better than how it handled the roundabout yesterday. The only question is how it would have handled a car already in the roundabout as it approached, since it didn’t even pause today. Very impressive drive through the round about to the exit lane without the random stop inside the roundabout yesterday.
DDotJ
Member
For sure, I appreciate that he shows the shortcomings of FSD and doesn't gloss over the flaws.
At the 41:00 mark it looks like FSD has issues with cones either on the land boundary or in next lane over. It was on his other videos too, where the vehicle starts to swerve when it is driving next to cones.
quadroplexor
Member
Around the 41:00 mark it struggles to drive in a straight line with cones on the right. A few minutes before it pulls out at a stop sign nearly into another car. FSD beta looks great but has a lot of edge cases to improve upon.
Silicon Desert
Well-Known Member
Exactly what I was thinking. Hey if a guy that runs a DeathStar is nervous about it, then I'll be more chicken. That was informative, but got more nerve than I do. I'm never going to trust a beta to go at 75MPH especially near a barrier, nor allow it to turn in a double turn lane with traffic. I'll wait until everyone else takes the risk.
Thanks for sharing. Nr 7/8/9 was interesting. My Audi has rear cross traffic alerting radars (one at each corner) alerting with noise and visuals in the backup camera of cars, pedestrians, cyclist etc. One can buy that for the front also. I do believe it also has a rear looking radar. I guess this could be a needed extra sensor set? If one see some of the fails so far? Especially fast European highways is a use case that could use some more radars to spot fast oncoming vehicles earlier than the cameras. When cars travel around 40 m/s faster than you, I guess the camera range is a bit to low to hold/abort a lane change in due time.
It’s not really only lack of funding...sometimes the highest area of potholes are the highest areas of traffic, and in a normal world the areas where there is very little appetite to shut down through ways just to repaid potholes, so they sit... and get bigger and wait till holidays, or notational shut downs for repair.
After watching some of the videos it is obvious that much progress has been made for City Streets. But I still don’t trust either of my cars in much simpler autopilot situations after nearly four years. When will I ever feel comfortable letting FSD handle much more complicated situations like intersections? Maybe in a year, maybe never.
One other observation: Elon’s choices of the safest and best drivers to test his software of course are slanted to the YouTubers who are often risk takers and use their privilege to push their “friends and family” referral codes. Obviously that is Elon’s right just as it is his right to let Electrek cash in multiple roadsters from Fred’s non-existent friends. But I think it merits pointing out.
One other observation: Elon’s choices of the safest and best drivers to test his software of course are slanted to the YouTubers who are often risk takers and use their privilege to push their “friends and family” referral codes. Obviously that is Elon’s right just as it is his right to let Electrek cash in multiple roadsters from Fred’s non-existent friends. But I think it merits pointing out.
None of these guys are the more popular you tubers, except for maybe TeslaRaj who didn’t actually get the update and only got to ride along with the actual guy who got the beta test.
Daniel in SD
(supervised)
Waymo
Google's original plan was a level 2 driver assist system like FSD. They dumped it after they decided that people would become complacent and not monitor it well enough.
Both are electromagnetic radiation. Both generate a depth information by measuring transit time. The advantage of microwave radiation is that it can penetrate fog, smoke, etc. , the disadvantage is the long wavelength makes it hard to get sufficient resolution. The whole silly argument was about whether they generate the same type of information which of course they do.
AlanSubie4Life
Efficiency Obsessed Member
Are we watching the same videos?
Similar threads
- Question
