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

FSD Beta 10.69

This site may earn commission on affiliate links.
Today is the first sunny day after several cloudy and rainy days, after cleaning all the cameras, and calibrating the cameras, In the first very short FSDb, I am the leading car and stopped at the busiest home depot/Lowes cross junctions, as traffic are crossing from different directions but mine is still red, so should stand still, but out of expectation, it decided to move forward a few feet, not sure why but I brake it immediately and take over.🧐
 
Today is the first sunny day after several cloudy and rainy days, after cleaning all the cameras, and calibrating the cameras, In the first very short FSDb, I am the leading car and stopped at the busiest home depot/Lowes cross junctions, as traffic are crossing from different directions but mine is still red, so should stand still, but out of expectation, it decided to move forward a few feet, not sure why but I brake it immediately and take over.🧐
and then...?
 
  • Like
Reactions: FSDtester#1
Here is one roundabout FSDb struggles with - (47.567945441354084, -122.05498772278888).Note that there has been some changes to this roundabout recently - the weird lane that is going around the roundabout (close to East Lake Trail) is closed now.

The main problem is - there are two lanes that enter the roundabouts. One from 43rd way is particularly troubling because the main traffic direction is the East Lake N/S (left/right in the picture - D to C). So, the vehicles coming from the top on 43rd (A,B) have to wait for vehicles coming from the right (D) on East Lake.

FSD struggles with this most when I have to turn right (A to C). The car correctly takes the right lane (A) - but usually there will be other vehicles that are in the left lane (B), partially blocking the view of vehicles from the left (D) of the roundabout. I've to almost always brake here because it seems to lose track of vehicles coming from the left (occluded by vehicle standing on the left lane, B, waiting to enter) and starts to enter the roundabout.

When I've to go from East Lake (Noth/left) to 43rd - again we have two entry lanes. FSD correctly takes the left lane F, after entry goes directly to the inner roundabout lane - goes around the roundabout and exits to 43rd correctly most of the time. This is probably because there are no vehicles occluding the view to the left and moreover as I said vehicles are mostly going north-south on East lake (left/right in the picture), so not many vehicles usually come from 43rd. BTW, this used to fail most of the time before 69 - but has been working consistently in 69.2, one of the reasons I've a lot more zero disengagement drives.

When I have to go from East Lake (right) to 43rd - there is a short-cut lane that doesn't use the roundabout - G. This is essentially a protected right. But FSD will slow down to a stop almost every time - needing me to push the accelerator.

1669581268927.png
 
Here is one roundabout FSDb struggles with - (47.567945441354084, -122.05498772278888).Note that there has been some changes to this roundabout recently - the weird lane that is going around the roundabout (close to East Lake Trail) is closed now.

The main problem is - there are two lanes that enter the roundabouts. One from 43rd way is particularly troubling because the main traffic direction is the East Lake N/S (left/right in the picture - D to C). So, the vehicles coming from the top on 43rd (A,B) have to wait for vehicles coming from the right (D) on East Lake.

FSD struggles with this most when I have to turn right (A to C). The car correctly takes the right lane (A) - but usually there will be other vehicles that are in the left lane (B), partially blocking the view of vehicles from the left (D) of the roundabout. I've to almost always brake here because it seems to lose track of vehicles coming from the left (occluded by vehicle standing on the left lane, B, waiting to enter) and starts to enter the roundabout.

When I've to go from East Lake (Noth/left) to 43rd - again we have two entry lanes. FSD correctly takes the left lane F, after entry goes directly to the inner roundabout lane - goes around the roundabout and exits to 43rd correctly most of the time. This is probably because there are no vehicles occluding the view to the left and moreover as I said vehicles are mostly going north-south on East lake (left/right in the picture), so not many vehicles usually come from 43rd. BTW, this used to fail most of the time before 69 - but has been working consistently in 69.2, one of the reasons I've a lot more zero disengagement drives.

When I have to go from East Lake (right) to 43rd - there is a short-cut lane that doesn't use the roundabout - G. This is essentially a protected right. But FSD will slow down to a stop almost every time - needing me to push the accelerator.

View attachment 878831
I have a similar roundabout here. Lucille has so many different names now because of this.

One day, from the PoB to the PoE, everything from the NoA to the FSD, will be golden.


Or is it AP? I named my one of my dogs years back, (phonetically: DeeOhGee). Chum asked what language that is and what's it mean.
 
Forget about beta, that scares me!
Its not that bad. I've never seen any accidents there ... however - the city would do well to convert that to regular traffic lights. There is a traffic light just 1/2 mile from that roundabout - so - its not like the traffic light there will slow down the traffic any more.

The north/south traffic there is quite dense - esp. during peak hours. You have to be really on your toes when entering the roundabout from 43rd - you only get an opportunity once in several minutes ... for a split second.
 
  • Informative
Reactions: FSDtester#1
Disappointing FSD drive. 6 months ago FSD approached this intersection where it should have kept to the left, instead the car jerked back and forth so badly I had to disengaged. First drive on 10..69.3.1 and it was just as bad if not worse. I let FSD continue since no cars were near me and FSD just stopped in the red box right in the intersection

The mapping in this old Massachusetts mill town (Lowell) is just awful. One of the big reasons there is so much variability in how FSD performs.
Major error.

So today after yesterday's failure I tried the same section of the drive where the car had completely screwed up. The car is supposed to stay straight on Gorham Street. As you can see first FSD starts to swerve over the center line on Gorham Street, FSD then does a few jerks of the steering wheel and moves way to the right only to move back to the left centering on the gas station. The car was coming to a complete stop in the middle of the intersection until I saw the vehicle approaching on Carlisle Street so I disengaged and moved to the right even though navigation was supposed to stay straight on Gorham Street. I have now driven thru this intersection 3 times and the bad behavior is the same every time.

At least today I saved the dashcam video.


Google Maps
 
Honestly, I can’t imagine they ever reviewed anything we sent them. Maybe select clips that people followed up on with emails.
Can you imagine how Google searches the entire web to find stuff for you?

It is not just the video Tesla got, but a load of data about the event. GPS, for example. A simple search would find locations which got reported repeatedly. Intervention type, speed, jerk, temperature, windshield wiper status, lord knows what all else. Their "auto labeler" can find video with common features too.

Presumably they still do log disengagements, interventions and collisions. But some of the failures we see are probably not automatically reported, like what nnvx described. For such events, some kind of report process would be nice. Otherwise, how are they going to be improves?
Funny, I had a major fail in Lowell today as well. FSDb botched this intersection pretty badly, but it's just a typical weird intersection here in New England:

I was heading southbound on Aiken Ave. Aiken and Essex both have stop signs, and first person to stop has the right of way. Irrelevant today as there was no one around but me. Car stopped at the line, then proceeded to accelerate normally to continue on Aiken. Suddenly it slams on the brake right in the center of the intersection. I can see a creep wall on the visualization that represents the edge of Essex. Basically just like on the map view where the car is in the intersection, that's where my car stopped, thinking it was crossing Essex. In reality, it's just continuing on Aiken.

Maybe it's a map issue. But I wouldn't have rated this intersection hard for FSDb.

SW
 
  • Informative
Reactions: FSDtester#1
Can you imagine how Google searches the entire web to find stuff for you?

It is not just the video Tesla got, but a load of data about the event. GPS, for example. A simple search would find locations which got reported repeatedly. Intervention type, speed, jerk, temperature, windshield wiper status, lord knows what all else. Their "auto labeler" can find video with common features too.

Presumably they still do log disengagements, interventions and collisions. But some of the failures we see are probably not automatically reported, like what nnvx described. For such events, some kind of report process would be nice. Otherwise, how are they going to be improves?


SW

Web crawlers searching information is pretty objective where whatever situation you were in when you pressed the button is usually subjective.
 
  • Like
Reactions: FSDtester#1
69.3 was sent "wide" to prepare for "real wide" V11.

Thanks for the stats. About what I expected, 20% growth for now.
This is what it takes to do a wide release, which only Tesla has:

Definitely going to be interesting to see if they go wide, maybe with V11 or maybe a follow up of 69.3…depends on highway performance of FSD stack I guess whether V11 keeps getting pushed.

Would be quite a milestone!
 
  • Like
Reactions: FSDtester#1
FSD already saw the danger and reacted omit.
Perhaps even superhuman reaction time!

We’ll never know if FSD would have stopped; takes real guts to determine the maximum brake force that FSD will apply.

Se seriously, this performance is completely unsurprising, and it makes you wonder about how the system misses these things.
 
Last edited:
Tried this sorry mess of a release again. It failed 100% of attempted left turns and right turns and botched stopping distance for one stop sign. So basically still pre-alpha level software where work hasn’t really begun in earnest for the car to make its own turns.
Try it outside parking lots... it works better on mapped streets..lol
 
  • Funny
Reactions: Pdubs