Autopilot saved our lives

[Az_Rael]

But that was versions ago.

Nope, incident was just a few months ago. So, over three years and lots of new hardware, the problem hasn’t been solved yet.

In engineering we have a term called “regression testing” that means with each new release we test everything that had been tested before as well as the new stuff. So I have tested as much as budgets allow all previous reported problems.

Bob Wilson

I am an engineer who works with safety critical systems and software, so I am very familiar with regression testing. If you had ever worked safety critical systems in your prior career, you would much more concerned with the current state of Tesla’s regression testing based on what gets released to the fleet. And as someone familiar with regression testing, you know that it is impossible for you to perform adequate regression testing on a Tesla software release, so I assume I am just missing the joke.

 

[AlanSubie4Life]

In engineering we have a term called “regression testing” that means with each new release we test everything that had been tested before as well as the new stuff. So I have tested as much as budgets allow all previous reported problems.

Bob Wilson

I do hope you are not serious about all of this! By not following Tesla’s very clear instructions to pay attention to the road and be ready to take over from Autopilot at all times, you are putting lives at risk.

The latest build of the Tesla software is extremely unlikely to have completely solved the stationary obstacle/crossing truck problem (it may not have even changed the behavior since March). All you have to do is encounter a single edge case (which may not be much of an edge case), and you or someone else could die.

Please use it as Tesla intends for it to be used, and heed the numerous warnings in the manual. When Tesla releases FSD with true full autonomy, you should buy that! Sounds like the product you need.

 

[bwilson4web]

The cited NTSB report starts "On Friday, March 1, 2019, about 6:17 a.m. eastern standard time, . . ." I then searched for "Tesla" and found only accidents dated in early 2019 and older. Are there any more recent accidents that have yet to be NTSB investigated?

I don't fault those who choose to live within their understanding of the manufacturer documentation. Go with God and be happy. Just I'm terribly curious and want to know how systems work including the limits.

AutoPilot is ground breaking, safety technology. Yet AutoPilot has limits and by testing, I can recognize edge cases soon enough to keep things boring. For example I'm desensitized to ordinary events like variable width lanes and AutoPilot 'indecision' when one lane becomes two or lane changes. So I was and remain pleased that AutoPilot seamlessly handled my unexpected micro-sleeps. But it makes sense and I'm not the first (nor will be the last!)

My only regret was not saving the dash cam video of AutoPilot in the severe rain storm event. It was at night and my attention was focused on the immediate driving requirements. AutoPlot handled both fog and heavy rain very well:

Notice how the video has enhanced the lane markers and reaches deep into the fog.

Being able to drive safely through foul weather is an important AutoPilot characteristic. It allows 'making miles' instead of being held up waiting for the weather to pass. Also avoiding being part of a multi-car pileup.

Bob Wilson

 

[Daniel in SD]

The cited NTSB report starts "On Friday, March 1, 2019, about 6:17 a.m. eastern standard time, . . ." I then searched for "Tesla" and found only accidents dated in early 2019 and older. Are there any more recent accidents that have yet to be NTSB investigated?

How long ago was March 1st 2019? You do realize that there are numerous autopilot accidents not investigated by the NTSB?

 

[Beeroclock]

A retired engineer you say?
Have you always been retired?

 

[bwilson4web]

How long ago was March 1st 2019? You do realize that there are numerous autopilot accidents not investigated by the NTSB?

Perhaps you might cite one or two that you find more interesting?

BTW, version 16.1 appears to be very interesting with improvements in "Lane Departure Avoidance."

Bob Wilson

 

[AlanSubie4Life]

Are there any more recent accidents that have yet to be NTSB investigated?

There are the several that have been reported here - autopilot running into a curb, autopilot hitting flexible road stanchions. Etc.

I am fine with users experimenting with Tesla Autopilot, even outside the zones recommended by the manual, but please do it always with an extremely high degree of vigilance.

There is a hypothesis in the autonomy world: as autonomous driving systems become more capable, they may become less safe (I argued about this ad nauseum with someone here), due to drivers trusting the system too much. You are kind of an outlier on that capability curve, but your behavior is tending to suggest that this hypothesis may turn out to be true...

 

[bwilson4web]

There are the several that have been reported here - autopilot running into a curb, autopilot hitting flexible road stanchions. Etc.

Thanks!

I missed the curb incident but the window falling out with the flexible road stanchions still has me scratching my head. I wish the OP had posted more photos. After re-reading the OP, it isn't clear that AutoPilot played a role in the glass falling out. I've posted some follow-up questions as it sounds more like a latent manufacturing defect in the window glass mechanism. I'll check my car in the morning.

The closest I came to was a mile or more of barrier barrels in Memphis that separated the repaving lane from the drive lane. Autopilot 'split the difference' and gently rocked the car going barrel-to-barrel. It was fun.

Bob Wilson

 

[Daniel in SD]

Perhaps you might cite one or two that you find more interesting?

Why?

 

[bwilson4web]

Why?

Works for me.

Bob Wilson

 

[bwilson4web]

The cited NTSB report starts "On Friday, March 1, 2019, about 6:17 a.m. eastern standard time, . . ." I then searched for "Tesla" and found only accidents dated in early 2019 and older. Are there any more recent accidents that have yet to be NTSB investigated?

• March 1 - fatal accident occurs, Version ??
• March 26 - took delivery of my Model 3 Standard Range Plus, Version 5.15
• April 4 - first references to Model 3, Version 8.5

Bob Wilson

 

[bwilson4web]

Bob Wilson

 

[Daniel in SD]

Works for me.

Bob Wilson

Haha. Let us know how it works out for you. I hope that’s your real name so we can hear about it in the news if you’re unable to post your research results.
I’m pretty convinced that you’re just trolling us though.

 

[bwilson4web]

So this morning I ran at the side of two semi truck trailers running 8.5. One trailer without the lower aero skirt and the other with:

• At no time did the trailer show up as a shadow unless I pulled along side, parallel.
• The trailer with aero skirt would kick the car out of cruise control but no emergency braking or alarm.
• The trailer without the aero skirt was never detected except in parallel on the side.

I tried both 1 and 7 distance settings.

Bob Wilson

 

[vikefan]

Basic Autopilot

In a rush, I left my electric razor and Continuous Positive Airway Pressure (CPAP) machine at home. In a world of disposables, the razor was easily fixed (Barbasol still tastes crappy) but a CPAP is not an off the shelf item. Used to treat sleep apnea, it ensures my sleep is not interrupted by short airway blockages. Then my wife spent seven days in the hospital.

Easily confused by strangers, my presence anchors her in a safer reality to avoid more aggressive restraints. But her 24 hour care requires my sleeping in that abomination called a 'reclining chair.' It has two positions, seat parallel to floor or raised ~20 degree back. There is an extendable leg lift but the 90 degree, seat-to-back angle is fixed. Reclined, it is like sleeping in a bucket with the legs and knees above the rim. To mitigate, I filled the seat-to-back hole with spare sheets, blankets, and a pillow filler so I can 'cat nap' and be available to help the staff. My wife was discharged after IV antibiotics reduced the hand swelling and pain. So I took a nap at my Mom's home, packed the car, and we were off to home in Huntsville.

'Micro sleep' is a thing that my narcoleptic wife (and late aunt) suffered. Involuntarily, the victim goes to sleep for 2-5 seconds with no warning and the most effective treatment is uninterrupted sleep. After seven days cat napping in a bucket chair and one post release nap, I was primed.

Tesla's Autopilot has dynamic cruise control and automatic emergency braking to avoid running into things. The optical and ultrasonic sensor based, lane steering keeps the car from darting into a ditch or into on-coming traffic. It was on the leg to Decatur there were at least five micro sleep events that I was aware of and Autopilot kept us safe. I stopped in Decatur for a pee, stretch, and coffee break before driving the last 20 miles home.

Micro sleeps are involuntary and detected only when the individual recognizes having woken up. When driving, it becomes either a head-on collision or single car accident. Autopilot gives them a chance to survive.

Home again, I have my CPAP sleeping aid, my wife has her dogs, and me in familiar settings. Life has returned to what passes for normal ... but it is life thanks to Tesla's Autopilot.

Bob Wilson

You’re an idiot

 

[Bill Price]

You’re an idiot

HEY! Don't sugarcoat it like that! Tell it like it is!

 

[bwilson4web]

More trailer testing using 8.5. One trailer without aero skirts and one without. Both were approached at 90 degrees. I used traffic aware cruise control (TACC) to set the approach speed. Due to parking lot dimensions, speeds were limited to 18-28 mph. It is forward detection that has the problem.

The aero skirted trailer was reliably detected showing a shadow on the center display and TACC bringing the car to a controlled stop. In contrast, the bare trailer was invisible to the car unless aimed at the rear wheels.

Bob Wilson

 

[bwilson4web]

I've been impressed with AP handling of moderate to heavy rain and fog. However, rain puddles splashed on the car from leading traffic in the next lane can, as expected, blind the optical sensor temporarily. After about 5 seconds, it runs off and you can re-enable AP but it is amusing.

I'm also moving toward the "frequent, shorter charging" from "adjust speed to reach destination charger" camp. You need both but there is nothing wrong with taking more frequent stops especially with wife and her dogs aboard.

Bob Wilson

 

[AlanSubie4Life]

More trailer testing using 8.5. One trailer without aero skirts and one without. Both were approached at 90 degrees. I used traffic aware cruise control (TACC) to set the approach speed. Due to parking lot dimensions, speeds were limited to 18-28 mph. It is forward detection that has the problem.

The aero skirted trailer was reliably detected showing a shadow on the center display and TACC bringing the car to a controlled stop. In contrast, the bare trailer was invisible to the car unless aimed at the rear wheels.

Bob Wilson

Perhaps this will still be acceptable for microsleeps since most trucks have aero skirts? Just to be 100% clear, I am “kidding.” About a very serious topic.

 

[Watts_Up]

I am an engineer who works with safety critical systems and software, so I am very familiar with regression testing.

... I hope you are not working for Boeing!

By the way, I was watching with interest a city road test using EAP.
It seems that there is still some work to handle narrow streets, in particular when cars are parked in the street.

- For example, The author explains that the EAP would stops behind an isolated parked car,
when there is no visible painted lines, like waiting for a traffic light or a stop sign, instead of passing the car.