Truck Lust and the myths that accompany it!

[Drone Flyer]

Reading through the comments about truck lust and the "so called" (A Donald Trump vernacular) attraction to big trucks while passing them, makes me think that it is all in one's imagination.
I've tried on several occasions to see if this happens and I have to say that the power of suggestion can drive someone to drink if one is not careful.
I think the issue here is that we tend to think that our Teslas will merge with the truck because we are told it will.
I pass big rigs all the time, and although I'm thinking about "Truck Lust", the car remains where it is on the road.
One has to only read ' Between the Lines' of the AP screen. LOL!

 

[Dax279]

I never heard or knew about it until it happened to me and I can assure you that it is for real. It has only happened once in a bunch of long distance trips and I can assure it is not my imagination.

You might want to check out a video someone posts of it happening to their MS.

 

[Hugh Mannity]

I think it happens Drone Flyer. There have been many accounts of it happening and many of us that it hasn't happened to. Some type of weird and maybe specific event(s) that it happens for many.

 

[HX_Guy]

OP you may want to look at these...

 

[BerTX]

In previous threads, I came to the realization that there are two separate phenomena that people are experiencing when passing large trucks. This leads to confusion and disagreements and arguments about whether it is real or not real, important or not.

1) as the Tesla is passing the truck, the car seems to drift toward the truck. This is the one that people say may or may not be real or just a perception.

2) as the Tesla approaches the truck, but is still behind it, the car takes a sudden, violent dive toward the rear wheels of the truck.

Both of these have been reported as "truck lust", but are very different experiences.

Please be sure what others are talking about before having an argument. Anyone who has experienced #2 above will get really upset if you suggest they imagined it!

 

[Dax279]

Number two was what I experienced.

 

[brucet999]

Reading through the comments about truck lust and the "so called" (A Donald Trump vernacular) attraction to big trucks while passing them, makes me think that it is all in one's imagination.
I've tried on several occasions to see if this happens and I have to say that the power of suggestion can drive someone to drink if one is not careful.
I think the issue here is that we tend to think that our Teslas will merge with the truck because we are told it will.
I pass big rigs all the time, and although I'm thinking about "Truck Lust", the car remains where it is on the road.
One has to only read ' Between the Lines' of the AP screen. LOL!

You haven't experienced "truck lust", therefor it has never existed?

I experienced it on 5 separate occasions during a 1,000 mile trip last spring; in every case while overtaking a big rig in the lane to my right. AP1 swerved my car SHARPLY to the right on a track to collide with the left rear corner of the semi trailer. Would it have continued to actual collision had I not had hands on the wheel and taken over? I can't say, but I was truly scared.

 

[mblakele]

Related to what @BerTX just posted, I'd say it's part myth and part fact. There's a perceptual illusion where we tend to think another vehicle is getting closer as we pass by. Without AP we may subconsciously give it more room in the lane, and anyway we're probably keeping most of our attention on the road ahead — I hope. But AP tends to stay centered in the lane, and with AP we have more leisure and more cognitive bandwidth to notice these details. I think that accounts for some of the anecdotal reports of "truck lust". We could call these "category 1". But on the face of it we can't push all the reports into that first category.

I've seen AP HW1 apparently mistake a shadow or a guard rail for a lane line. If that can happen, then it's easy to see how it might mistake part of a long trailer for a lane line. The car may swerve suddenly to try to center itself in what it thinks are the new lane lines. Call these reports "category 2". An abrupt swerve like that wouldn't be unreasonable if the lane lines really did change. Construction zones often have abrupt changes like that, and I drive past a couple of more permanent examples on a regular basis. In those locations drivers should swerve in a fairly abrupt way — if they're paying attention and following the lines.

If all this is correct, category #1 may be worrying but should be safe to ignore. Category #2 is an error and potentially dangerous. Can it be fixed? Maybe, but not necessarily. Keep in mind that the AP HW1 camera is very limited. It's like a one-eyed person, wearing blinders and staring at a fixed point not too far in front of the hood. Tesla (and Mobileye) did an amazing job with that limited vision, but software can only do so much when the input data is poor. Say the camera and its machine-learning backend can't be taught to distinguish between a real, abrupt lane change and a truck-induced artifact. Then the programmers are left with a choice between failing on the real, abrupt lane change and failing on the truck-induced artifact. Neither failure would be safe, but they may have enough data to decide which one is likely to be safer.

Limitations like that throw more of the burden back onto us, the drivers. We have to understand the system's weaknesses and supervise it closely.

HW2 has potential to be much better: three front cameras, each with a different focal length and angle of view. That should give the software enough information to avoid some, perhaps all, of these errors.

 

[brucet999]

I've seen AP HW1 apparently mistake a shadow or a guard rail for a lane line. If that can happen, then it's easy to see how it might mistake part of a long trailer for a lane line. The car may swerve suddenly to try to center itself in what it thinks are the new lane lines. Call these reports "category 2". An abrupt swerve like that wouldn't be unreasonable if the lane lines really did change. Construction zones often have abrupt changes like that, and I drive past a couple of more permanent examples on a regular basis. In those locations drivers should swerve in a fairly abrupt way — if they're paying attention and following the lines.

How, exactly, could the camera mistake the rear of a truck trailer for a lane line suddenly moving farther to the right? All parts of a truck trailer, except for its non-lane-line-resembling wheels, are at least 3 feet higher than the pavement. Are you saying that the camera sees the bottom rail of a trailer as a line, and by parallax angle interprets it as a lane line somewhere to the right of the actual lane line?

I suppose that might explain how the phenomenon has largely disappeared, if we assume that AP1 has since been taught to compare lane line camera data with proximity sensor data and maybe retained data tracking the radar perception of the truck in the next lane from several seconds before coming abreast of it.

 

[BerTX]

My opinion from observations is that they are still fine-tuning how and when the car centers itself in the lane. When approaching some freeway exits, the right side lane marker gets further away from the left lane marker as the exit starts to widen. I think for a while the car just sees this as a wide lane, and tries to stay in the center of the two lines. This looks to the driver like it is trying to exit, when actually it is just staying in the center of what it perceives as the lane. Then it sees the proper right lane marker re-appear (now right in front of it), and veers back into the lane. It seems to dive toward the exit.

I think this is similar to what happens UNDER CERTAIN CONDITIONS when passing a truck. I think, when the light is right and the truck is casting a shadow, the right lane marker is lost in the contrast. The car loses that line, and sometimes it is fine and follows the left marker. Sometimes, though, I think it is picking up the right lane marker from the next lane over, thinks that is the right lane marker for its lane (which it has actually lost), and tries to center itself in the lane. Centering itself in the middle of two lanes looks to the driver like it is diving toward the back of the truck.

Just a theory.

 

[mblakele]

How, exactly, could the camera mistake the rear of a truck trailer for a lane line suddenly moving farther to the right? All parts of a truck trailer, except for its non-lane-line-resembling wheels, are at least 3 feet higher than the pavement. Are you saying that the camera sees the bottom rail of a trailer as a line, and by parallax angle interprets it as a lane line somewhere to the right of the actual lane line?

I suppose that might explain how the phenomenon has largely disappeared, if we assume that AP1 has since been taught to compare lane line camera data with proximity sensor data and maybe retained data tracking the radar perception of the truck in the next lane from several seconds before coming abreast of it.

The HW1 camera doesn't have stereo vision. That makes it difficult to judge the height of a line. Humans can sometimes manage this trick if they grow up with stereo vision: we can often judge relative heights in a flat image, and people can adjust somewhat after losing an eye. A sufficiently well-trained computer could probably do about as well — but we don't always get it right, and it's not clear to me how good AP with HW1 is at that sort of thing.

You're suggesting sensor fusion as a way around that problem? Could be... if HW1 is up to it. I get the impression that the new camera-radar fusion is pushing the limits, and it seems fairly primitive.

 

[mrElbe]

On a 3000 mile return trip to Florida in January I made a point of looking for this. My AP1 car did not exhibit one instance of "Truck Lust" an I passed hundreds of trucks of all shapes.
It may be due to having the radar unit re-aligned over a year ago.

Now point 2. mentioned by BerTX could be that the car thinks it is in follow mode and tries to align itself behind the truck.

 

[brucet999]

I think, when the light is right and the truck is casting a shadow, the right lane marker is lost in the contrast. The car loses that line, and sometimes it is fine and follows the left marker. Sometimes, though, I think it is picking up the right lane marker from the next lane over, thinks that is the right lane marker for its lane (which it has actually lost), and tries to center itself in the lane. Centering itself in the middle of two lanes looks to the driver like it is diving toward the back of the truck.

Just a theory.

But not a valid one, as it turns out. During my 5-incident trip, 3 were while traveling NNW before noon, so any shadows would have been clear of the lane line. One was eastbound and the last NNE in mid afternoon. Those last two would have placed shadows well into my lane.

 

[22522]

Once a lane reaches standard width, the vehicle should offset off the left line a fixed distance to control the lane and not invite busses to pass (recall Google SUV bus accident). Good lane discipline means controlling the lane in the same way a cyclist does.

 

[John Green]

LAWSUIT: NHTSA IGNORED FREEDOM OF INFORMATION ACT REQUEST
Government sued for allegedly not providing data about Tesla's Autopilot and Autosteer systems.
By David A. Wood, CarComplaints.com Posted in News
Lawsuit: NHTSA Ignored Freedom of Information Act Request
June 29, 2017 — The National Highway Traffic Safety Administration (NHTSA) has been sued for allegedly ignoring a Freedom of Information Act (FOIA) request filed by a research company that wants to know details about Tesla's Autopilot system.

Plaintiff Quality Control Systems Corp. (QCS), based in Maryland, says it concentrates in "computer-intensive, statistical research with large databases," and wants to know more about the data used to investigate the Tesla Model S and Model X.

Safety regulators took a look at Tesla's Autopilot system in 2016 after a crash that killed former Navy SEAL Joshua Brown. Mr. Brown was driving his Model S with Autopilot engaged when the car slammed into a tractor-trailer, killing Brown on a Florida highway.

NHTSA eventually closed its investigation and claimed the data showed Autosteer, a function of the Autopilot system, reduced airbag deployments and crash rates.

QCS says it sent a FOIA request to NHTSA in February 2017 to obtain crash data allegedly withheld from the public by the government. According to the lawsuit, scientific researchers need the data to "assess the validity of the remarkable claim made by NHTSA that airbag deployments in Tesla vehicles dropped by almost 40 percent after the installation of a component of the Tesla's Autopilot system, Autosteer."

QCS says the documents specifically say that NHTSA calculated airbag deployment crashes in the Tesla vehicles before and after Autosteer was installed and found the reduction in crash rates. However, QCS says it is concerned the alleged reduction in crash rates is associated with the "installation" of Autosteer, rather than the actual use of Autosteer.

QCS wants to know if NHTSA used scientific methods to validate the tests and if the results can be replicated. In addition, the company wants to know if the alleged reduction in crash rates is due to Autosteer itself and if the decreased crash rates are expected to continue.

"The surprising claim by NHTSA of an extraordinary reduction in crash rates associated with the installation of Autosteer must be carefully considered in the context of the Agency's failure to allow public access to the underlying data. Such an important conclusion by the Agency should not be based on data that the government is withholding from researchers who want to examine NHTSA's results." - Quality Control Systems

In March 2017, NHTSA responded to QCS by saying the agency was issuing “an interim response to your FOIA request dated February 24, 2017," and was “extending by ten working days the time period by which the agency must provide a response.”

Further, NHTSA said it needed additional time "to search for and collect the requested records from field facilities or other establishments that are separate from the office processing the request.” Finally, the letter stated that NHTSA expected to provide a response by April 14, 2017.

However, QCS says that communication was the last it heard from NHTSA.

The Quality Control Systems Freedom of Information Act request lawsuit was filed in the U.S. District Court for the District of Columbia - Quality Control Systems Corp. v. U.S. Department of Transportation.

 

[AWDtsla]

It's very real. Some of it may just be lack of adequate timely negative feedback. Trucks produce suction under their trailers, normally drivers bias away to begin with, but if you approach it center lane it's already dragging the car in, if AP is slow to counter that, it's going to feel like it's diving in.

 

[brucet999]

It's very real. Some of it may just be lack of adequate timely negative feedback. Trucks produce suction under their trailers, normally drivers bias away to begin with, but if you approach it center lane it's already dragging the car in, if AP is slow to counter that, it's going to feel like it's diving in.

How does a truck produce suction under its trailer? Even if there were suction under the trailer, I doubt if it would have much effect on a 4500lb low-profile and very aerodynamic Tesla. It would also not explain the Tesla steering wheel turning toward the truck.

 

[croman]

Trucks are apparently black holes. Who knew?

 

[5YJSA]

I think the issue here is that we tend to think that our Teslas will merge with the truck because we are told it will.

I can confirm your theory is incorrect.

I had never heard of "truck lust" on this forum or anywhere else.*

I drove from Toronto to Tampa and back using an AP1 Model S and noticed it constantly. The first time I thought it was weird and a one-time thing. After I saw it repeatedly, I realized it was something else. I did not have my hands on the wheel many times. It was either 100% consistent or extremely close to it.

Since I did not know this phenomenon did or even could exist, yet I saw it happen repeatedly for thousands of kilometres, we can conclude that I did not think it happened because I was told it would happen.

My trip was months ago (March?). I found out about truck lust and the fact that it is discussed here only yesterday.

*(Assuming "truck lust" means the Tesla on autopilot pulls toward the left rear corner of 18-wheelers as it passes them in the left-adjacent lane.)

 

[5YJSA]

You haven't experienced "truck lust", therefor it has never existed?

I experienced it on 5 separate occasions during a 1,000 mile trip last spring; in every case while overtaking a big rig in the lane to my right. AP1 swerved my car SHARPLY to the right on a track to collide with the left rear corner of the semi trailer. Would it have continued to actual collision had I not had hands on the wheel and taken over? I can't say, but I was truly scared.

Just a note:

I experienced "truck lust" repeatedly from Toronto to Tampa.

Recently from Toronto to NS, I experienced it once only, but it was such a severe swerve into the truck I was passing that AP1 actually autocorrected itself, but OVER-autocorrected so it broke its sudden swerve toward the truck with an even MORE severe swerve to the left, which would have put me straight into a concrete median if I had pulled out of autopilot any more slowly.

So I think AP1 WILL eventually autocorrect out of its truck lust (at least sometimes) but that autocorrection may be just as dangerous.