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

Autopilot a fraud? The linked author bets Elon he'll eat a hat if it changes lanes!

This site may earn commission on affiliate links.
This is nonsense and I hope whoever enjoys eating hit hat.

The car will be able to do lane changes but IMHO Tesla should have added rear and side radar along with a rear camera for added security the car will only make a lane change if it is safe to do. Maybe when Autopilot 2.0 sensors are unveiled it will have greater rear and sideways radar coverage. Being limited to 16 feet of sensor coverage is a bit limiting to what they can later do with software.

I've heard that LIDAR sensors are also becoming a lot cheaper so I hope the next revision of the Autopilot sensors will offer a more enhanced suite of sensors with some level of redundancy built it.
 
I'd like to see Tesla implement sensors like Subaru has for their Rear Cross Traffic Alert & Blindspot protection which can sense cars quickly approaching from the adjacent lane and when backing out of a parking spot it warns of approaching cars from the side. This would enable lane changing more safely, the only other risk is if another car 2 lanes over tries to merge over to the lane you are merging into at the same time, they could possibly add sonar to the side of the car to monitor over the next lane and a half but it would probably still require the driver to pay attention.

I would also like to see TACC improved to be as effective as Subaru's second generation Eyesight cameras for their adaptive cruise / emergency braking & lane tracking, which were rated the best of any manufacturer for preventing an accident and worked much better when I had that than Tesla's camera, radar & sonar so far. One issue I have noticed is in cases where the Tesla camera does not see well such as sun glare it doesn't seem to utilize radar well to track the car slowing down in front. Another thing that would help if it has color vision is to monitor red brake lights or google traffic if multiple cars in front or to the side are simultaneously braking so it knows to react more quickly to slowing highway traffic, and if it could see the blinking turn signal of a car about to merge into your lane that would help to determine whether to quickly speed up or slow down to avoid a collision. Subaru's system prevents a collision even without adaptive cruise enabled if you are within 31mph of the speed of the car in front.
 
Last edited:
I've heard that LIDAR sensors are also becoming a lot cheaper so I hope the next revision of the Autopilot sensors will offer a more enhanced suite of sensors with some level of redundancy built it.

Although he didn't directly address this, the engineer from the Autopilot team at Reach The Beach suggested that LIDAR was probably still too expensive for the next generation. Reading between the lines, I think he was suggesting that the next revision would add more radar and camera units.
 
+1 on cross traffic alerts. Out of necessity I have to back out of my garage partly blind because I can't see around the car and cross traffic alert would be such a useful safety feature. I feel sooner or later Autopilot 2.0 will be coming as the current implementation is lacking sensor coverage offered by other cars. With the sensors the Model S has, it is making maximum use of and i think by this summer they will have a better autopilot system than anyone but the system can definitely benefit from better radar coverage especially in cross traffic situations where you can see around the front or rear bumper...

I'd like to see Tesla implement sensors like Subaru has for their Rear Cross Traffic Alert & Blindspot protection which can sense cars quickly approaching from the adjacent lane and when backing out of a parking spot it warns of approaching cars from the side. This would enable lane changing more safely, the only other risk is if another car 2 lanes over tries to merge over to the lane you are merging into at the same time, they could possibly add sonar to the side of the car to monitor over the next lane and a half but it would probably still require the driver to pay attention.

I would also like to see TACC improved to be as effective as Subaru's second generation Eyesight cameras for their adaptive cruise / emergency braking & lane tracking, which were rated the best of any manufacturer for preventing an accident and worked much better when I had that than Tesla's camera, radar & sonar so far. One issue I have noticed is in cases where the Tesla camera does not see well such as sun glare it doesn't seem to utilize radar well to track the car slowing down in front. Another thing that would help if it has color vision is to monitor red brake lights or google traffic if multiple cars in front or to the side are simultaneously braking so it knows to react more quickly to slowing highway traffic, and if it could see the blinking turn signal of a car about to merge into your lane that would help to determine whether to quickly speed up or slow down to avoid a collision. Subaru's system prevents a collision even without adaptive cruise enabled if you are within 31mph of the speed of the car in front.
 
I'd really like to see a video of the "turn signal triggered automatic lane change" in snow, rain, standing water in only one lane, etc. before or soon after they deliver the feature to customers.
 
I'd really like to see a video of the "turn signal triggered automatic lane change" in snow, rain, standing water in only one lane, etc. before or soon after they deliver the feature to customers.
I think snow is out of the question. It can block the radar, ultrasound and camera. Rain... it depends on how much. I believe the standard disclaimer for Tesla Driver Assistance Functions is fair weather only.
 
I'd really like to see a video of the "turn signal triggered automatic lane change" in snow, rain, standing water in only one lane, etc. before or soon after they deliver the feature to customers.
Simple answer, if you can see the lane markings clearly and unambiguously, so can the car, if you can not, neither can it. If it can't see the lane markings, it won't change lanes automatically.

There's really no risk to this feature assuming customers use it as intended and advertised (look and make sure it's safe, touch the turn signal, away you go.) Everyone expecting a self driving car in all weather and road conditions has not been paying any attention whatsoever to anything that Tesla has ever said.
 
I think this thread has seriously derailed from the core topic: hat eating. How do we, as a community, make absolutely sure that this guy in fact does eat his [a] hat as the lane-change-if-set-if-good-visibility-on-turn-signal-activation-when-car-in-assisted-driving-mode feature rolls out OTA in less than a year?
 
Simple answer, if you can see the lane markings clearly and unambiguously, so can the car, if you can not, neither can it. If it can't see the lane markings, it won't change lanes automatically.
Point of clarification:
I'm not concerned/interested in the visibility aspect but rather how it deals with slip and such while turning. For example, when I-90 signs report "standing water" it's very easy to get cruise control and traction control to fight -- and it's very uncomfortable for the driver. I hope assisted lane changes don't have this problem.

- - - Updated - - -

I think this thread has seriously derailed from the core topic: hat eating. How do we, as a community, make absolutely sure that this guy in fact does eat his [a] hat as the lane-change-if-set-if-good-visibility-on-turn-signal-activation-when-car-in-assisted-driving-mode feature rolls out OTA in less than a year?
Keep the "bet" alive via weekly retweets?
 
Point of clarification:
I'm not concerned/interested in the visibility aspect but rather how it deals with slip and such while turning. For example, when I-90 signs report "standing water" it's very easy to get cruise control and traction control to fight -- and it's very uncomfortable for the driver. I hope assisted lane changes don't have this problem.
This sounds to me like one of those situations that every car manufacturer explicitly calls out as a situation to NEVER use cruise control in. TACC would be no different.

Cruise control is not meant for slippery road surfaces, or those covered with snow or water. Using ANY cruise control in those situations is a very bad idea, and explicitly listed in the owners manuals of every vehicle I've ever seen as something never to be done.
 
I think this thread has seriously derailed from the core topic: hat eating. How do we, as a community, make absolutely sure that this guy in fact does eat his [a] hat as the lane-change-if-set-if-good-visibility-on-turn-signal-activation-when-car-in-assisted-driving-mode feature rolls out OTA in less than a year?
Well, I posted some sample hats that might be easy to eat. However, others may want to post pictures of hats that they would like him to eat. I think we can post these to his blog/twitter/etc. to keep the bet alive.
I agree, we should focus on getting him to eat his hat.
 
This sounds to me like one of those situations that every car manufacturer explicitly calls out as a situation to NEVER use cruise control in. TACC would be no different.

Cruise control is not meant for slippery road surfaces, or those covered with snow or water. Using ANY cruise control in those situations is a very bad idea, and explicitly listed in the owners manuals of every vehicle I've ever seen as something never to be done.
And that's why I turn it off when I see those signs.

I'm still interested to see how well/poorly the vehicle behaves from a technology perspective though.
 
I think the confusion tends to be in what people expect vs what has actually been promised. People love to claim that Tesla said the car would completely drive itself in every conceivable situation, and then make the (very accurate) claim that this isn't possible with the existing hardware or today's technology, when in fact they have been very clear on exactly what it will and won't do.

Elon originally said the car would go from on-ramp to off-ramp on a freeway unassisted. He did not mention lane changes at all, nor did he say the car would merge or exit. Adding lane keeping to the existing TACC would fulfil that promise completely. At the event they also demonstrated user initiated auto lane changing functionality, there is nothing to indicate that this can't happen with the current hardware, and in fact is a very simple extension of lane keeping once they get that going.

In every discussion of any of the "auto pilot" suite of features Tesla has also been extremely clear that none of this is in any way "self driving" and that at all times a fully qualified, competent, and attentive driver must be at the controls ready to take over in an instant, just like in an airplane where the term "autopilot" was taken from.

Self driving cars are the future, not the present.

If Tesla had been completely clear, this thread would not exist.
 
I just want to know, is it physically possible to eat a hat without having some sort of medical complications? I vote he eats one of the douchey Ed Hardy trucker hats. That would be cool.

As an MD I'll bite :)

First off it's already clear this guy is going to eat a TM baseball cap. It will consist primarily of cotton fiber, with small amounts of plastic likely. It would first of all have to be cut in to very small pieces or preferably ground up to a fine powdery substance. I think eating the whole thing (200-300 grams?) in one sitting could cause complications. Likely not life threatening (all though it could get pretty bad) he would probably have a lot of gastrointestinal discomfort and there could be a chance of severe constipation following the meal. To be safe I think he should be allowed 30-60 days to consume the power, and spread it out evenly eating 5-10 grams per day. That would be completely safe.
 
Hahah I'm dying over here... Best reply on TMC ever ;)

As an MD I'll bite :)

First off it's already clear this guy is going to eat a TM baseball cap. It will consist primarily of cotton fiber, with small amounts of plastic likely. It would first of all have to be cut in to very small pieces or preferably ground up to a fine powdery substance. I think eating the whole thing (200-300 grams?) in one sitting could cause complications. Likely not life threatening (all though it could get pretty bad) he would probably have a lot of gastrointestinal discomfort and there could be a chance of severe constipation following the meal. To be safe I think he should be allowed 30-60 days to consume the power, and spread it out evenly eating 5-10 grams per day. That would be completely safe.