Investigation into Tesla after fatal crash

Stallingrad

It was only a matter of time, will be interesting to see the outcome.

http://www.bbc.com/news/technology-36680043

Shannon757

It's the first fatal accident in 130 million miles. Not bad for new technology but tragic all the same.

Raging_Ninja

Stallingrad wrote: »

It was only a matter of time, will be interesting to see the outcome.

http://www.bbc.com/news/technology-36680043

Should never have been called "autopilot" - really it's just assisted driving.

seamus

Not really an investigation into Tesla, just an investigation into the crash.

Any road, this was indeed inevitable. The upside here is that once the data is processed and the issue corrected for, it's a scenario that will never happen again in an automated vehicle.

Unlike with human drivers, where this exact same incident will continue to happen multiple times a day, every day.

kyote00

In autopilot mode, it tried to drive under a truck that it did not "see" due to a brightly lit sky ......

"The company said in a statement: "The high ride height of the trailer combined with its positioning across the road and the extremely rare circumstances of the impact caused the Model S to pass under the trailer, with the bottom of the trailer impacting the windshield of the Model S."
In a statement, Tesla said it appeared the Model S car was unable to recognise "the white side of the tractor trailer against a brightly lit sky" that had driven across the car's path.


I wonder what other unique circumstances lie in wait ?

seamus wrote: »

Not really an investigation into Tesla, just an investigation into the crash.

Any road, this was indeed inevitable. The upside here is that once the data is processed and the issue corrected for, it's a scenario that will never happen again in an automated vehicle.

Unlike with human drivers, where this exact same incident will continue to happen multiple times a day, every day.

GreeBo

kyote00 wrote: »

In autopilot mode, it tried to drive under a truck that it did not "see" due to a brightly lit sky ......

"The company said in a statement: "The high ride height of the trailer combined with its positioning across the road and the extremely rare circumstances of the impact caused the Model S to pass under the trailer, with the bottom of the trailer impacting the windshield of the Model S."
In a statement, Tesla said it appeared the Model S car was unable to recognise "the white side of the tractor trailer against a brightly lit sky" that had driven across the car's path.


I wonder what other unique circumstances lie in wait ?

Does the Telsa not use simple distance measuring lasers for this or purely image analysis?

A simple distance laser wouldnt care about the colour of objects...

bertie4evr

From reading (Or misreading!) between the lines it sounds like the driver was decapitated, which is absolutely horrible.

But it all sounds, as they said, unique. Having glaring sunlight obscuring the camera, and a white trailer that's elevated so the radar couldn't pick it up is a perfect storm that'd fool the system.

Edit: Seems the radar did pick it up, but because it was elevated and not at road height it was recognised as an overhead sign.

Tweet

eeguy

kyote00 wrote: »

In autopilot mode, it tried to drive under a truck that it did not "see" due to a brightly lit sky ......

"The company said in a statement: "The high ride height of the trailer combined with its positioning across the road and the extremely rare circumstances of the impact caused the Model S to pass under the trailer, with the bottom of the trailer impacting the windshield of the Model S."
In a statement, Tesla said it appeared the Model S car was unable to recognise "the white side of the tractor trailer against a brightly lit sky" that had driven across the car's path.


I wonder what other unique circumstances lie in wait ?

An infinite amount I imagine.

I think there's a misconception that autonomy will bring crashes to zero. That's never going to happen, even the most advanced autopilot can't escape physics.
People will continue to die on the roads, but hopefully less and less as the tech advances.

I will say that it's great that Tesla can pull so much info from the cars and identify exactly what happened. It's absolved them from quiet a few legal cases so far.

GreeBo

bertie4evr wrote: »

From reading (Or misreading!) between the lines it sounds like the driver was decapitated, which is absolutely horrible.

But it all sounds, as they said, unique. Having glaring sunlight obscuring the camera, and a white trailer that's elevated so the radar couldn't pick it up is a perfect storm that'd fool the system.

Edit: Seems the radar did pick it up, but because it was elevated and not at road height it was recognised as an overhead sign.

Tweet

Bit of an issue if overhead isnt actually over someones head....

bertie4evr

GreeBo wrote: »

Bit of an issue if overhead isnt actually over someones head....

Absolutely, the radar I assume is pointed down, if it's in the rear-view mirror, or straight out if it's in the grille. The camera probably looks up to read speed limits and recognise other obstacles etc.

I don't understand how the sun did block it out though, as even some of the cheaper dash cams I've seen do fairly well with bright light.

Maybe it wasn't the camera being completely obscured and more that the computer vision was confused by it? I have heard of autopilot throwing warnings when driving into the sunlight and not being able to pick up road markings...

keyboard_cat

GreeBo wrote: »

Does the Telsa not use simple distance measuring lasers for this or purely image analysis?

A simple distance laser wouldnt care about the colour of objects...

Like the Google car uses however that Lidar system costs double what the car does

Del2005

kyote00 wrote: »

In autopilot mode, it tried to drive under a truck that it did not "see" due to a brightly lit sky ......

"The company said in a statement: "The high ride height of the trailer combined with its positioning across the road and the extremely rare circumstances of the impact caused the Model S to pass under the trailer, with the bottom of the trailer impacting the windshield of the Model S."
In a statement, Tesla said it appeared the Model S car was unable to recognise "the white side of the tractor trailer against a brightly lit sky" that had driven across the car's path.


I wonder what other unique circumstances lie in wait ?

Might be time the US put barriers under their HGVs like Europe has had for years, plenty of human driven vehicles go under HGVs.

GreeBo

bertie4evr wrote: »

Absolutely, the radar I assume is pointed down, if it's in the rear-view mirror, or straight out if it's in the grille. The camera probably looks up to read speed limits and recognise other obstacles etc.

I don't understand how the sun did block it out though, as even some of the cheaper dash cams I've seen do fairly well with bright light.

Maybe it wasn't the camera being completely obscured and more that the computer vision was confused by it? I have heard of autopilot throwing warnings when driving into the sunlight and not being able to pick up road markings...

I'd say it was more that the sun shining straight onto the white side of the truck from behind, "dazzled" the camera.

I still think that a simple distance measuring device fitting to the top of the car pointing straight ahead would solve this.

Similar to simple parking sensors, but obviously working a little faster.

Raging_Ninja

GreeBo wrote: »

I still think that a simple distance measuring device fitting to the top of the car pointing straight ahead would solve this.

Similar to simple parking sensors, but obviously working a little faster.

That's what the radar did - that's the "distance-measuring device". It's just that it was programmed to classify anything over X-height as an overhead sign and not apply the brakes. In this case, the X was too low.

bertie4evr

Raging_Ninja wrote: »

That's what the radar did - that's the "distance-measuring device". It's just that it was programmed to classify anything over X-height as an overhead sign and not apply the brakes. In this case, the X was too low.

I think the radar and camera work together, and in this case the radar passed below the trailer, which would typically be fine because the camera would be able to pick up anything over the radars height limit. But obviously this didn't happen.

Raging_Ninja

bertie4evr wrote: »

I think the radar and camera work together, and in this case the radar passed below the trailer, which would typically be fine because the camera would be able to pick up anything over the radars height limit. But obviously this didn't happen.

Elon Musk in the tweet above said that the radar filtered what looks like overhead signs - implying the trailer was detected, but ignored because it was over a certain height.

eeguy

Raging_Ninja wrote: »

Elon Musk in the tweet above said that the radar filtered what looks like overhead signs - implying the trailer was detected, but ignored because it was over a certain height.

Yep, I'd say the height cut off was just too low.

There's plenty of barriers, such as those on the Car Only lane at the M1 toll, that are set just above the roof of a car.

I've seen some of the American artic trucks and their ground clearance could be well over a metre, especially those who haul building materials.

Sabre Man

Article on Teslarati about the accident
http://www.teslarati.com/tesla-model-s-driver-using-autopilot-killed-crash-tractor-trailer/

It's crazy that it's possible to cross a dual carriageway like that.

Sabre Man

Sabre Man wrote: »

Article on Teslarati about the accident
http://www.teslarati.com/tesla-model-s-driver-using-autopilot-killed-crash-tractor-trailer/

It's crazy that it's possible to cross a dual carriageway like that.

The police description doesn't make it sound very appealing!

The top of Joshua Brown’s 2015 Tesla Model S vehicle was torn off by the force of the collision... the tractor-trailer was traveling west.. Brown’s car was headed east in the outside lane of U.S. 27A. When the truck made a left turn onto NE 140th Court in front of the car, the car’s roof struck the underside of the trailer as it passed under the trailer. The car continued to travel east on U.S. 27A until it left the roadway on the south shoulder and struck a fence.... Brown died at the scene.

This might sound silly, but how come the car also managed to leave the road and hit fences (and a 'power pole')? Whatever about the truck surely it'd notice these?

I'd have thought that once the car was hit (crashed into, etc) it'd bring the car to a stop (unless the driver pressed a pedal,etc. to over-power the autopilot)?

kyote00

I think the point is that the technology failed here.

Its no different than the 'sticking accelerator pedals' cases in normal cars etc ....

This time, its a design flaw in the software - presumably if the same conditions arise again, the same result will happen. It really needs to be able to detect an articulated truck in front and perpendicular of it

Quoting stats on safety so early on into the life is also a bit daft. Later on they offer the gem "requires explicit acknowledgement that the system is new technology and still in a public beta phase" --- like the washing machine shrunk your clothes....

pablo128

Deleted User wrote: »

The police description doesn't make it sound very appealing!



This might sound silly, but how come the car also managed to leave the road and hit fences (and a 'power pole')? Whatever about the truck surely it'd notice these?

I'd have thought that once the car was hit (crashed into, etc) it'd bring the car to a stop (unless the driver pressed a pedal,etc. to over-power the autopilot)?

Or unless it was absolutely flying down the road. Does anyone know what speed the car was travelling at?

RIP to the guy too. Hopefully his legacy will be an even greater improvement in safety for the worlds motorists.

bertie4evr

kyote00 wrote: »

I think the point is that the technology failed here.

Its no different than the 'sticking accelerator pedals' cases in normal cars etc ....

This time, its a design flaw in the software - presumably if the same conditions arise again, the same result will happen. It really needs to be able to detect an articulated truck in front and perpendicular of it

Quoting stats on safety so early on into the life is also a bit daft. Later on they offer the gem "requires explicit acknowledgement that the system is new technology and still in a public beta phase" --- like the washing machine shrunk your clothes....

The safety stats people are quoting (1 in 130 million) are bullsh*t anyway. It's almost the exact same fatality rate as traditional driver operated cars (Not to say it won't go down in future).

H3llR4iser

Deleted User wrote: »

This might sound silly, but how come the car also managed to leave the road and hit fences (and a 'power pole')? Whatever about the truck surely it'd notice these?

Shot in the dark, probably the impact that tore the roof off also took a bunch of fundamental equipment with it (sensors et all). At that point the system was blind, or trying to feed off wrong/discording data.

Deleted User wrote: »

I'd have thought that once the car was hit (crashed into, etc) it'd bring the car to a stop (unless the driver pressed a pedal,etc. to over-power the autopilot)?

This is more important, I believe. One'd expect a failsafe system to be there to immediately stop the vehicle if a collision/incident is detected. Also, I don't know the Tesla system, does it have a "dead man switch" like trains have, stopping the vehicle in case the driver becomes unresponsive for longer than a set time? It should be a required feature for any "autopilot" system.

kyote00 wrote: »

Quoting stats on safety so early on into the life is also a bit daft. Later on the offer the gem "requires explicit acknowledgement that the system is new technology and still in a public beta phase" --- like the washing machine shrunk your clothes....

Yep, as an engineer, I facepalm every time I hear triumphalist statements about statistics, safety and so on. It's a public beta test, things will go wrong, and it's insane to instill people with the idea the system is infallible at this point in time. But hey, it's a commercial success, so...

eeguy

kyote00 wrote: »

I think the point is that the technology failed here.

Its no different than the 'sticking accelerator pedals' cases in normal cars etc ....

It's totally different. In your example the car has forcibly taken control from the driver.

In Teslas example, the driver gave control to the car and didn't monitor his surroundings like he should.

Autopilot is a slightly more advanced cruise control and Lane assist. It's not a substitute for a driver and doesn't claim to be.

porsche boy

Deleted User wrote:

I'd have thought that once the car was hit (crashed into, etc) it'd bring the car to a stop (unless the driver pressed a pedal,etc. to over-power the autopilot)?

Does your car do that? Does any?
Reason I say that is because I have watched a few car crash compilations on you tube and the one thought I am left with is why the ABS system doesn't lock the brakes on in the event of an accident/airbag deployment. Cars seem to just wander off until they hit something that stops them, occasionally causing another crash.

kyote00

Its really is the same thing.

The flip side is saying that drivers didnt monitor the function of the accelerator pedal ....

In reality, the poor guy probably had only seconds to react...was probably expecting the car to do something....

I drove a tesla about 6 months ago in the US, albeit for a few hours only. My overwhelming memory of autopilot was permanently being on "tender hooks" waiting to see what the car would do.....

The whole tablet dash thing is really awful aswell - I'm a big fan :eek:

eeguy wrote: »

It's totally different. In your example the car has forcibly taken control from the driver.

In Teslas example, the driver gave control to the car and didn't monitor his surroundings like he should.

Autopilot is a slightly more advanced cruise control and Lane assist. It's not a substitute for a driver and doesn't claim to be.

2RockMountain

seamus wrote: »

Unlike with human drivers, where this exact same incident will continue to happen multiple times a day, every day.

I wonder if there will be the same level of investigation and press coverage into the 100-ish other people killed on the roads in the US today?

Raging_Ninja

2RockMountain wrote: »

I wonder if there will be the same level of investigation and press coverage into the 100-ish other people killed on the roads in the US today?

Where there is a systemic problem with the vehicle model there usually is such an investigation. Most traffic collisions though are caused by negligence - people either breaking the rules of the road or just not paying attention. There are very few honest-to-god genuine accidents which could not have been avoided.

porsche boy

porsche boy wrote: »

Does your car do that?

My car isn't claiming to do the driving for me, though. Surely a system like this should have a critical zone or such where it detects a collision and reacts accordingly (ie; stopping. Not, as you say yourself, rolling into another crash).

Just seems like, if safety is such a big concern on these things (and it obviously is) this would have been thought up.


In relation to HellRaiser's comments about a 'dead man switch', i did read in one of the articles that the system will acknowledge if there haven't been hands on the steering wheel in X amount of time, and it will 'gradually slow the car' until hands are detected again.

richy

I wonder how many people die per 100,000 kms driven in a manually operated car in comparison to a computer controlled car. Would be good to see a comparison.

gctest50

looking for likes maybe / bit too carried away with the tech




same lad that got killed :

2,348,565 views

eeguy

kyote00 wrote: »

Its really is the same thing.

The flip side is saying that drivers didnt monitor the function of the accelerator pedal ....

In reality, the poor guy probably had only seconds to react...was probably expecting the car to do something....
:

Nope. Drivers did monitor the accelerator, and it stuck and didn't know what to do. It was a fundamental flaw in the car.
If the poor guy actually saw the obstacle and waited for the autopilot then it's his fault. He should have done something.
In reality it looks like he was busy watching a movie while cruising down the road.
Then again, perhaps neither the driver or the car could have done anything if the truck driver swerved suddenly.
But we'll find out soon enough when Tesla review the footage recorded and release a statement.

At least we'll get a definitive answer, unlike every other car company who don't record the data.

kyote00

They already released statements where they said that 'autopilot' failed to recognise the articulated truck in front of and perpendicular to the tesla because the sun was shining and it mistook the truck for advertising....WTF.

The statement then goes on to say autopilot is on "beta" ..... again WTF ... my brakes are only beta

"In reality it looks like he was busy watching a movie while cruising down the road. "
Seriously dude, there is no mention of that anywhere

eeguy wrote: »

Nope. Drivers did monitor the accelerator, and it stuck and didn't know what to do. It was a fundamental flaw in the car.
If the poor guy actually saw the obstacle and waited for the autopilot then it's his fault. He should have done something.
In reality it looks like he was busy watching a movie while cruising down the road.
Then again, perhaps neither the driver or the car could have done anything if the truck driver swerved suddenly.
But we'll find out soon enough when Tesla review the footage recorded and release a statement.

At least we'll get a definitive answer, unlike every other car company who don't record the data.

bertie4evr

kyote00 wrote: »

They already released statements where they said that 'autopilot' failed to recognise the articulated truck in front of and perpendicular to the tesla because the sun was shining and it mistook the truck for advertising....WTF.

The statement then goes on to say autopilot is on "beta" ..... again WTF ... my brakes are only beta

"In reality it looks like he was busy watching a movie while cruising down the road. "
Seriously dude, there is no mention of that anywhere

They found a portable DVD player in the car. I'm guessing that's where it came from, but of course we have no idea what he was doing at the time. What we do know is that he didn't hit the brakes, or intervene in any way.

Autopilot has been advertised as a beta from the start, the user manual page on it is littered with warnings on how you need to pay attention and it lists a load of situations where it could go wrong.

H3llR4iser

porsche boy wrote: »

Does your car do that? Does any?
Reason I say that is because I have watched a few car crash compilations on you tube and the one thought I am left with is why the ABS system doesn't lock the brakes on in the event of an accident/airbag deployment. Cars seem to just wander off until they hit something that stops them, occasionally causing another crash.

Completely different, a "normal" car doesn't "drive" for you.

If the system doesn't have a "stop immediately in case X Gs are exceeded" feature, it's a gigantic oversight. A case where the occupants might be seriously injured and in need of urgent medical assistance, while the vehicle is still in workable conditions is not far fetched - and an automated driving one might keep on going, with little or no way of stopping it. I'd be completely surprised if the engineers in Tesla didn't implement anything along these lines - something that doesn't require remote intervention.

wil

richy wrote: »

I wonder how many people die per 100,000 kms driven in a manually operated car in comparison to a computer controlled car. Would be good to see a comparison.

Think you need to be looking at several orders of magnitude more than that, ie per billion km driven.
Ireland would be between 3 or 4 per billion, the worst countries over 50 per billion.
Would take a few years before you could expect to have any kind of meaningful stats from autonomous cars, but at the same time technology should also be improving.

marno21

Sabre Man wrote: »

Article on Teslarati about the accident
http://www.teslarati.com/tesla-model-s-driver-using-autopilot-killed-crash-tractor-trailer/

It's crazy that it's possible to cross a dual carriageway like that.

Lots of places like that on the A1 Newry-Belfast and the N25 in Cork. Not uncommon at all

Renegade Mechanic

seamus wrote: »

Not really an investigation into Tesla, just an investigation into the crash.

Any road, this was indeed inevitable. The upside here is that once the data is processed and the issue corrected for, it's a scenario that will never happen again in an automated vehicle.

Unlike with human drivers, where this exact same incident will continue to happen multiple times a day, every day.

Aw yeah, because humans can't see a ginormous juggernaut crossing the road in front of them. And even if it's close, it won't ever dawn on them to brake or aim for a wheel, so they don't get decapitated under the trailer's skirt :rolleyes: :rolleyes:

Worrying dependence developing on technology to do things as simple as driving imo. Probably linked to people simply wanting to be held responsible for themselves less and less with each passing year...

It's alright in the air with miles of space between ye. Not on the ground. I'll take control over automation of ground transport, thanks.
If it becomes phased out on cars, then I'll get a bike.

gctest50

Renegade Mechanic wrote: »

..............
It's alright in the air with miles of space between ye. Not on the ground. I'll take control over automation of ground transport, thanks.
If it becomes phased out on cars, then I'll get a bike.

GM ignition switches killed 124 + , loose control of steering


http://money.cnn.com/2015/09/17/news/companies/gm-recall-ignition-switch/

pippip

Rip first off.

Unfortunately I feel this is a classic case you see with any automation. For the first few hundred miles I'm sure the guy was going round with hovering clenched grip of the wheel and foot an inch from the brake. And then as the miles went into the thousands you'd naturally just trust the system more and more until that point where you would possibly trust it enough to start doing other things while on your journeys.

I'm all for automated driving but only where dedicated lanes are made for it. Don't feel you can mix and match non automated and automated.

eeguy

Renegade Mechanic wrote: »

I'll take control over automation of ground transport, thanks.
If it becomes phased out on cars, then I'll get a bike.

I don't understand this thinking.

Computers run safety critical operations every day and we accept this no problem.
You won't trust a computer to drive your car, but you trust your airbag, ABS and a multitude of other features to keep you safe if you make a mistake and crash.

Cookie_Monster

Is there anything to suggest that he would have had the time to react and avoid the truck in a normal car? If you're doing 60-70mph and a truck pulls out directly in front of you there'd be very little time to do anything anyway...

eeguy

pippip wrote: »

I'm all for automated driving but only where dedicated lanes are made for it. Don't feel you can mix and match non automated and automated.

Of course you can. The only way forward is mixed lanes. The cars need to fully rely on their own sensors to deal with the infinite problems they may encounter.

The issue is that a model S is not an autonomous car and doesn't claim to be. It's a beta cruise control and Lane assist.
People using it as an autonomous car are putting their own lives at risk.

frozenfrozen

Cookie_Monster wrote: »

Is there anything to suggest that he would have had the time to react and avoid the truck in a normal car? If you're doing 60-70mph and a truck pulls out directly in front of you there'd be very little time to do anything anyway...

forward planning. Would someone texting while driving have been able to avoid it? Maybe not. Would someone paying attention and actually driving properly? Probably

Chuchote

God rest him.

This seems to have been a failure of technology in how the laser was programmed. Perhaps the makers need to reprogram it so that it will recognise a body that starts 1.5 metres above the ground and continues up to 3 metres above the ground as being a vehicle, not a sign. Perhaps it should recognise a moving body as a vehicle. There are probably a dozen factors that will be used in the next version to prevent this.

I have to say that when I was driving (I now only cycle or take public transport) there were a good few times when I was squinting into harsh early-morning or late-afternoon sunlight when going directly southeast into dawn sun or directly southwest into setting sun, during which I did not have the ability to see safely enough. I should have had sense enough to pull in and wait an hour, but never did. There is no question but many accidents - and many fatal accidents - must happen in these circumstances.

Renegade Mechanic

gctest50 wrote: »

GM ignition switches killed 124 + , loose control of steering


http://money.cnn.com/2015/09/17/news/companies/gm-recall-ignition-switch/

Still better than not being in control. 124 died, how many lived because they could react in ways a computer could not?

eeguy wrote: »

I don't understand this thinking.

Computers run safety critical operations every day and we accept this no problem.
You won't trust a computer to drive your car, but you trust your airbag, ABS and a multitude of other features to keep you safe if you make a mistake and crash.

I trust my brain to avoid situations where they're required to begin with. And in situations out of my control, where they are required, it's still better to be in control. Crashes are rarely one impact, quite often they're a series.
The first impact can easily knock out sensors/radars. If your car is still careening and you're only hope is now ineffective sensor arrays, you're done... Meanwhile, the lowly pleb with only his brain to help him is still making decisions, likely reducing the total outcome.

And this is ever before we get into hacking.

Buffman

kyote00 wrote: »

"In reality it looks like he was busy watching a movie while cruising down the road. "
Seriously dude, there is no mention of that anywhere

bertie4evr wrote: »

They found a portable DVD player in the car. I'm guessing that's where it came from, but of course we have no idea what he was doing at the time. What we do know is that he didn't hit the brakes, or intervene in any way.

Ye, It's being reported by witnesses that 'Harry Potter' was still playing in the car after the crash.

The 62-year-old driver of the tractor trailer, Frank Baressi, told the Associated Press (AP) that the Tesla was driving so quickly that it “went so fast through my trailer I didn’t see him.” Combined with the alleged high rate of speed the Model S was traveling, Baressi tells the AP that he witnessed the Tesla “playing Harry Potter on the TV screen” though he acknowledged that he only heard the movie and couldn’t see it. “It was still playing when he died and snapped a telephone pole a quarter mile down the road,” Baressi tells in an interview with the AP.

Also reports that the Tesla passed another car which was doing 85mph.

My own take on these 'autopilots', is that it is the future, but there'll be a massive learning curve for the industry. It'll be similar to the way aviation has evolved over the last few decades. Motor vehicles will end up with a 'TCAS' type system to see each other in real time. But 'Murphy's law' still applies, 'Anything that can go wrong, will go wrong'.

bertie4evr

Buffman wrote: »

My own take on these 'autopilots', is that it is the future, but there'll be a massive learning curve for the industry. It'll be similar to the way aviation has evolved over the last few decades. Motor vehicles will end up with a 'TCAS' type system to see each other in real time. But 'Murphy's law' still applies, 'Anything that can go wrong, will go wrong'.

V2V is the solution to that, and it's been approved in the US. So hopefully in the future these cars will have no blind spots and could anticipate an accident sooner.

One example is I'm behind a BMW, with a Toyota in front of that. The Toyota brakes, sends the signal to the BMW that it's doing an emergency brake and the BMW sends the signal back to me. All the cars would know what the others are doing.

Atomic Pineapple

I can't find it now but I had come across a post on Twitter that linked a video claimed to be from the driver of the car that crashed showing him previously abusing the autopilot function. I didn't actually get to see the video so now sure how you would abuse the system?

pablo128

bertie4evr wrote: »

One example is I'm behind a BMW, with a Toyota in front of that. The Toyota brakes, sends the signal to the BMW that it's doing an emergency brake and the BMW sends the signal back to me. All the cars would know what the others are doing.

All well and good until you have a Renault in front. I'd be a bit nervous putting my life in the hands of a French electrical system!:pac:

eeguy

bertie4evr wrote: »

V2V is the solution to that, and it's been approved in the US. So hopefully in the future these cars will have no blind spots and could anticipate an accident sooner.

One example is I'm behind a BMW, with a Toyota in front of that. The Toyota brakes, sends the signal to the BMW that it's doing an emergency brake and the BMW sends the signal back to me. All the cars would know what the others are doing.

The new haulage trucks have that system so they can organise convoys where only the driver at the front does any work.