Autonomous cars and cyclist salmon.
#76
Senior Member
Join Date: Sep 2014
Location: SE Michigan
Posts: 106
Bikes: 2014 Genesis GS29 (Yellow Fork)
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
Frankly, I've never heard of a car where there weren't still direct physical links for steering and brakes. Nor have I been in anything where the brakes couldn't overpower the engine if applied right. Still, it is nice to know that my manual transmission allows me to just plain break the connection between engine and wheels in two different ways. (Clutch and just pulling it out of gear.)
#77
Senior Member
Join Date: Sep 2014
Location: SE Michigan
Posts: 106
Bikes: 2014 Genesis GS29 (Yellow Fork)
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
I've just thought of something else: insurance. If an accident happens with a robocar, who is liable if, despite every precaution, it hits someone? Legal precedent is that, even of the pedestrian is drunk and falls into the road, the driver is at fault, so what happens if a drunk falls into the road in front of a robocar and doesn't have time or distance to stop or swerve before hitting him? Is the passenger of the car held accountable? The car's owner? The manufacturer?
#78
Senior Member
Join Date: Jul 2010
Location: Stephenville TX
Posts: 3,697
Bikes: 2010 Trek 7100
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 697 Post(s)
Likes: 0
Liked 3 Times
in
3 Posts
Electric steering assist; just like power steering, except that the power source is electric. Same for the brakes since there's no full-time vacuum system for regular power brakes. Both can be overridden by simply using more force on the wheel or pedal. I'd still want something to be able to physically break the link between engine and wheels, but that really hasn't been available with any automatic transmission made in the last several years; the lever just tells the transmission what to do, rather than physically taking it out of gear or engaging the park pawl.
#79
genec
Join Date: Sep 2004
Location: West Coast
Posts: 27,079
Bikes: custom built, sannino, beachbike, giant trance x2
Mentioned: 86 Post(s)
Tagged: 0 Thread(s)
Quoted: 13658 Post(s)
Liked 4,532 Times
in
3,158 Posts
That involves a long training regimen... Therefore you are not likely to get hardly anyone off the ground... hey, problem solved.
#80
20+mph Commuter
Join Date: Jul 2007
Location: Greenville. SC USA
Posts: 7,517
Bikes: Surly LHT, Surly Lowside, a folding bike, and a beater.
Mentioned: 31 Post(s)
Tagged: 0 Thread(s)
Quoted: 1434 Post(s)
Liked 331 Times
in
219 Posts
So engineers have taught the software to predict the behavior of cyclists based on thousands of encounters during the approximately 10,000 miles the cars have driven autonomously on city streets, Hohne said. The software projects a cyclist's likely movements and plots the car's path accordingly - then reacts if something unexpected happens.
#81
Senior Member
Join Date: Oct 2013
Location: Singapore
Posts: 1,143
Bikes: Fully customized 11-spd MTB built on 2014 Santa Cruz 5010 frame; Brompton S2E-X 2014; Brompton M3E 2014
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 17 Post(s)
Liked 0 Times
in
0 Posts
I've just thought of something else: insurance. If an accident happens with a robocar, who is liable if, despite every precaution, it hits someone? Legal precedent is that, even of the pedestrian is drunk and falls into the road, the driver is at fault, so what happens if a drunk falls into the road in front of a robocar and doesn't have time or distance to stop or swerve before hitting him? Is the passenger of the car held accountable? The car's owner? The manufacturer?
#82
Senior Member
Join Date: Jul 2008
Location: Austin, TX
Posts: 3,040
Bikes: Bacchetta Giro, Strada
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 6 Post(s)
Likes: 0
Liked 1 Time
in
1 Post
However, if the circuit/computer that controls that motor doesn't lose power but instead commands the wheel to turn at full strength when that's not what it should do ... you probably won't be strong enough to stop it.
That said, I haven't really heard of this happening, so I guess they've done a good job at designing the system so this doesn't happen.
But people certainly have been killed by software before. Maxillius tells us that "Anyone who knows anything about cars doesn't buy a car with drive-by-wire", so I guess that means the people driving Toyotas don't know anything about cars. (That said, it's not quite clear what the problems really were. I imagine the floor mats were the problem in some cases, operator error in others ... but they seem to think that software problems could have caused it too.)
Other cases where people were killed by bad software? The Therac-25 killed a few people due to a software problem.
#83
Senior Member
Join Date: Jul 2010
Location: Stephenville TX
Posts: 3,697
Bikes: 2010 Trek 7100
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 697 Post(s)
Likes: 0
Liked 3 Times
in
3 Posts
There is an electric motor connected to steering system, and so is your steering wheel. If you lose power, you can still steer, though with considerable effort.
However, if the circuit/computer that controls that motor doesn't lose power but instead commands the wheel to turn at full strength when that's not what it should do ... you probably won't be strong enough to stop it.
However, if the circuit/computer that controls that motor doesn't lose power but instead commands the wheel to turn at full strength when that's not what it should do ... you probably won't be strong enough to stop it.
#84
Senior Member
Mechanical systems fail. Electrical systems fail. Mechanical/Electrical/Hydraulic/Pneumatic systems fail.
I've had carberators stick before, throttle cables break, hydraulic systems fail, pedals break, tires go flat, shifters and transmissions not work. Wait are we talking about cars or bikes again.
The key to containing catastrophic failures is designing a proper limp or safe mode for each system. Seperating key systems into smaller autonomous systems, that pass information through a main system with backup and a safe shutdown, etc.
I've had carberators stick before, throttle cables break, hydraulic systems fail, pedals break, tires go flat, shifters and transmissions not work. Wait are we talking about cars or bikes again.
The key to containing catastrophic failures is designing a proper limp or safe mode for each system. Seperating key systems into smaller autonomous systems, that pass information through a main system with backup and a safe shutdown, etc.
#85
Senior Member
Join Date: Jul 2010
Location: Stephenville TX
Posts: 3,697
Bikes: 2010 Trek 7100
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 697 Post(s)
Likes: 0
Liked 3 Times
in
3 Posts
The problem is what it usually is; hubris on the part of the engineers. If they build in a failsafe at all, it's often still dependent on electronics rather than good old fashioned physically airgapping the failed system from the essential ones that could still work without it. To use the example of the electric power steering, you need a way to completely physically disengage that motor from the steering system, not just tell the software to stop turning it or cut power and have to drag against it.
#86
Senior Member
Join Date: Sep 2014
Location: SE Michigan
Posts: 106
Bikes: 2014 Genesis GS29 (Yellow Fork)
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
A shrink would have a field day trying to figure out how you saw fear in what I posted. Besides, someone has to ask those questions before the problem happens.
Change for change's sake isn't always better.
Change for change's sake isn't always better.
#87
Senior Member
Join Date: Sep 2014
Location: SE Michigan
Posts: 106
Bikes: 2014 Genesis GS29 (Yellow Fork)
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times
in
0 Posts
The problem is what it usually is; hubris on the part of the engineers. If they build in a failsafe at all, it's often still dependent on electronics rather than good old fashioned physically airgapping the failed system from the essential ones that could still work without it. To use the example of the electric power steering, you need a way to completely physically disengage that motor from the steering system, not just tell the software to stop turning it or cut power and have to drag against it.
#88
genec
Join Date: Sep 2004
Location: West Coast
Posts: 27,079
Bikes: custom built, sannino, beachbike, giant trance x2
Mentioned: 86 Post(s)
Tagged: 0 Thread(s)
Quoted: 13658 Post(s)
Liked 4,532 Times
in
3,158 Posts
The problem is what it usually is; hubris on the part of the engineers. If they build in a failsafe at all, it's often still dependent on electronics rather than good old fashioned physically airgapping the failed system from the essential ones that could still work without it. To use the example of the electric power steering, you need a way to completely physically disengage that motor from the steering system, not just tell the software to stop turning it or cut power and have to drag against it.
#89
What happened?
Join Date: Jun 2007
Location: Around here somewhere
Posts: 7,927
Bikes: 3 Rollfasts, 3 Schwinns, a Shelby and a Higgins Flightliner in a pear tree!
Mentioned: 57 Post(s)
Tagged: 1 Thread(s)
Quoted: 1835 Post(s)
Liked 292 Times
in
255 Posts
Since I first heard of this I could only speculate that the slow, super cautious vehicles were a sitting duck for gangs of highwaymen (pun intended).
__________________
I don't know nothing, and I memorized it in school and got this here paper I'm proud of to show it.
#90
Senior Member
Join Date: Apr 2011
Location: Alpharetta, GA
Posts: 15,280
Bikes: Nashbar Road
Mentioned: 71 Post(s)
Tagged: 0 Thread(s)
Quoted: 2934 Post(s)
Liked 341 Times
in
228 Posts
I've just thought of something else: insurance. If an accident happens with a robocar, who is liable if, despite every precaution, it hits someone? Legal precedent is that, even of the pedestrian is drunk and falls into the road, the driver is at fault, so what happens if a drunk falls into the road in front of a robocar and doesn't have time or distance to stop or swerve before hitting him? Is the passenger of the car held accountable? The car's owner? The manufacturer?
But this does pose a legal question and it's not so easy. Usually the driver is responsible, and it occurs to me that "driver" is usually defined (in the statutes) as the person causing the vehicle to move. He doesn't have to be steering it, he doesn't have to be in control, it doesn't even have to be in powered motion. Push a wagon down a hill and you're the driver of the vehicle. So does sitting in the car and telling it to take you to the store make you the driver? Or can the software be the driver? After all, I'm not driving when I tell a cab where to go, so would the same actions with an AI make me a driver? It could be argued both ways; it's not clear.
#91
genec
Join Date: Sep 2004
Location: West Coast
Posts: 27,079
Bikes: custom built, sannino, beachbike, giant trance x2
Mentioned: 86 Post(s)
Tagged: 0 Thread(s)
Quoted: 13658 Post(s)
Liked 4,532 Times
in
3,158 Posts
My first reaction was superficial: you'll still need liability insurance, and for product liability whoever has the deepest pockets.
But this does pose a legal question and it's not so easy. Usually the driver is responsible, and it occurs to me that "driver" is usually defined (in the statutes) as the person causing the vehicle to move. He doesn't have to be steering it, he doesn't have to be in control, it doesn't even have to be in powered motion. Push a wagon down a hill and you're the driver of the vehicle. So does sitting in the car and telling it to take you to the store make you the driver? Or can the software be the driver? After all, I'm not driving when I tell a cab where to go, so would the same actions with an AI make me a driver? It could be argued both ways; it's not clear.
But this does pose a legal question and it's not so easy. Usually the driver is responsible, and it occurs to me that "driver" is usually defined (in the statutes) as the person causing the vehicle to move. He doesn't have to be steering it, he doesn't have to be in control, it doesn't even have to be in powered motion. Push a wagon down a hill and you're the driver of the vehicle. So does sitting in the car and telling it to take you to the store make you the driver? Or can the software be the driver? After all, I'm not driving when I tell a cab where to go, so would the same actions with an AI make me a driver? It could be argued both ways; it's not clear.
If you leased the car, with an aforementioned card, from some company (the robot equivalent of a taxi), likely the company that owns the vehicle would be responsible (especially since such cars may be offering service to folks who themselves are not qualified drivers... such as blind folks or elderly). This latter situation might actually encourage both the service company to maintain the vehicles and to discourage private ownership of self driving cars.
Of course any collision of a lease type vehicle might be contested by by the lease company and they would have the deep pockets to go after the manufacture... at which point, a lawsuit would determine the ultimate responsibility... for something like a design glitch, such as the Toyota gas pedal situation.
So the answer is "it's complicated." But if you own it, likely finger pointing will start with you.
#92
Senior Member
Join Date: Apr 2011
Location: Alpharetta, GA
Posts: 15,280
Bikes: Nashbar Road
Mentioned: 71 Post(s)
Tagged: 0 Thread(s)
Quoted: 2934 Post(s)
Liked 341 Times
in
228 Posts
Good point... what if someone with a robot car lease card called a robot car... who is responsible if it collides in that case? I think it may come down to ownership of the vehicle... if you own the car, and have commanded it, then likely you would be responsible. (it may also depend on the actual driving conditions at that point... which no doubt will be recorded... the google car has collided, but it was being driven by a human at that point)
If you leased the car, with an aforementioned card, from some company (the robot equivalent of a taxi), likely the company that owns the vehicle would be responsible (especially since such cars may be offering service to folks who themselves are not qualified drivers... such as blind folks or elderly). This latter situation might actually encourage both the service company to maintain the vehicles and to discourage private ownership of self driving cars.
Of course any collision of a lease type vehicle might be contested by by the lease company and they would have the deep pockets to go after the manufacture... at which point, a lawsuit would determine the ultimate responsibility... for something like a design glitch, such as the Toyota gas pedal situation.
So the answer is "it's complicated." But if you own it, likely finger pointing will start with you.
If you leased the car, with an aforementioned card, from some company (the robot equivalent of a taxi), likely the company that owns the vehicle would be responsible (especially since such cars may be offering service to folks who themselves are not qualified drivers... such as blind folks or elderly). This latter situation might actually encourage both the service company to maintain the vehicles and to discourage private ownership of self driving cars.
Of course any collision of a lease type vehicle might be contested by by the lease company and they would have the deep pockets to go after the manufacture... at which point, a lawsuit would determine the ultimate responsibility... for something like a design glitch, such as the Toyota gas pedal situation.
So the answer is "it's complicated." But if you own it, likely finger pointing will start with you.
#93
genec
Join Date: Sep 2004
Location: West Coast
Posts: 27,079
Bikes: custom built, sannino, beachbike, giant trance x2
Mentioned: 86 Post(s)
Tagged: 0 Thread(s)
Quoted: 13658 Post(s)
Liked 4,532 Times
in
3,158 Posts
Right, and there's something else that makes it even more complicated. I haven't studied law and I don't know the legal jargon or precedents, but the basic concept is that if a product has enough general benefit the producer isn't held liable, even when the damage or fatality is foreseeable. The extreme examples are pharmaceutical or drug companies. They know that the drugs are going to kill people, or cause some horrendous side effects to a certain number of people, but if the normal use is helpful enough in general, those casualties are acceptable. They aren't liable. On the other hand you can't sell a toaster or lawnmower that you know will kill or injure someone during normal or even reasonable use. Where will the autonomous car fit into that spectrum, between the toaster and prescription drugs? It's going to save lives and property, is that enough to shield them from liability lawsuits? If so, is anyone at all liable? If you're not liable, you wouldn't even need insurance since proving that you're able to cover your financial responsibility is the sole purpose of requiring insurance. This could wind up being a huge benefit to owning (or using) one of these cars - no liability, no insurance. I don't know the answers to any of these questions.
At this time there are no states that permit driverless cars without drivers... thus the driver still has liability and the responsibility to push the "panic button."
But no doubt as self driving cars prove to be better drivers than humans, this situation change. If self driving cars prove safer than human driven cars, then that will open up all sorts of new issues.
#94
Senior Member
Join Date: Jun 2008
Location: Rochester MN
Posts: 927
Bikes: Raleigh Port Townsend, Raleigh Tourist
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 36 Post(s)
Likes: 0
Liked 11 Times
in
8 Posts
For more on self-driving vehicles, Australia leads the way. Admittedly, the trucks are in a mine, but there is nobody in the cab.
Australia has monstrous self-driving trucks | Stuff.co.nz
Australia has monstrous self-driving trucks | Stuff.co.nz
#95
Senior Member
Join Date: Jul 2010
Location: Stephenville TX
Posts: 3,697
Bikes: 2010 Trek 7100
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 697 Post(s)
Likes: 0
Liked 3 Times
in
3 Posts
At least in the case of a broken belt, you've still got the ability to go a few miles or more (often lots more in the daytime) with complete control to get to a safe place or a mechanic. Plenty of the other failures result in very dangerous conditions.
#96
Senior Member
Join Date: Jul 2010
Location: Stephenville TX
Posts: 3,697
Bikes: 2010 Trek 7100
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 697 Post(s)
Likes: 0
Liked 3 Times
in
3 Posts
Which defeats the main arguments in favor of them for most drivers; you won't be legally permitted to be drunk, asleep or otherwise absorbed in a way that prevents noticing a problem and hitting the button. Honestly, it sounds pretty annoying to me too; I would have trouble staying awake on a lot of trips if I was just staring out the windshield waiting for the car to do something wrong.
#97
genec
Join Date: Sep 2004
Location: West Coast
Posts: 27,079
Bikes: custom built, sannino, beachbike, giant trance x2
Mentioned: 86 Post(s)
Tagged: 0 Thread(s)
Quoted: 13658 Post(s)
Liked 4,532 Times
in
3,158 Posts
Which defeats the main arguments in favor of them for most drivers; you won't be legally permitted to be drunk, asleep or otherwise absorbed in a way that prevents noticing a problem and hitting the button. Honestly, it sounds pretty annoying to me too; I would have trouble staying awake on a lot of trips if I was just staring out the windshield waiting for the car to do something wrong.
And no doubt as the cars prove themselves and the technology gets stronger, the laws may change.
Of course if automakers keep doing stuff like the GM key problem, and the Toyota gas pedal problem... people may never trust self driving cars.
#98
Been Around Awhile
Join Date: Oct 2004
Location: Burlington Iowa
Posts: 29,974
Bikes: Vaterland and Ragazzi
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 12 Post(s)
Liked 1,536 Times
in
1,045 Posts
A little dose of reality that might wake up the day dreamers; but I doubt it:
"Perhaps one day tech enthusiasts will be able to visit a Museum of the Future That Never Was, where the Jetsons’ hover car and the Google super-robocar will sit side-by-side as showcase exhibits. Expect long lines for both, because the demos will be sensational."
Google self-driving car: It may never actually happen.
"Perhaps one day tech enthusiasts will be able to visit a Museum of the Future That Never Was, where the Jetsons’ hover car and the Google super-robocar will sit side-by-side as showcase exhibits. Expect long lines for both, because the demos will be sensational."
Google self-driving car: It may never actually happen.
#99
genec
Join Date: Sep 2004
Location: West Coast
Posts: 27,079
Bikes: custom built, sannino, beachbike, giant trance x2
Mentioned: 86 Post(s)
Tagged: 0 Thread(s)
Quoted: 13658 Post(s)
Liked 4,532 Times
in
3,158 Posts
A little dose of reality that might wake up the day dreamers; but I doubt it:
"Perhaps one day tech enthusiasts will be able to visit a Museum of the Future That Never Was, where the Jetsons’ hover car and the Google super-robocar will sit side-by-side as showcase exhibits. Expect long lines for both, because the demos will be sensational."
Google self-driving car: It may never actually happen.
"Perhaps one day tech enthusiasts will be able to visit a Museum of the Future That Never Was, where the Jetsons’ hover car and the Google super-robocar will sit side-by-side as showcase exhibits. Expect long lines for both, because the demos will be sensational."
Google self-driving car: It may never actually happen.