Driverless Cars

[sd455dan]

This technology still has a long way to go. The rudimentary first steps clearly show that it encourages complacency in drivers. After the system successfully maneuvers a little ways, the "driver" starts to lose focus and gets distracted into other activities. Being perpetually ready to second-guess and override the self-driver just isn't going to happen, so the systems need to be more nearly foolproof than they are now.

I read that and immediately thought of the old phrase "several hours of boredom punctuated by moments of pure terror..."

driver less system breeds Complacency and then the car beeps and "hands" control over at 70 mph with only a second or two to avoid a bad accident.

A few experiences like that and there will be a whole new category of PTSD

 

[cqaigy2]

As long as sensors fail, there will be major challenges to this. I don't understand how the companies will be able to handle the lawsuits that will results. There is a younger generation that is working on this. So there is a lot of thought and energy going in, but my experience says it's a LONG way off for fully automated. How much sensor redundancy will you need? How much road salt, snow covering, and dirt before the car can't function? Living in NY tells me no way will this be workable. Most sensors can't work through snow.

I was shocked when a friend told me his F350 cut out on him when he was on a one lane bridge last year. He luckily was able to glide off the bridge to the side of the road with the momentum his truck had. The truck would not restart and needed to be towed. The problem was an oxygen senor in the engine. There was NOTHING wrong with the engine. Just the senor stopped working. The geniuses decided that if this senor goes off or fails, the truck should immediately stop! I certainly don't trust a company that fails here to be successful at fully automating a car.

Good point. Seems like newer cars should be held to higher safety standards something like they have for aircraft. I bet an equivalent of an O2 sensor failure, doesn't shut down the engine on an aircraft. Not that being stalled on a one lane bridge is the same as stalling during take off.

 

[Soundguy]

anyone know how well driverless cars handle temporary road surfaces from diverted roads due to construction.. cones.. .. what about a traffic director / flagman.. etc.

 

[Soundguy]

As long as sensors fail, there will be major challenges to this. I don't understand how the companies will be able to handle the lawsuits that will results. There is a younger generation that is working on this. So there is a lot of thought and energy going in, but my experience says it's a LONG way off for fully automated. How much sensor redundancy will you need? How much road salt, snow covering, and dirt before the car can't function? Living in NY tells me no way will this be workable. Most sensors can't work through snow.

I was shocked when a friend told me his F350 cut out on him when he was on a one lane bridge last year. He luckily was able to glide off the bridge to the side of the road with the momentum his truck had. The truck would not restart and needed to be towed. The problem was an oxygen senor in the engine. There was NOTHING wrong with the engine. Just the senor stopped working. The geniuses decided that if this senor goes off or fails, the truck should immediately stop! I certainly don't trust a company that fails here to be successful at fully automating a car.

Something sounds fishy about that story. an 02 sensor is an emissions device, fuel computers can run open loop using canned data and just take a fuel milage hit. I've seen plenty of vehicles with bad 02 sensors.

PS, that would be on a gas vehicle.

Now.. if it was a diesel, there is a CPS... and that for sure will pull a diesel f350 right over to the side of the road and leave it there till that 25$ sensor is replaced.

 

[3930dave]

Current tech - losing a crank or sometimes even cam sensor is a hard failure on a gas engine too. Coasting to the side of a highway is a pain, and definitely a hazard due to passing traffic.

Outright sensor failures on a driverless car doesn't concern me as much as partial or intermittent failures. In a past life, I did a fair bit of test-engineering - usually, soft or intermittent problems are some of the last ones to be successfully locked down, every time.

This thread alone lists various situations where existing software in driverless cars didn't perform adequately. So far, I haven't heard of hardware failures as the root-cause, but these cars are relatively young.... hardware failures can and will happen.

The software-driven tech world is famous for "Just Build It; We have Customers for Beta testing.....", largely driven by the race to be first-mover in the market. When my laptop blue-screens, I get irritated. When the systems we are discussing mess up, people die.....

Rgds, D.

 

[sd455dan]

I read that and immediately thought of the old phrase "several hours of boredom punctuated by moments of pure terror..."

driver less system breeds Complacency and then the car beeps and "hands" control over at 70 mph with only a second or two to avoid a bad accident.

A few experiences like that and there will be a whole new category of PTSD

looks like Uber monitor driver just experienced this and now a person crossing a street is dead.

 

[3930dave]

looks like Uber monitor driver just experienced this and now a person crossing a street is dead.

I believe that was the "accident" in Tempe.

I used to joke (about harmless common snafus) "What did we blame things on before computers were around ?". Sad event in all respects, but esp. when the only use for the human "driver" is to have someone to blame when the artificial "intelligence" bails out.....

Someone died. Does the monitor driver get charged ? This is progress ?

Rgds, D.

 

[Travelover]

I believe that was the "accident" in Tempe.

I used to joke (about harmless common snafus) "What did we blame things on before computers were around ?". Sad event in all respects, but esp. when the only use for the human "driver" is to have someone to blame when the artificial "intelligence" bails out.....

Someone died. Does the monitor driver get charged ? This is progress ?

Rgds, D.

It was all caught on video from the Uber car. Let's wait until the facts come out. It is entirely possible that someone just ran out in front of the car and human or robot, it was unavoidable.

 

[3930dave]

It was all caught on video from the Uber car. Let's wait until the facts come out. It is entirely possible that someone just ran out in front of the car and human or robot, it was unavoidable.

While the video may never be released, I expect that that Uber will be strenuously emphasizing the cause if that does turn out to be true.

Rgds, D.

 

[bcp]

Uber May Not Be to Blame for Self-Driving Car Death in Arizona | Fortune

"It's very clear it would have been difficult to avoid this collision in any kind of mode [autonomous or human-driven] based on how she came from the shadows right into the roadway," Moir told the paper, adding that the incident occurred roughly 100 yards from a crosswalk. "It is dangerous to cross roadways in the evening hour when well-illuminated managed crosswalks are available, she said."

"The driver said it was like a flash, the person walked out in front of them," Moir said. "His first alert to the collision was the sound of the collision."