How Autonomous Cars' Safety Features Clash With Normal Driving

An anonymous reader writes: Google's autonomous cars have a very good safety record so far — the accidents they've been involved in weren't the software's fault. But that doesn't mean the cars are blending seamlessly into traffic. A NY Times article explains how doing the safest thing sometimes means doing something entirely unexpected to real, human drivers — which itself can lead to dangerous situations. "One Google car, in a test in 2009, couldn't get through a four-way stop because its sensors kept waiting for other (human) drivers to stop completely and let it go. The human drivers kept inching forward, looking for the advantage — paralyzing Google's robot." There are also situations in which the software's behavior may be so incomprehensible to human passengers that they end up turning it off. "In one maneuver, it swerved sharply in a residential neighborhood to avoid a car that was poorly parked, so much so that the Google sensors couldn't tell if it might pull into traffic."

Comments

Poor example

[fred911]

"One Google car, in a test in 2009,..."

One would think that in 6 years some improvements would have been made. Do we have a more current example?

Re:Poor example

[PolygamousRanchKid]

it will be always be a challenge to have these control systems anticipate what human drivers intend to do.

This is complicated by the fact that some human drivers do not even know themselves, what they intend to do. So how should a computer control system be able to anticipate what a human driver intends to do, when the human drivers don't even know themselves?

I really don't think it is that many . . . maybe only 1% of all human drivers. However, one clueless driver can confuse and tie up 99 drivers who know where they want to go, and can communicate it to other drivers.

It's like being on a escalator at the airport or train station. Two folks don't know where they are going. So they stop dead in their tracks at the end of the escalator, blocking the path for all the other folks on the escalator. An accordion affect ensues, with all the folks on the escalator getting squished together. The two people doing the blocking, are totally oblivious to this fact. Their field of vision ends at their own noses. They are entirely engulfed in themselves, and can't even conceive that there are other living beings around them.

This is what happens on the road, as well. The driver of the car parked halfway into the street, is just not capable of thinking, that other drivers might be confused by this. Is the car really parked? Or is the driver trying to park? Or maybe trying to drive away . . . ? At any rate, some drivers need to be taught that it is terribly important to anticipate how others might interpret their actions.

In other news

[Anonymous Coward]

Millions of people on the road today deserve to have their license taken from them because they can't follow simple rules like signaling, not parking halfway out into the street and leaving enough room to brake in case the car in front of you brakes.

Re:Actually, it IS the software's fault

[bickerdyke]

The article summary isn't very good. If the software is programmed in a way that causes a car to behave in a way that's dangerous, it IS the software's fault.

That's trivial but true.

It becomes interesting when the software has the car behaving in a way that is SAFE, but unexpected.

Re:Best solution:

[Anonymous Coward]

No surprise this is up modded insightful. I'm betting many slashdotters are horrible technologists who assume the world needs to bend to technology. Simply put, that's not the case. If these cars can't handle driving around humans they are not ready for consumption. The fact that they can't properly work with and adapt with humans on the road means that these cars are unsafe. They may be "safe" from the definition of the laws, but they are not safe if they are causing or instigating traffic accidents. It seems it's blind luck that these cars haven't been the clear cut cause of an accident yet.