I agree with you, Junko, that autonomous cars are probably decades away. But that might also depend on how we define autonomous. To me, it means you enter a destination address, and then the car does its own thing while you're taking a nap or eating a three-course dinner. Or sound asleep lying down in the back seat.
Other people, in particular those who benefit from the hype, instead seem to define active ADAS as autonomous driving.
In your series of articles, and comments written responding to them, I've come to this conclusion. We're using different definitions. And never mind the refrain about lawsuits.
Specifically on this turning left at a stoplight issue, this is another one of those things that creates road rage.
Everywhere I've been, the correct way to turn left at a stoplight is, when it turns green, you move up to the middle of the intersection, blinkers on, and wait for a clearing in oncoming traffic. Then, worst case, the light turns red in your direction, and the two or perhaps three cars that are waiting patiently to turn left skidaddle through the intersection before being blocked by traffic.
Instead, you get people who forgot their driver training, and plant themselves in your way, potentially going through several cycles before they make it through.
But very good point on local customs. The first time I drove in Cyprus, years ago now, I was surprised to hear everyone honking their horns at me, because I was waiting for the light to turn green. Turns out, Cypriots look for a yellow in the other direction, and then they start moving through the red light. Okay, I thought, I can work with this. Never had a problem since.
My guess is that within about five years, aftermarket self-driving kits will start to show up in the backs of car magazines. They will be advertized as "for off road use only", just like a lot of the aftermarket ROM chips are.
The biggest thing holding back the car manufactureres will be liability concerns. Eventually, the NTSB will start requiring greater and greater levels of autonomy.
As far as the complexity involved and issues go; consider just how much information a phone collects and manages today. Consider that, while not yet fully refined, some cars have self parking, lane holding, emergency braking, adaptive cruise control, etc. It won't be long at all before all of these seemingly insurmountable challenges will be better handled by computer.
One of the things I think about are 'unwritten rules'.
Similar to my drive home in suppertime rush hour. There are a couple busy intersections which are preceeded by driveways into a post office and a community college. These friendly north Texas drivers NEVER block these driveways while waiting for a red light, thay have all learned to hang back and let traffic in and out of the driveways. Everybody gets to move.
Would a google car have enough smarts to give other drivers a break?
There are definitely different phases of automatic driving and I agree that it will be a long time before we get fully automatic cars.
One of the things I think about are 'unwritten rules'.
As an example, I remember driving in this New England town where there was a left turn lane at the light but no left turn light and the light didn't have car sensors. It would get busy during rush hour, which means that it would be near impossible to ever make a left turn. So what drivers would do is turn left immediately after the light turns green and before oncoming traffic would cross the street. At least two or three cars could make it through this way. Everyone knew the situation, so the oncoming cars would always start slow to let the few cars get through.
If they didn't do this, then the left turning cars would never be able to make their turn. A Google self-driving car would be sitting there forever. On the other hand, if you tried this in California, you'd probably get yourself killed.
Would even a semi-autonomous car let you do this or would it slam on the brakes when it sees the oncoming cars slowly approaching?
Every state, and especially every country, has a large and different set of unwritten rules that we don't conciously think about when we drive. We usually pick them up after being in an area for a while. I don't know how a fully autonomous car could do this.
Seems to me that they have had their fair share of turkeys already.
I am with you on that one. As you pointed out, the comparison of autonomous cars with autopiloted aircrafts is appropriate. And yet, we have not seen or heard yet, at this point, something equavalent to FAA is to do the oversight of autonomous cars.
Thats an interesting consideration that the industry will get it right? Seems to me that they have had thier fair share of turkeys already. It suprising with automated landing and take off that they still require a captain to fly a plane?
Seriously speaking, I am actually puzzled because many industry analysts I talked to believe that insurance companies would "lower" the rate if you drive a car with ADAS features. The assumption is that by getting the human out of the equation, there will be fewer driving errors, thus more safety is assured.
Certainly, the NCAP star rating also indicates it.
I wonder if the assumption is that if carmakers get ADAS features correctly implemented, surely they will make much "safer" autonomous cars?
Autunomous cars. Will the car have insurnace cover in it's own right. I can see getting what we call Third party liability cover in England could be interesting. The discussion with the Insurance company could sound almost as good as a Bob Newhart comedy recording.
"Hi Mr Crusty you wnat insurance, yes please, are you the named driver, Well no it's Google." " How long has Mr Google held a driving licence? Hows Mr Googles eye sight"?
The class actions against the autonomus software could be fun and go on for years.
What are the engineering and design challenges in creating successful IoT devices? These devices are usually small, resource-constrained electronics designed to sense, collect, send, and/or interpret data. Some of the devices need to be smart enough to act upon data in real time, 24/7. Are the design challenges the same as with embedded systems, but with a little developer- and IT-skills added in? What do engineers need to know? Rick Merritt talks with two experts about the tools and best options for designing IoT devices in 2016. Specifically the guests will discuss sensors, security, and lessons from IoT deployments.