Tuesday, March 7, 2017

The ANTOAVTYCAD: A New Type Of Autonomous Vehicle That You Can Actually Drive

Autonomous, or "self-driving" cars are much in the news today, and, judging from the hype, they are just around the corner. According to most experts: not yet. And, in my humble opinion: not ever.


Yes, the technical innovation behind the development of such enormously sophisticated systems is extremely impressive. I would not have thought that such tremendous strides in visual pattern recognition, artificial intelligence and machine learning were possible within such a relatively short time. Who would have predicted even 5 years ago that driverless cars would be capable of so much, and come so close to being truly autonomous? On that basis huge sums of money, literally in the hundreds of billions, is being invested in this technology by several car companies, not to mention tech giants like Google, Tesla, Amazon and Microsoft, among others. Enthusiasm runs high. The money is flowing. The research is exploding. The level of sophistication is steadily increasing. And according to some sources, the fully autonomous car, i.e., the car that will literally drive itself, with no need for steering wheel, accelerator or brakes, what's often referred to as "Level 5," is just around the corner.

So what's the problem? In the words of one of Toyota's top researchers, Gill Pratt,
The most important thing to understand is that not all miles are the same. Most miles that we drive are very easy, and we can drive them while daydreaming or thinking about something else or having a conversation. But some miles are really, really hard, and so it’s those difficult miles that we should be looking at: How often do those show up, and can you ensure on a given route that the car will actually be able to handle the whole route without any problem at all? Level 5 autonomy says all miles will be handled by the car in an autonomous mode without any need for human intervention at all, ever. . . .
Because the “full autonomy” phrase can mean such a wide range of things, you really have to ask the question, “What do you really mean, what are the actual circumstances?” And usually you’ll find that it’s geofenced for area, it may be restricted by how much traffic it can handle, for the weather, the time of day, things like that. So that’s the elaboration of why we’re not even close.
The heart of the problem is the fact that even at 99.99% reliability, which is still a long way off, such a vehicle would be potentially dangerous, because in the long run at least some would be destined to fail. The vehicle could lose its GPS connection and get lost; it could mistake a shadow for a pedestrian and stop suddenly in the middle of fast-moving traffic; it could go blind in heavy rain or snowfall; it could veer to avoid a child, or a deer, and slam into a nearby truck; without clear lane markers it could easily drive off course; and, as we've seen lately in a San Francisco test, it could even get confused and run a red light. Not to mention the inevitability of system failure. Sooner or later, a sensor will go blind, a valve will wear out, a CPU will fail, a hidden software bug will emerge from the shadows to crash the computer. Etc. And even a single such failure could be catastrophic. An out of control automobile moving at high speed could be as dangerous as a guided missile, plowing through traffic, slamming into a crowd of pedestrians or school children, or into a nearby building.

What sounds at first like a reasonable alternative is the "level 4," or, by some accounting, "level 3," vehicle, essentially the same as a fully autonomous one but with a human driver ready to take over the controls at any instant if and when something goes wrong. Big Mistake! It's unreasonable to expect people to continuously monitor the road in such a manner, mile after mile, without getting distracted, daydreaming or even falling asleep.  If and when something goes wrong in a fast moving car, an instantaneous response is almost always necessary. Unless the driver is continually vigilant, however, the response will be far from instantaneous. And if total vigilance is required, then why bother with such a car in the first place, why not just take the wheel and drive? This alternative is clearly a non-starter and most companies have already abandoned the idea. It's "fully autonomous" or nothing. Go For Broke!

In view of all the many "challenges," it's no wonder that many in the industry are admitting that autonomous vehicles are still "a long way off." Realistically, however, when we consider all the above issues, plus the many "unknown unknowns" we can't yet imagine, it seems clear, at least to me, that such vehicles will never be viable.

Which is not to say that they won't be manufactured en masse. Far too much money has been invested by far too many of the rich and powerful, with far too much political influence, for any government to put the brakes on this out of control industry, or even meaningfully regulate it. The cars will be manufactured, they will be advertised, they will be distributed, they will appear at automobile dealerships. And they will sit there, unsold. Because, first of all, most people will be too nervous to ride in one; and second of all, there is really no great demand for such a vehicle, aside from the relative few who always like to get their hands on the "latest and greatest" gadgets. If I want to take a cab I have no problem sharing it with a driver. If I need to commute or go on a trip, I see no problem with taking the wheel myself rather than entrusting it to some robot I don't know anything about. Will the robot get me there sooner? Not really. Will the robot get me there safer? I feel confident in my driving skills, thank you -- and not even the most intelligent robot can save me from some inebriated human (or robot) honking his horn behind me and passing in the wrong lane.

So. Countless billions will have been invested in this new and exciting technology. But hardly anyone will actually want to buy one of these things. And those who do will soon find themselves in unreliable vehicles that will be unmanageable in anything but the most ideal conditions. What to do? Will all these marvelous devices have to be junked?

Not necessarily, no. (To be continued . . . )

No comments:

Post a Comment