Some day, you'll probably be driving--or be driven by--an autonomous robot vehicle.
Rapid progress being made in development of the sensors, computer programs and code, and other technology necessary to power autonomous robot-cars--vehicles that can drive themselves.
In 2004, DARPA (Defense Advanced Research Projects Agency--the Defense Dept's futuristic research funding arm) sponsored a challenge in the desert to see if a robotic vehicle could finish an off-road course within a reasonable time. None of the entrants got more than a few miles before wrecking or otherwise failing miserably. It looked like robo-cars were still well off in science fiction land.
Then, in 2005, DARPA repeated the challenge, with a markedly different outcome. This time, four entrants finished the race course in under the 10-hour limit, led by a car (pictured above) created by a team from Stanford (, which finished the 132 mile course in around seven hours. (Two Carnegie-Mellon robo-cars and one from a private consortium followed closely behind.)
This year, DARPA will move the challenge to an urban environment, or at least a mock one, to see how robotic vehicles perform in a more typical driving environment. Up to 20 entries will compete for a shot at the $2 million first prize for finishing the course fastest (without wrecking or running over a mock pedestrian, of course).
These sophisticated robo-races present an excellent opportunity to develop and prove next-generation technologies likely to be incorporated in your car of the future. Some of the technologies are already appearing in select upscale models, in the form of front, rear and side sensors that can detect when another vehicle is too close, either sounding a warning or even making a course correction or brake adjustment.
Fully robotic cars are, no doubt, still many years away, but we could very well see them in specialized environments within a decade, followed by expansion to the broader population.
Of course, all of us would like to keep our manually driven vehicle while EVERYONE ELSE switches to a robo-car (because we all think we drive better than anyone else).
But apart from keeping jerks off the road--or at least off the steering wheel--one advantage of robo-cars will be their ability to communicate with each other. This could allow greater density of vehicles on the road, as well as elimination of traffic signals. For example, if a vehicle can sense that no other vehicle is at (or approaching) an intersection, there is no need to stop. Even with multiple vehicles at an intersection, computer protocols could optimize the order in which cars proceed, or even vary their speeds so that none need to stop. If vehicles no longer need to stop and go in urban driving, they will have significant energy savings as well.
Robo-cars may also be able to sense and find empty parking spaces, without driving around. Indeed, your robo-car could drop you off at the entrance to wherever your going, and then go park itself. Robo-valet! This will be great, since by 2020 every citizen of our country will be at least 100 pounds overweight and unable to walk more than a few feet at a time.
Here's another possibility: robo-car sharing. Let's face it, 99% of people who own large SUV's and pick-ups do so for the 2% of the time they actually need one. Suppose, instead, you joined a robo-car sharing co-op, which had mostly smaller cars--even tiny single passenger commuter ones--and a few larger ones for the rare occasion you need one. The requisite vehicle shows up on its own at your doorstep at the desired time and off you go.
And, of course, we like the idea of the robo-car that takes the auto thief to the local police station.
The Curmudgeon will be pretty old and even more curmudgeonly by the time this all happens, but we predict our grandchildren may not even learn how to drive. Perhaps that will be a good thing.
No comments:
Post a Comment