Moving incrementally towards robot cars


Safety on U.S. roadways has been steadily improving since the 1970’s. In 2010, 32,885 persons were killed in crashes involving motorized vehicles, the lowest number recorded since 1949. The largest number of traffic fatalities was recorded in 1972, when 54,589 persons were killed. We have made substantial improvement in reducing the overall number of fatalities.

As much as I want to take credit for myself and other traffic engineers for this improvement, I can’t. We have made great strides as a profession. We’ve created infrastructure programs targeted at improving known safety issues. We’ve created safety campaigns and completed a huge amount of research aimed at understanding how to design safer roadways. However, most engineers are quick to admit that the majority of the credit belongs to the automobile manufacturers for making safer cars (or to the regulatory bodies that forced the automobile manufacturers to make safer cars).

However, cars are not doing enough to keep us safe. Not anywhere close. And we have been negligent in demanding that automobile manufacturers continue to develop technologies to keep us more safe. Robot cars that completely eliminate humans from the equation are the answer, and we’re slowly getting there, but too slowly.

We can not wait for Google to unveil the full robot car. We can’t wait for Google to perfect the full driverless vehicle before we begin making progress in that direction. We’ve already waited too long. There are a number of ways we should be implementing various robot car technologies that eliminate human control of the vehicles at critical times. Here is a list of a few things cars should be doing (but aren’t) to keep us safer that move us in the direction of robot cars:

  • Cars should be required to be GPS equipped. Cars should know the speed limits of the roadways on which they are driving. Cars should not be capable of exceeding those speed limits.
  • Cars should be able to sense when a vehicle is rapidly approaching a stopped or slowing object (like another car). When the car senses that the driver has been negligent in slowing to avoid hitting the stopped object, cars should apply the brakes for us. Rear end crashes should not happen.
  • Cars should be able to  identify the solid white fog line (edge line) and sense when vehicles are about to drive off the road. Cars should either correct the steering for us, or at least sound an alarm to alert us (e.g. wake a sleeping driver). Oh, and in those cases, cars should also electronically prohibit drivers from wildly over-correcting.
  • Cars should not be permitted to accelerate aggressively (except for cases when the GPS recognizes that you are, say, on a freeway on-ramp and may need to accelerate quickly). Quick acceleration should not be permitted in, say, parking lots.

This is just a quick list off the top of my head. I’m sure there are other ways cars should be doing a better job keeping us safe. The technology necessary to accomplish these goals is not beyond our capabilities. I am not certain why we have not already implemented these technologies, but I think we probably could have deployed these technologies a decade ago had we wanted to do it.

Automobile manufacturers will not embrace these changes until we force them to. The time is now to require automobile manufacturers to begin to implement these (and other) measures to keep us safe.


Reuben Collins

About Reuben Collins

Reuben lives in South Minneapolis with his wife and kids. He authors the cycling blog and tweets at @reubencollins. In his spare time, he enjoys renovating his 1939 tudor home and riding bicycles.