The idea is to remove the mechanical linkages between the controls of a car and the devices that actually do the work. Instead of operating the steering and brakes directly, the controls would send commands to a central computer, which would instruct the car what to do.
The great advantage being put forward for this is that the computer is able to make the steering, suspension and brakes work together to give the car better handling, especially in bad road conditions, to give better fuel consumption, and to react to emergencies faster than a human driver could. Though the vehicle would look much the same, it would be transformed into what one industry expert calls “a computer network with a car wrapped round it”.
The problem for car manufacturers, who are actively researching the systems, is that getting them right is likely to be much less of a problem than convincing drivers — familiar with crashing computers at home — that their cars won’t do the same.
It’s no accident that the term sounds like fly-by-wire, which is a method of controlling commercial aircraft that has been in use for more than a decade. The term drive-by-wire has been around since the 1980s, though in early examples it could instead refer to methods of automatic steering using circuits embedded in the road surface.
As far as the industry is concerned, it is only a matter of time before drive-by-wire becomes standard. But some safety experts are questioning the wisdom of this radical change. They point out that fly-by-wire has a bumpy track record. Will the car industry learn from these mistakes, they ask, or make them all over again?
New Scientist, 8 Nov. 2003
Drive-by-wire may have to be proven first in a secondary system, such as the parking brake, before consumers grow more comfortable with the idea of replacing the traditional primary control systems in their vehicles.
Business Wire, 12 Dec. 2002