Take trains, for example. The driver or engineer in a train does little more than apply the throttle and brakes. They don't steer, they have no direct control over switches or routing, and they don't have to monitor traffic. Certainly algorithms, a few more sensors, and computer controls can be developed to handle almost any foreseeable situation.
In fact, the technology to automate trains has been around for at least 20 to 30 years, according to David B. Clarke, director of the Center for Transportation Research at the University of Tennessee. Need proof? At least 48 fully automated metro trains operate in 32 countries. Positive train control, a system mandated for most trains by last year, uses track sensors and onboard controls to prevent train-to-train collisions, trains breaking local speed limits, and trains running through construction/maintenance zones or travelling through mispositioned switches.
Commercial airliners would be next on the list of transportation modes to go pilot-less. Autopilots already do much of the flying, and there's relatively little traffic compared with what most drivers handle daily. If the Soviet Union could design a space shuttle that doesn’t need pilots, surely airliners could be developed in the same vein.
Mind you, this wouldn't be a simple task, nor would it eliminate aircraft accidents. But however difficult coming up with robust emergency procedures may be, once they’re in place, the autopilot will follow them—unlike human pilots, who have an alarming trend of ignoring them either on purpose or in the confusion. Currently, 60% to 80% of aircraft accidents are attributed to pilot error.
One step on the route to pilot-less planes is already said to be onboard many Boeing airliners: the Boeing Honeywell uninterruptible autopilot system. It lets the plane be flown remotely, much like a drone. It even prevents any onboard pilot or hijacker from controlling the plane.
The most difficult transportation mode to automate is probably the personal car and truck, though it's the one that gets the most attention. The hype is likely generated by companies invested to the tune of several billion dollars in driverless cars. But can driverless cars handle weather? Do fog, snow, and dust storms confuse the tracking and detection systems? Does snow built up on the sensors render them useless? Can vehicles recognize the difference between a broad two-inch-deep puddle and an underpass with six feet of water in it?
Currently, driverless cars must have a driver ready to take control at a moment’s notice. That's hardly anyone's idea of a self-driving car. If the driver can’t read, make phone calls, watch a movie, or sleep, and must pay attention to the car and its surroundings, what's the purpose?
Many of the benefits of driverless cars only materialize if everyone has one and they are all networked. How long will that take? Who sets the standards? Will it be mandatory? Can the average person afford it? Will cars that need a driver be outlawed? And when your driverless car rear-ends my driverless car, who (or whose insurance) pays? These decisions need to be made before autonomous cars become common, and there probably needs to be federal laws and standards rather than state or local ones.
Perfecting and testing the technology takes time, as do passing legislation and educating the public. Let’s hope it's done right if it's going to be done. And why jump to cars before tackling trains and planes? I have my suspicions.