Why U.S. Transit Systems Are Still So Far Away From Converting to Driverless Trains

With Google priming Nevada to be the first state to allow driverless cars on its roads, transit fans could be forgiven for asking: Where are the driverless trains?

The technology is relatively simple and has been around for decades. Unlike cars, which are autonomous and proceed on sight alone, railways must be centrally controlled to prevent collisions. So while a driverless car is limited by how far its sensors can "see," the central computer that directs driverless trains is fully aware of all trains on its tracks, removing much of the guesswork.

The United States doesn't yet have any fully automated trains outside of a few airport shuttles and small-scale "people movers," but Europe and Asia have adopted the technology quite readily. European firms like Italy's AnsaldoBreda and France's Matra (now owned by Siemens) pioneered the technology, and which operates on six continents. Africa's first system, in Algiers, opened in November of last year.

The obvious advantage of driverless trains over their manned counterparts is that transit agencies don't have to pay drivers. Upgrading to driverless requires a large upfront investment, and still requires humans to act as engineers, maintenance workers, janitors, and station managers. But after the investment is made, eliminating the driver position offers agencies flexibility and riders much more frequent service.

During rush hour, the main impediment to more service is the number of trains an agency owns, and in some cases tracks that simply can't handle any more traffic. But during off-peak hours, it's the cost of putting drivers on each train that determines how often the trains come.

Driverless service eliminates these costs, "break[ing] the connection between frequency and labor costs," as transit consultant Jarrett Walker put it. Vancouver's driverless SkyTrain network, for example, has off-peak headways that would make Americans drool with envy. Riders on the Expo and Millennium trunk line never have to wait more than 5 minutes for a train, even late at night – a frequency that would be prohibitively expensive without driverless trains.

While pouring money saved on labor costs into better service is a boon for riders, it's anathema to transit unions. When London Mayor Boris Johnson proposed a driverless Tube within a decade, he sought to preemptively counter criticism from what he termed "hardline union barons," going so far as to request changes to labor laws to prevent some strikes. (ASLEF, the Tube drivers' union, did not disappoint. "How many votes are a young child's life worth?" the union's general secretary asked of Johnson's proposal.)

Explaining the benefits of Johnson's plan, Daniel Moylan, former deputy chairman of Transport for London, downplays the reduction in wages. "You save a bit of money," he tells The Atlantic Cities, "but not a lot." The real savings, he says, come from eliminating staffing inefficiencies.

Managing people as opposed to computers comes with "additional and unpredictable factors," with Moylan citing occasional sickness and lateness as examples. "Add to that the perfectly understandable need for drivers to have agreed work schedules, meal breaks, and respites to go to the lavatory," he write over email, and scheduling drivers becomes "as large a logistical exercise, almost, as timetabling trains."

"All this would be true in a non-unionized environment," Moylan added, "though unionization often adds further inflexibility."

Don't hold your breath for driverless trains in the United States. While America generally lags behind peer nations in transit, the prospects for driverless trains are especially slim.

The Washington, D.C. Metro is one of the best candidates for retrofitting, but the system's deplorable state makes such an upgrade unlikely in the foreseeable future. Even Metro's automatic train operation system (ATO) – a halfway step to fully driverless operation, where a computer drives the train while the operator still usually opens and closes the doors – was turned off after the fatal Fort Totten crash three years ago.

Being forced to switch off an ATO system for years may be a first in the world of automated trains, but systems across the U.S. face similar challenges of underinvestment and mismanagement. Driverless trains save money in the long-run, but they require upfront capital upgrades and ongoing careful maintenance that most American transit agencies simply cannot provide.

And New York? Fuggedaboutit. Upgrades to Paris's Métro have proven that retrofitting century-old subways for driverless operation is possible (and arguably safety-enhancing, with platform screen doors installed at stations that restrict access to the tracks until a train is safely docked and aligned), and Glasgow's subway, which predates New York's by almost a decade, is only five years away from driverless capabilities.