Background
Freight railroads, manufacturers and technology companies are exploring driverless locomotives, autonomous freight cars and automated track inspections to improve labor efficiency and manage infrastructure. The effort is unfolding alongside a longstanding dispute over train staffing. A single freight train commonly has two crew members, while crew sizes have historically been shaped through collective bargaining. Railroads have explored one-person crews to reduce costs, while unions and some lawmakers have supported a two-person minimum on safety grounds.
Driverless transit trains already operate in cities, airports and other controlled settings. Long-haul freight is different. Freight trains move mixed cargo across long distances and may share networks with passenger trains. Their routes can include branches, interchanges and road-rail crossings.
One long-haul example discussed in the Congressional Research Service report is Rio Tinto’s AutoHaul system in Western Australia. The company operates a roughly 1,200-mile rail network linking 18 iron ore mines and two seaports. Trains are centrally monitored from an operations center in Perth, nearly 1,000 miles away, and generally need people in the locomotive cab only when entering or leaving the ports. Rio Tinto has credited AutoHaul with increasing average train speed and reducing fuel consumption.
That success has important boundaries. The network is largely self-contained, has few branches or interchanges, runs through a sparsely populated area, has few road-rail crossings and carries almost entirely iron ore, with no passenger traffic. It offers evidence that autonomy can work in a predictable, dedicated environment. It does not establish that the same model will transfer cleanly to the U.S. freight network.