Rail transport is already making significant strides toward autonomous trains. In fact, in Western Australia the heavy-haul railway system has been completely transferred to autonomous driverless operation, thus making it world’s first fully-automated mainline rail network. The operator Rio Tinto is already reaping the advantages of full automation in terms of cheaper operational costs, shorter route times, bottleneck elimination and increased productivity.
Negotiations with North American railways about automating freight train operations have been started by Hitachi STS, creator of the West Australian system. The challenge will be significantly bigger in North America, where many freight and passenger operators share the same rails, trains of all types and weights run on the same lines, and there are numerous connections, yards, and as well links to industrial units.
In Europe similar challenges also exist. Nonetheless, progress is being done here as well. After 18 months of study, the French National Railways (SNCF) conducted its first test run with a remotely-controlled locomotive-hauled autonomous train, as part of a mission to build driverless passenger and freight train prototypes by 2022.
Metros have been fully automated since 1981, when the first line opened in Kobe, Japan. According to the International Association of Public Transport (UITP), there were 64 fully-automated metro lines totaling 1026km in service in 42 cities at the end of last year. UITP predicts a significant acceleration in the development of automated metros with total length expected to triple by 2023 to exceed 3000km. Fully automated metro lines are expected to account up to 48 percent of greenfield metro projects by 2022, which is 10% increase compared to 2018.
However, there are critics about slow progress in automation of existing lines. Converting a line to automated operation is a challenge. The fundamental issue is not cost but difficulty of administering the project and suppliers must offer operators more confidence. Automation benefits include greater capacity, availability and operational flexibility, elimination of human error, reduced capital and operating costs and lower energy consumption.
Despite the thirty-eight years of expertise with autonomous operation there is still room for improvement. Failures on driverless lines are extremely rare though when such happens it could be dramatic. Operators must be able to reduce incidents and recover operations rapidly than now to avoid stranding trains in tunnel. Artificial Intelligence (AI) has to be projected to improve the operation of automated metro lines with implementation of predictive maintenance. Self-driving cars technologies are also introduced to rails. Self-position detection, sensor fusion for environment recognition, multiple driving assistance systems, sensors for human behavior prediction and platooning are among them. Rail transport could preserve its lead over road transport in autonomous operations only by introduction of advanced automation technologies.