Better regulation of trains, better information for travellers, and a better mobility offer: this will all be possible with the application of artificial intelligence to rail operations.
Rail operations: a complex machine
Rail operations comprise managing 15,000 trains a day before, during, and after their commercial service; defining timetables; rotating rolling stock and supervising personnel; providing travel information and recommending itineraries; and managing traffic flows into, out of, and at stations.
With passenger volumes continually growing, transport offers expanding to keep pace with demand, and infrastructures sometimes at full capacity locally, operating a rail system is becoming increasingly complex.
Travellers expectations are changing too: they want quicker responses, more personalised service, and greater reliability.
Under these circumstances, the operation of rail systems is evolving. Since 2016, the teams at TECH SNCF have focused on finding ways to optimise the capabilities of the rail operations “machine” for the rail operator’s benefit. The project is called “Augmented Operations”, and its aim is to improve the rail system’s operational performance and the quality of service provided to its users.
Leveraging artificial intelligence
The TECH SNCF teams are exploring the potential of artificial intelligence in three areas: traffic optimisation, response to unforeseen events, and simultaneous management of traveller flows and rail traffic.
Artificial intelligence is going to “augment” the rail operator’s capacity to deal with the growing complexity of the rail system in three ways:
1. Real-time processing of all rail operations parameters with resource and flow management algorithms
2. Detailed visualisation of a situation at a given moment and its projected evolution
3. Decision-making tools that do predictive simulations, make comparisons, and recommend solutions.
Augmentation, not substitution
Artificial intelligence will never replace rail operators, but it will help them to make better and faster decisions by analysing a much larger set of parameters than the human mind can process and doing so without ever compromising safety.
For artificial intelligence is to deliver its full potential, data visualisation and new man/machine interfaces are crucial.
A complete simulator
That is why the TECH SNCF teams have created and developed a complete virtual rail environment, the “Network Shop”, to simulate operational situations in a digital twin of the actual rail network.
With this simulator, SNCF is able to test and refine the sets of algorithms as well as take into account the human factor (the operators’ comprehension of and confidence in the data, ergonomics, etc.).