26 March, 2015
If you are used to take the metro, you might sometimes have thought you were packed in like sardines. Other times, you might have found yourself virtually all alone in the vehicle. All this happens because it is difficult to plan the metro service as it is affected by different disruptions as the day progresses:
– Randomness in passenger demand (both in spatial and temporal)
– External factors that also interfere demand, as the weather or big city events
Developing flexible control strategies is the way to reduce the negative effects of service disturbance as they allow the operator reacting dynamically to real-time issues. This is possible if responsibility for operation management of the trains is transferred from the driver to the train control system, is what is known as metro automation and there exist various degrees: a Grade of Automation 2 would correspond to a system in which trains run automatically from station to station but a driver is in the cab, with few responsibilities, as for examples, door closing. Grade of Automation 4 would refer to a system in which vehicles are run fully automatically without any operating staff onboard.
Metro automation is not a futuristic approach, but a reality; there are today 48 automated metro lines in operation in 32 cities of all sizes and demographic environments. As Ramon Malla (Director of Automated Lines in Transports Metropolitans de Barcelona) explains in this article “automated metro lines are no longer a trend to be considered or debated, they are an indisputable reality and their benefits are clear”; hence now is time to talk about how companies must be adapted to get the most of it.
As the main objective within metro automation is offering a service capable of reacting dynamically to passengers behavior is crucial that supply and demand are permanently connected. Enormous progress has been made from the side of supply so far, and is time now to react from the side of demand. It is therefore essential to monitor this demand all the time, as well as analyze historical footfall traffic of passengers and identify which variables affect its fluctuation. Only by guaranteeing that (dynamic) service matches real-time demand, metro automation will be attractive enough to overcome the obstacles to convert traditional lines.
In this link you can check an interesting project that the Cambridge University is doing for the London Underground in which Counterest People Counters are being used to monitor passengers flow in real time. The main purpose is to develop a predictive model based on how people counting fluctuates along the time. This will make possible to design a service aligned to real demand and, eventually, build up a dynamic scheduling based model.