Algorithm predicts and prevents train delays two hours in advance

Springwise: “Transport apps such as Ototo make it easier than ever for passengers to stay informed about problems with public transport, but real-time information can only help so much — by the time users find out about a delayed service, it is often too late to take an alternative route. Now, Stockholmstag — the company that runs Sweden’s trains — have found a solution in the form of an algorithm called ‘The Commuter Prognosis’, which can predict network delays up to two hours in advance, giving train operators time to issue extra services or provide travelers with adequate warning.
The system was created by mathematician Wilhelm Landerholm. It uses historical data to predict how a small delay, even as little as two minutes, will affect the running of the rest of the network. Often the initial late train causes a ripple effect, with subsequent services being delayed to accommodate new platform arrival time, which then affect subsequent trains, and so on. But soon, using ‘The Commuter Prognosis’, Stockholmstag train operators will be able to make the necessary adjustments to prevent this. In addition, the information will be relayed to commuters, enabling them to take a different train and therefore reducing overcrowding. The prediction tool is expected to be put into use in Sweden by the end of the year….(More)”