Algorithms to predict train delays due to fog

NEW DELHI: With fog likely to settle in soon and train delays also to shoot up, Northern Railway is planning to use an empirical algorithm-based system to accurately predict train timings and inform passengers accordingly.

To prevent congestion on platforms, a railway official said, it would make use of SMS alerts to inform passengers on delays and a probable good time to reach the station.

“We used a similar system last year and now this has been perfected to ensure accuracy, which is over 80% currently. The system puts in place the distance and compares the speed and time a train took last year in a similar fog setting to arrive at a probable time,” said RN Singh, divisional railway manager, Delhi division.

In addition, fog safety devices are also being utilised to alert the loco pilot of an approaching signal which is almost 2km away. “If the driver gets no alert, he knows that there is no signal for around 2km, which can help him run at a higher speed. Last year, while the trains were running at a speed of around 60km/hr, they can now run at around 75km/hr in fog, which will further reduce delays,” Singh added.

The railways had introduced the fog-based SMS alert system last year in November, which used mobile numbers of all the reserved passengers from the Passenger Reservation System (PRS) and information from the National Train Enquiry System (NTES) to send alerts. Passengers were also updated in case of rescheduling or cancellation.

“This will ensure people not only get an accurate idea of when to come to the station, but also to not come too early and crowd the station,” said a senior railway official. In addition, RPF personnel have also been directed to guide passengers to the waiting rooms to ensure platforms stay relatively less congested, the official added.