Analytics, in layman terms, is the process of examining past data to gain insight to
the future and take decisions based on the same. Hence the obvious necessity for
analytics is historical data. The Indian Railways is a state-owned enterprise and is
one of the worlds largest railway networks. It serves almost 30 million passengers
on a daily basis and hence it goes without saying that there is a massive data bank
within the organization. This in turn implies there is immense potential to use
analytics to generate information that can benefit both the organization and the
passengers.
The limited number of seats in trains usually leads to a large number of people
being waitlisted. This uncertainty regarding whether the tickets would ultimately be
confirmed or not means that the passengers have to make a decision, whether to
wait in hope of the ticket getting confirmed or to arrange for an alternate mode of
travel. In such a scenario, analytics can come in and help passengers in making the
decision. Based on the past data regarding the waitlist movement on the same day
the previous year or month or week, a system can come up with a predicted
number of waitlisted tickets that would get confirmed. This will help the customer
get an idea of the possibility of the ticket being confirmed and enable him or her to
plan the future course of action accordingly.
The arrival time of trains, a critical component of multi-leg journeys, is often a point
of concern for passengers. Analytics can help out in two aspects in this regard.
Firstly, based on available data of arrival times, an application can give passengers
the most probable expected arrival time other than just the earliest arrival time
which is usually given. Secondly, this probable expected arrival time can also be
determined real time by attaching a GPS in each train which will relay the current
location of the train. This can be coupled with the historical data regarding the time
taken by the train from that point to the destination to determine the expected
arrival time. Another dimension can be added to this by using the data collected
from another train that might have gone through the same route in the very recent
past (maybe an hour or two ago). Trains running on the same route in the opposite
direction present a possibility of route conflict and that too can be factored in. This
information can also be made available to passengers through a mobile app.
Every year in the railway budget, the rail minister announces some new trains to be
introduced. In this context too, analytics can be used to determine which new routes
should be introduced. The data about passengers undertaking multi leg journeys
can be analyzed and based on the results, the demand for specific new routes can
be forecasted and trains can be introduced accordingly. Some routes may have high
demand only during certain times of the year. This period of demand can be
forecasted using past data and special trains can be announced which only run
during those periods of high traffic. For example, there is huge demand for tickets
from Kolkata to Puri during the Christmas-New Year season. Special trains are often
announced a few days prior to the journey. But this arrangement means that some
passengers are lost who are not willing to wait till the last moment and therefore
References:
http://blogs.wsj.com/cio/2012/03/30/union-pacific-using-predictive-software-toreduce-train-derailments/
http://www.railnews.co.in/smart-railway-market-it-to-drive-the-future-railwaysystems/