It is convenient to have a train stop at a station nearest your home. Since trains are crowded, it is even better if the train starts at a station near the home.
All this results in trains having to change tracks. We need synchronization of threads. Every time the train stood waiting for the signal, I kept wondering if there was a better way.
Sometimes a little more discomfort may be the better option.
Suppose, we get rid of some more stops. E.g. after Bombay Central, straight to Churchgate. The fast train from Churchgate to Borivali used to take about an hour to cover. The slow train took around an hour and fifteen minutes. The Virar train which did not stop on stations between Andheri and Borivali took 45 minutes.
So, if we got rid of slow trains and we reduced the number of stops, we could cover the distance from Borivali to Churchgate in about 35 minutes. Since there were very few Virar trains, we would virtually be halving the travel time - or, more importantly, doubling the capacity.
Since the signaling system determines the gap between the trains, faster travel would imply that the time gap between the trains reduces by half. That is two tracks could cover the load.
We could use an additional track to run from Borivali to Churchgate stopping only at Andheri in the morning and in the reverse direction in the evening. This could help in increased peak time capacity.
The 4th track could run a shuttle service stopping at each station between neighboring stops of fast trains.
A little refactoring and a little redesign can go a long way in improving performance and capacity of an application.
Saturday, April 12, 2008
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment