Indian Railways dilemma

Present passenger reservation system is time tested and very reliable. It handles over 1 million transactions per day. The original PRS software was developed in FORTRAN in 1985. The original PRS went through three versions. Thereafter, in 1997, the software was largely rewritten in C by CRIS. So the Indian Railway has changed the PRS earlier too but in a phased manner to make sure system gets enough time to stabilize with each change and does not impact existing business. It took around 14 years to roll out all India implementation for current PRS from 1985 to 1999.

On average the system collects over Rupees 200 million in daily revenue.  Average uptime of the system is around 99. 2% and average transaction latency is less than 3 seconds. PRS has complex business logic. Complexity arises from the combination of handling types of trains 8, types of quotas 40, types of classes 9, types of concessions 199, and types of coaches (Railway Carriages) 123. Indian railway handles around 16 million passengers per day in over 9500 passenger trains. Nearly 1 million passengers per day get their accommodation reserved through the PRS.

As of now only 15% of journeys are undertaken using reserved accommodation, it still adds up to over 1 million transactions per day on the PRS. The IRCTC developed web based portal for internet based reservation is available at http://www. irctc. co. in. This is among the top e-commerce sites in India. CRIS developed the back-end interface from PRS to IRCTC booking portal. 10. CRIS has set up a very successful website for passenger inquiry, which gets over 6 million hits per day during the peak travel season.

It should be noted that these inquiry hits PSR at backend to get the latest status so overall it is handling around 7 millions hit per day. Major decision issues in the case Which option to choose, incrementally enhancing the capabilities of the existing PRS to serve the emergent dynamic requirements of Indian Railway customers or go in for a wholesale switchover to a new system. Recommendation/inferences on the major issues 1. Indian Railway need not look outside beyond its own experience. They are the pioneer in implementing technology out of all Indian government departments.

They should look back at their own experience as how it took around 14 years to reach where they are today. The existing system is considered a very successful because it was allowed to take time to grow slowly and become stable instead of making it available overnight. Rushing to develop a new system will definitely not help Indian railway as it will run into a high risk of failure. 2. It is not worth taking risk of 200 millions of daily revenue. No organization in the world will go for a big bang by replacing a very stable and tested system with new system without proper backup plan.

If any major failure happens, there will be wide spread public backlash. 3. To have 99% plus uptime, any new system takes couple of years and it does not happen in weeks. So this point strongly supports that CRIS should develop new system incrementally. 4. Indian Railway should definitely choose to develop a new system, which is more flexible in handling changing business rules, easy to integrate with partner organizations in future, scalable enough to interface with different devices like mobile, hand held devices, or any other devices.

New system should be rolled out in phases similar to how it was done for old system. Starting with one station and few trains should be made available on new system and then slowly migrating more trains and stations. This will gives new system enough time to stabilize and get tested for increasing load rather than testing it for 7 millions plus transactions and taking a huge risk of a big failure. 6. Indian railway should continue with its existing system by enhancing it incrementally for only must needed features till new system is developed and proved to be reliable over time.

This will make sure Indian Railway is never on the risk of loosing serious business and can always use existing system as backup till new system stabilizes. 7. Given the new system has to develop complex business logic it will not be easy to implement everything right at first attempt. So IR should go for incremental development of new system. 8. Indian railway should definitely re-use some of the best features of existing system to control cost of developing new system. The existing features like proprietary networking protocol to share information between 5 clustered servers connected through its own leased lines.

This will reduce the security threat since it is not common standard platform. 9. Reuse network protocol for security, flat file based database management, server clusters rather than one centralized deployment. File based database system are still in use for large ticketing system in worldwide and proving a batter option compare to costly DBMS. 10. The existing system can take the load of 7 millions hit per day as mentioned in key observations. This proves that existing system is capable of handling huge transaction volume without impacting performance.

So keeping it up and running till new system is fully tested and found reliable will always give railway a upper hand and allow them to have sufficient time to test new system till it stabilizes. Evaluative Criteria: – Benefits: what Indian Railways can get for remains the existing system or adopts the new system. – Impacts: What will happen if IR decides to keep the old system (advantages or disadvantages for keeping it); what will happen if IR decides to adopt a new system (advantages for adopting it or disadvantages if it lead to further problems)

– Strategic: The system that might fit the management Strategic and business plan. Recommendation: Although there are a lot of benefits if IR chooses to adopt the new PRS, I will recommend that IR to remain and enhance the existing system. The old system is considered very successful as it is stable and time-tested after 14 years of operation. Compared to the new PRS, the old system is familiar to both managers and staffs and is able to minimize uncertainty.

While the new PRS brings an opportunity into IR, the threat is obvious as well. It is risk taking for the company to replace a stable and tested system with new system without proper preparation. If any failure happens, there will be many inconveniences for travelers and even lead to public backlash. In addition, almost every new system requires years to fully develop. IR might lose millions or even billions during these years. I total agree with you Asya, adopting the new PRS seems a little bit risk taking to me as well.

As the change of system is not a problem that needs to solve immediately, IR could slow down the change process by running both systems (Starting the new PRS with one station and few trains and then slowly migrating to more trains and stations). This might give time for new PRS to stabilize, fix any small or major problems before it completely replace the old one, and minimize the chances of failure. MS: As long as it doesn't increase costs by running both systems or lead to overbooking / double booking if the two systems can not communicate properly.

If they can keep the old system and the new system synchronized properly it could keep them from having a giant catastrophe with the entire collapse of the system. Yep, the cost is a big issue in this situation. It is interesting because I just associate this to Riley's idea of funding (which I believe is a very good point). If IR has support from government, the cost of operating tow system should not be a problem. If not, I think I will still consider the smooth changing process to be the first priority as IR should be responsible for all travelers.

Attitude/willingness of IR employees to adopt a new PRS and following that, their perceived ability to do this efficiently (with a reasonable amount of training). JC: I really like the idea of that as well and I think the attitude or willingness of employees is very important for any company which planning to make a change. To further this point, I am thinking how about the attitude of the public as they might be another group of people that influence by both systems. IR might concern whether travelers are willing to accept the inconvenience during the system changing period or they are hoping to maintain the existing situation.