Computers go off track
The National Informatics Centre (NIC) and Indian Railways" Centre for Railway Information Systems (CRIS) have locked horns over freight control software. CRIS has chosen the Traffic Reporting and Control System (TRACS) -- a freight operation information system (FOIS) developed by the Canadian National Railways as long ago as 1967 -- over the systems with voice and data transmission options offered by NIC. The results of the ongoing first phase of trials of TRACS are expected soon.
This is the second time in the past 6 years that CRIS has rejected an NIC proposal for a freight operations information systems. In 1987, NIC"s proposal was turned down on the ground that it had limited voice access. NIC claims that the chosen Canadian package is not only outdated, it has critical limitations and is unsuitable for Indian freight operations. Adding to that is the fact that it is outrageously priced and has a high day-to-day operation cost. CRIS, on the other hand, alleges that NIC lacks the wherewithal to handle a complex, giant system.
NIC director-general N Seshagiri retorts, "In the Indian Railways, most information is shared through voice communication (telephones). That is an outdated way of exchanging information. In 1987, we created a package that could send most freight operation information in the form of data. Our proposal was rejected solely on the ground that it had limited voice access. But in the latest proposal submitted by us, the limitation was overcome."
The NIC proposal is still databased, but the channels that were earlier used to transmit only data have been modified to accommodate voice. Says Seshagiri, "Using techniques of compression and decompression, voice can now be sent through channels carrying data." He says that there is a need to do away with the old system of shouting orders down the phones, which is more expensive than sending data. "For the same information, the cost of data communication can well be a hundred times less than voice communication," he adds.
Decentralised operations
The package is divided into 2 systems -- the central system and the zonal operations information system (ZOIS). ZOIS provides a stored programme interface to zonal users. The responsibility for obtaining timely and accurate reports from the field user and for its delivery to the central system lies with ZOIS. Information from the central system is received, stored, processed and delivered by the zonal system. To suit the requirements of individual zones, 2 categories of zones were considered. The first includes central, eastern, northeastern frontier, northern, northeastern and southeastern railways. The second consists of the rest of the divisions.
According to Seshagiri, "The NIC system uses a virtual private network, which is like an independent dedicated network without interference from other global users." Says Keval Kishan of NIC, "Given the flexibility of the NIC package, it is possible to develop value-added features for assistance in planning, monitoring and decision-making."
Seshagiri says, "The backbone of the NIC proposal, which is hooked to the NICNET Information Highway (a network of communication carriers set up by NIC), is satellite communication. This is an advantage over the Canadian package, which is dependent on a telecom system. The satellite communication facility makes the network immune to natural calamities as well as the quirks of weather."
The NIC package was offered to the railways with speeds varying between 16 kilobytes per second and 2.2 megabytes per second. The package has the potential to process more than 6 million instructions per second. Seshagiri says, "The technology offered by NIC can deliver low-cost solutions for high-speed requirements."
Antiquated language
Experts say the Canadian package uses an outdated language. Dhiman Dasgupta, a software engineer with the Computer Maintenance Corporation, points out, "TRACS is written in TOPSTRAN, which is a version of FORTRAN. The system was developed by IBM for the Canadian Railways, which was highly containerised and piecemeal even in the "60s. It commits the user to a technology that is compatible only with IBM mainframes and requires an extensive telecom network that will cost more than Rs 1,000 crore." And a CRIS engineer confesses, "TRACS does not maintain a trackwise inventory of rolling stock within yards and other activity centres, something that is necessary for getting a complete picture of the yard and reporting of movement."
A senior railway official says, "The centralised system accepted by CRIS in a way goes against the emphasis on decentralisation. It also makes the system prone to total collapse, while in the distributive pattern, the individual unit that develops a problem can be isolated and the rest of the system can continue functioning."
Critical anomaly
CRIS officials refute NICs contentions. They say that the voice factor is important in India, as people are not used to databased communication and add that the claim that TRACS would be much more expensive than the Rs 267-crore estimated cost of the NIC system is hogwash. R Sharma, a software engineer at CRIS, and Dewang Mehta, head of the National Association of Software and Computers, say, "NIC is not considering the huge investment the government has made in creating its facilities. The cost of research and the use of existing software should be added to the cost estimates."
CRIS executive director A J Kumar said, "I am not bothered about the NIC claims as they do not have the wherewithal to handle such a project and the Canadian package is not outdated. Even if NIC modifies its package now to a voice-dominant one, we cannot switch back."
Such trading of charges is old hat. However, Mehta says, "It is surprising that CRIS and NIC do not come together to create an indigenous system instead of buying from outside and then squabbling. Imported software will always be plagued by problems of ready applicability and high costs. It is also a great disincentive for the private sector, which has made rapid strides over the years."
Work on a computerised freight system for Indian Railways has been going on for the past 10 years, with the objective of making available data about freight operations, facilitating better yard management, terminal management, locomotive repair, wagon repair, fuel management, safety management and revenue accounts at both the activity centre level and along the network.
In 1982, it was decided that the freight information system would be a combination of centralised and decentralised systems and that functions like the control of wagons and locomotives and trains would be implemented by adapting a proven FOIS. Decentralised functions like yard management, locomotive and wagon repair, crew and fuel management would be developed and implemented indigenously with minimum foreign support. In July 1985, the Central Organisation for Freight Operation Information System (COFOIS) was set up to prepare the project report and finally, in 1987, CRIS was created.
However, over the years CRIS has been more in the news because of the wrangling with NIC than for its achievements. The squabble has delayed the installation of FOIS. User preference is changing in favour of road transportation because of the pressure on the system, according to S K Harit, an official with the Northern Railway"s traffic section. Further, there are wastages and loss in revenue which is transferred to the user every year during the budget. This is where a FOIS system that was proven and simple was required. One just has to wait and see whether the Canadian package lightens the load or adds to the deadweight.
Related Content
- Poverty mapping: innovative approaches to creating poverty maps with new data sources
- Locust collective motion and its modeling
- Wristband from e-waste to track carbon footprint
- Climate Change And Corporations: Apple, Dell, HTC And Others Launch Effort To Reduce Carbon Footprint Of Global Supply Chain
- ‘Project Lakshya’ streamlines LPG supply
- Eye in the sky' spies on illegal waste dumpers