Making a mark on the scientific software map of the world
Benedict Paramanand

"India must do more in software creation. At the most basic level, you must try and own your intellectual property. Now, intellectual property created through the work of Indians is owned by others," Acer chief Stan Shih said recently. Shih's thoughts are shared by many but their attempt to translate it into action seems feeble in the midst of companies clamouring to make a quick buck from Y2K and Euro conversion.

Can India dream of becoming a software superpower in the true sense of the word in the foreseeable future or will it continue to play second fiddle to multinational companies? Unfortunately, experts say, the Information Technology Action Plan has not given á strategic focus for promoting world class high-end software products.

While developing business software products seems a long haul, India seems to have a better chance of making a mark on the scientific software map of the world. The major inhibiting factor for developing business solutions software packages and products are high development (20 per cent) and marketing costs (80 per cent) according V Chandrasekaran, president, Wipro Systems.

The scenario for scientific software products is just the opposite. According to Dr M Vidyasagar, director of the Centre for Artificial Intelligence & Robotics (CAIR): "Marketing plays a smaller role in scientific software compared to business software. This is because the clientele is very discriminating and the base is small." The major cause for optimism seems to stem from the fact that a lot of scientific application software is lying idle, waiting to be packaged and sold.

Under Dr R A Mashelkar, the market savvy scientist and head of the Council for Scientific & Industrial Research (CSIR), India is making a beginning with a dedicated centre in Pune for identifying and packaging scientific software into marketable products. The high-end software products developed in the 40-odd labs around the country finally could have a chance of hitting the marketplace in a big way.

Scientific solutions software and services constituted 9.6 per cent of the $ 2 billion software industry (a mere 2 per cent of the world share). The world average is said to be just over 10 per cent.

But how does one get scientific software off the lab and into the market? Knowing the pressures on the private sector for generating quick returns and that scientific software development is a long drawn out process, this task world over is generally undertaken by the scientists themselves with active support from the administration. "Indian scientists are in any case in the fore front of world technology unlike those associated with business software," Dr Vidyasagar adds.

Both Silicon Valley and Route 128 outside Boston turned out to be citadels of technology due to the system which allows professors/scientists from Stanford/Berkeley and MIT to pursue their cutting edge ideas to their logical conclusion in the marketplace. Says Vidyasagar: "It is not that Indian scientists are lacking in ideas having huge commercial potential but the present system of mistrust and the `government knows best' attitude prevents all such ideas from ever making it to the market."

Government rules governing the scientists' career seem outdated and out of tune with the global scenario. Scientists say there is not enough incentive for them to commercialise their expertise. Academics, in principle, are allowed to undertake consultancy on á revenue-sharing basis, but rules prevent them from starting private companies for commercialising their research findings.

Realising the need for a liberal compensation package, CSIR raised the scientist's returns to 40 per cent from inventions. "With this a scientist's income has seen a raise of 50 to 100 per cent a year," says Dr Mashelkar. Prof N Vijayan, chairman of the division of biological sciences, IISc, however, seems happy with the money. He says: "The returns are reasonable."

Scientists' bodies have asked the government to grant them some independence as is done in the US. Among other things:

* Allow them certain to undertake consultancy for at least one day a week.

* Permit them to set up private firms for commercialising technologies developed with equity participation by the institution.

* Encourage the filing of patents by fully underwriting the costs of patenting and sharing the patent revenues with the scientist.

The main problem seems to be lack of motivation, inadequate market intelligence and the poor understanding of the needs of the industry. Says Dr Srinivas Bhogle, head, information services and systems, National Aerospace Laboratories (NAL): "We often don't know who needs our software. This is partly because we are not trained to think that way and partly because we don't care. Also, we need companies which have faith in us."

Adds Dr Balakrishnan, head of the supercomputer research centre, (SERC) at IISc: "Primarily, we have not recognised our own worth." Scientists generally think that their job is over once a particular solution has been found. "When we think the job is over, it actually begins," Dr Bhogle adds. Says Dr Srinath Kumar, the hero of Control Law software for the Light Combat Aircraft: "Designers have no motivation in the business aspect. What we need is a system which takes over from where they leave."

The scope for commercialising scientific software is limited as á lot of these are in public domain. For example, a lot of good work on molecular modelling and crystallography done at the Indian Institute of Science is put on public domain. Dr Vijayan however believes there is ample scope for commercialising software developed at IISc if it is packaged well and is user friendly.

One of the main functions of the IT action plan is to offer market intelligence to whoever needs it. Says Dewang Mehta, executive director, National Association of Software and Service Companies (NASSCOM): "The market intelligence system which is being strengthened will be of great use for scientific software developers.''

The much talked about industry-lab interface is yet to start in real earnest. Says Prof S Chandrashekar, a former scientist at ISRO and now teaching corporate strategy at Indian Institute of Management, Bangalore: "There is a lot of software in the lab. How much of it is relevant to industry is debatable. The higher the technology fewer are the customers. The real challenge is to understand what the industry needs and the initiative should come from the industry itself."

Says Infosys chairman and managing director, N R Narayana Murthy: "We need higher interaction between the labs and the industry to bring their products to market. Productionising an idea is a long process and expensive. Only when the industry sees value will it show any interest."

Allowing scientists to join the boards of IT companies, which is not allowed now, is seen as a first step for greater industry-lab interface. Dr Vidyasagar, a member of Andhra Pradesh chief minister's IT task force, has even suggested that scientists be allowed to perform a dual role of doing research in a government lab and run a company. He notes: "This is required as it gives a sense of security to a scientist at a time when the mortality rate of hi-tech companies is very high."

The IT action plan seems to have showered financial incentives on business software companies. However, the availability of venture capital for high-risk-high-return firms is still one of the major handicaps. "The banks even question expenses on R & D and the auditors ask questions about foreign trips," a senior executive of a home-grown technology-driven company said.

Recognising the creative potential of Indian scientists, most global IT majors have already set up design and development centres in India and particularly in Bangalore. "But the catch is that MNCs will own the rights of the products created with Indian brain and sweat. We seem to have already lost the bus on IPR in software. To catch up we have to do it ourselves," says a senior ADA scientist.

ADA has so far commercialised three software that it developed for LCA and launched in the US first. "We prefer to launch our software first in the US after tying up with a company there. This way its worth in India goes up," says Dr Kota Harinarayana, director of the LCA programme. ADA plans to commercialise 10 software packages shortly.

The Hindustan Aeronautics Ltd (HAL) has set up a strategic business unit to commercialise software developed in its divisions. HAL's software joint venture with British Aerospace is aimed primarily at marketing its software abroad.

Several Indian IT majors have started interfacing between the labs and MNCs. While this seems good in the short term when the labs learn to operate on stringent quality and schedules, for sustaining innovation in the long run scientists need to get a good share of the cake.

A few success stories of scientific software hitting the market

---------------------------------------------------------------

1. Parallelisation of molecular dynamics code by NAL for Hitachi

2. Aircraft wake vortex modelling at NAL. Software sold to Civil Aviation Authorities of UK.

3. FEPACK of NAL - software for analysis of structures sold to many domestic customers.

4. Autolay Software by Aeronautical Development Agency (ADA): Bought by Boeing, Airbus and Hughes.

5. GITA - by ADA - Graphical interactive three dimensional analysis software. Technology Associates Inc of US and Boeing use it for CAD\CAM application

6. FEAST - ISRO - Application in structural analysis

7. MALAR - ISRO - Software helps translation from one language to another

8. SACIMAGE - ISRO - Remote sensing applications

9. Cyber-Profit-PRedict - C-DAC - financial trading simulator system