New Delhi: It’s not something most people would get excited about, but Hiranmay Ghosh is in quite a lather over an empty bottle of Mirinda. “How would a computer system identify this bottle?” he asks, waving it in the air, then tilting it to one side. “Not from its shape. Many other soft drink bottles have the same shape. Neither from the colour of the label, (nor) its texture.”
The answer, as he and his team of researchers at the Tata Consultancy Services Ltd (TCS) Innovation Lab in Delhi found out, lay in the edges of the letters of the Mirinda logo.
“The edges,” exclaims Ghosh, the principal scientist at the lab, “were unique.” Taken together, they constituted an almost foolproof signature for the product. Not only did the process work to electronically distinguish soft drink bottles, but also for a range of other products, from biscuits to movie CDs.
In a matter of a few months, his team put the finding to work. The outcome is “Shopping by Example”, a software that allows buyers to order products by simply clicking photos of wrappers with their cellphones.
So if you want to check out the reviews of a movie before you buy its DVD, all you have to do is take its photograph with your phone. The software will not only decipher the name of the film from the photograph and fetch a review, but also allow you to order it online. The light, the angle at which the photograph is taken or even how crumpled the pack is do not matter.
The software has the potential to revolutionize retail marketing in India, and is ready for commercialization, having been thoroughly tested. But Gautam Shroff, who heads the lab, has kept it aside for the time being.
It was created because it seemed like an interesting idea and a challenge, he says. “We’ll pick it up when either a client asks for something like this or when we can combine it with some of our other innovations.”
Also Read More articles in the series
The idea funnel
That’s the way innovation works at TCS, India’s largest exporter of software services. Ideas, if interesting enough, are never discarded. Ever so often, a lab working in a particular area will come up with a “tangential” innovation; they’ll see it through and then keep it aside while they continue working on focus areas.
Shopping by Example is a product of multimedia analytics, or the complicated task of analysing video, audio and text data to ferret out relevant information. This is one of the two broad areas that the Delhi lab is working on; the other being to make enterprise computing—or computing for business purposes—more accessible and user-friendly.
Out of the rut: TCS Innovation Lab’s Delhi team. Principal scientist Hiranmay Ghosh (centre) and Gautam Shroff (left, wearing a tie), who heads the lab.Madhu Kapparath/Mint
“Think of it as a funnel,” says chief technology officer K. Ananth Krishnan, who presides over the network of 20 TCS Innovation Labs spread across the country, as he explains how they operate. Each lab works on a few broad areas, decided in part by the needs of clients and in part by TCS’ long-term business interests.
Research teams are let loose in these areas to mull, ideate, create applications and write research papers. Hundreds of new ideas flow into the mouth of the funnel. Some are rejected at the outset, but most are let through. Researchers then have the time and freedom to develop these ideas.
“During this time, the researchers are measured only by the number of research papers they publish,” says Krishnan. Some of the ideas take on a life of their own and are developed into separate products; the rest evolve and merge into the focus area stream that flows out of the funnel.
Starting out in 1981 with a single lab that focused on process engineering and systems research, the network has now grown nationwide and employs 400 researchers. Their expertise runs the gamut, from materials engineering, genomics, wireless technologies and information technology (IT) systems for insurance applications to biometrics.
The company registered a consolidated net profit of around Rs7,000 crore in the year ended 31 March, up 33% from the year-ago period. Revenue went up 8% in the same period to Rs30,000 crore.
Krishnan divides the innovations coming out of these labs into three categories. The first includes derivative applications that “help existing businesses to be competitive”.
One of its best examples, he says, is InstantApps, a Web-based technology produced by the Delhi lab as part of its focus on enterprise computing. The technology allows business people and analysts to develop and use custom applications without the need for complicated programming.
This means users can create applications better suited to their needs, and also bring them much more quickly to the market. The technology, being Web-based, also does away with the need to manage hardware and software. At least 30 companies are using it already.
TCS Innovation Labs have come up with nearly 60 such tools, but InstantApps has been the most successful of them all. Shroff says it has led “to a three- to fourfold increase in productivity”.
By September, its simplified version will be available on Amazon.com Inc.’s cloud-computing platform, which offers shared resources, software and information to computers, and other devices on demand.
Breaking the mould
The second category of innovations, says Krishnan, includes those that lead to the creation of new platforms that provide better ways of doing business, as opposed to improving on existing methods.
An example is the TCS BaNCS platform, a comprehensive banking platform consisting of an array of preconfigured, yet customizable products that cover every aspect of banking, capital markets and insurance.
It is among the most well-integrated banking solutions available, and has recently been ranked one of the best in US-based research and consultancy firm Celent Llc’s Core Banking Solutions for Large Banks report, which evaluated the top vendors for global banks.
But the innovations that excite the researchers at TCS Innovation Labs the most are those that are, in Krishnan’s terminology, “disruptive” or “game-changing”. They ruffle existing patterns and hierarchies, or sidestep them altogether; and they lead to massive improvements. “And of these,” says Krishnan, “we have quite a few.”
The best known is the Tata Swach water filter, whose precursor, Sujal, was conceived at the Innovation Lab in Pune in the late 1980s. It came out of a project to find applications for rice husk, which is wasted in enormous quantities every year. “That project was part of our continuing focus on green and sustainable technologies,” says Krishnan.
The technologies initially used in Sujal were perfected in Swach, which uses rice husk ash with nano-silver particles bound onto it, as a base. It’s one of the cheapest filters available in the market; and requires neither running water nor electricity.
In 2009, the Mumbai lab created mKrishi, a mobile agro advisory system that offers agricultural and market advice to farmers using a mobile network that is integrated with local sensors and weather stations.
The sensor network provides farm, soil and crop information to experts, allowing them to give more specific advice to farmers than they otherwise would have been able to. It also provides farmers with micro-climate information and market prices on their mobile phones.
By circumventing the usual systems, mKrishi has brought farmers, research institutes, agro-industries and the government closer, to the benefit of all.
Krishnan rattles off the names of some more “disruptive” technologies. HIP, or the home infotainment platform, combines the functions of a television and a computer, blending visual content with graphics and allowing customers to use their television sets for low-bandwidth video chats and text messages.
At the other end of the spectrum is BioSuite, a bioinformatics application consisting of 79 programmes, which performs the highly complicated tasks of genome sequence analysis, three-dimensional modelling, structural analysis and drug design.
“The social impact of disruptive innovations are much more than their financial impact. We work on these with a 10-year time frame,” says Krishnan.
Avinash Vashistha, chief executive of Tholons, a Bangalore-based IT advisory firm, believes TCS’ focus on innovation differentiates it from other Indian IT companies.
“It’s definitely the farthest ahead in terms of R&D (research and development),” he says. “I haven’t seen such a concerted effort from any other company.”
Vashistha thinks the most important innovations at TCS have been in the areas of process automation and testing. “This has helped them lower costs significantly, and given them an edge over competitors,” he says.
Areas that the TCS Innovation Lab in Delhi is working on are becoming increasingly important worldwide.
Tools such as InstantApps, which have “in-memory analytics and ease of visualization”, have made it an “attractive proposition to bypass IT”, according to Dan Sommer, senior research analyst at IT research and advisory firm Gartner Inc.
Sommer says vendors in this segment continue to be the fastest growing in the business intelligence market. Gartner expects IT spending to grow 5.3% in 2010.
The next big thing
Back at the Delhi lab, Ghosh is on to the next challenge: How to extract information from video feed? But the task of discerning semantics, or meaning, from the muddle of voice, text and images is devilishly complicated.
They’ve created an application called Multimedia Explorer that builds a “summary” of the video footage by sampling it at regular intervals and selecting what it thinks are a representative collection of scenes. The current criterion for selection is transitions in colours across scenes. It works better than other criteria such as illumination, but Ghosh says, getting excited once again, “it’s just one of the many criteria that might work”.
The application, part of a larger project that the lab is doing for the government on the automated analysis of news broadcasts, is still at a nascent stage. Many other similar applications are likely to go into tackling the big question that Ghosh is trying to address.
This is also one of the areas Krishnan would like his teams to focus on over the next few years. “The big story for us,” he says, “is going to be in deep analytics, analysing information to gauge opinion, customer feedback and market behaviour.”
Some of the other areas he thinks will be important over the next couple of years are drug discovery and genetics, and the use of social networking for enterprise.
But this does not, he says, preclude research in other areas. Eclecticism is what seems to drive innovation at TCS. “My biggest goal,” as Krishnan puts it succinctly, “is to encourage a creative dissatisfaction with the status quo.”
Tata Consultancy Services Ltd Started operations in: 1981
Made in India: InstantApps, which allows users to develop customized business applications without knowing complicated programming; Tata Swach water filter, one of the cheapest filters available in India; mKrishi, which offers agricultural and market advice to farmers on mobile phones