Rediff.com« Back to articlePrint this article

India morphs into global R&D hub

Last updated on: July 21, 2007 12:56 IST

Srini Koppolu was delighted when Wal-Mart urged its top 100 suppliers to start using a Radio Frequency Identification microchip that is becoming increasingly important for product security and identification in the retail industry.

This was followed by Gillette deploying RFID chips to stop shoplifters from stealing razor blades. "Even Benetton is planning to weave RFID chips into its clothing products, to track them worldwide," Koppolu says excitedly.

Admittedly, that's a good thing for manufacturers and retailers, but why can't Koppolu stop smiling? Because, as vice president and managing director of Microsoft India Development Centre, he is one of the strongest expounders of RFID technology.

Get him talking on the work the India centre is doing on RFID and the Microsoft veteran responds animatedly.

"What if merchandise could talk? What if a shirt or hammer or CD player communicated automatically and not just when its barcode is scanned? Imagine every item talking about itself as soon as it pulls into the warehouse."

Talking? No, Koppolu isn't loony, not by far. But having kick-started the RFID project from Microsoft India Development Centre, he wants to tell you there's more to come. "The best part will be when products pass their expiry date and start to scream, 'I am too old, replace me'."

Koppolu wasn't always an IT innovator. Like any other aspiring IT professional he'd opted to go to the United States to do his masters in computer science for 'purely better job opportunities'.

But by the end of the first semester he was writing line after line of software code. Koppolu spent 10 years at Redmond, but in August 1998 he was back in India 'not because I was homesick but because I was convinced that India was ready for a global research outfit'.

He shifted roles from being a Microsoft technical manager to an organisational leader and his vision, the sprawling 42-acre Microsoft India Development Centre, is today a key part of Microsoft's future strategic direction.

At Hewlett-Packard, the technology giant estimates that innovative research and development has an addressable market of nearly $1.1 trillion. It invests nearly $3.5 billion annually on R&D, a major chunk of which finds its way to labs in HP (India).

Reasons Ajay Gupta, lab director, "The R&D emphasis is on countries like Russia, India and China because the next billion customers will be from these BRIC countries."

HP Labs in India has staked out areas like mobile and personal computing, networking the paper world, and sponsored services. Dr S Ramani, director for science and technology, and his team have developed 2D (two-dimensional) barcodes to prevent data forgery or manipulation of printed text.

In order to print digital signatures on paper documents, the document needs to be made machine-readable with the help of 2D barcodes where the data and a digital signature are encoded.

Ramani estimates the potential savings for paper documents with a 2D barcode as substantial. "Forgery is rated among the top three sources of fraud in India and only 50 per cent of instances are identified through internal audits," he points out.

The company is in talks with several government institutions and agencies for deploying machine readable documents. "A pilot project is underway to issue secure documents such as experience certificates or identity cards to employees," he adds.

Imagine, he enthuses, these documents can enable any citizen to get an authenticated government document from an Internet kiosk or in their own homes. How? The request for a document is transmitted from an Internet kiosk to a central document issuing authority. The request is processed and a digitally signed, barcoded document is returned to the kiosk.

Unearthing specific consumer needs in emerging economies definitely aids innovation for these R&D majors. For instance, how does one evaluate the language skills of hundreds of contact centre employees? How does one rate an individual's ability to understand English spoken in a myriad international accents?

For a burgeoning outsourcing industry like India's, it comes with associated problems. In an effort to solve the recruitment problems for its BPO arm, IBM Daksh, its India Research Lab researchers came up with a web-based interactive tool called Sensei (Japanese for teacher).

Researcher Om Deshmukh describes it thus: "Sensei checks and rates the grammar, pronunciation, diction and comprehension of voice-based call centre agents by requiring them to respond to simulated situations — then automatically rates their performance and highlights areas for improvement."

Likewise, HP Labs in India identified power outages as a key factor limiting the access and utility of computers in rural areas, so it designed a community PC that can run on car batteries.

While most Indian R&D outfits are yet to come up with significant innovations in their product lines, Indian development units contribute more than just a trickle to the process. Intel is probably the best epitome of this practical emphasis. Its India Development Centre is Intel's largest non-manufacturing site outside of the US. The chipmaker takes a highly disciplined approach when selecting the projects it backs from India.

"We do no research that can be classified as cheap," insists Rahul Bedi, a director in the India office. Most of Intel's latest chips and processors have been designed extensively by its researchers in Bangalore. Intel's India Development Centre employs 3,000 staff of which close to 2,900 are devoted only to R&D. Intel India is responsible for conducting over 800 invention disclosures and has filed 50 patents to date.

With academicians embracing the corporate world, there seems to be no dearth of professionals. For instance, an assistant professor of computer science for four years at Yale University, P Anandan is best recognised as the developer of video stabilisation technology for ground and airborne video surveillance.

During a research career of two decades, Anandan did pioneering research in video motion analysis before he joined as managing director of Microsoft Research India and operations commenced in January 2005. "I worked for months out of makeshift offices in Bangalore and struggled to get employees onboard who had PhDs," he says.

For Microsoft Research, innovations for markets like India have revolved around developing computers for an illiterate population. Principal researcher Kentaro Toyama and his team have developed the prototype of a system to connect illiterate domestic workers iwith families seeking their services.

"The system typically comprises pictures, video and voice commands to tell rural women what jobs are available, how much the jobs pay and where they are," Toyama elucidates.

The team hit a glass ceiling when they found that while these women understood how to use the technology, they had trouble understanding why a computerised system was better than traditional word-of-mouth methods.

"We created a video showing a woman complaining to her spouse that she needed another job, and using the computer to find it," and this, Anandan says, clicked. The tough part is implementing the system since most women who do domestic work do not own computers, rues Toyama.

Then there are young researchers like 27-year-old Sumit Mittal who chose to come back to India after completing his MS from Rice University, Texas. He is currently fine-tuning software called Business Finder.

A next-generation, real-time, presence-based mobile technology developed at IBM's IRL, it allows mobile phone consumers to locate and use the nearest service vendors -- whether plumbers, electricians, carpenters or doctors. Mittal and his team have invested a half-year on these dynamic 'yellow pages' for mobile phone users.

"It provides a uniform search capability for both 'mobile' businesses and vendors such as taxis and plumbers, as well as 'static' businesses and vendors including stores and gas stations," he adds.

Even as you read this, an IBM business development team is marketing the service as a 'great value proposition for micro-businesses such as plumbers and mechanics with little or no market reach'.

The technology, claim IRL researchers, could either be hosted by a telecom service provider ('as a differentiating value-added service to its consumer base', observes Mittal) or cut across different operators.

On the surface, Indian innovation has never been stronger. Globally, multinationals spent close to $200 billion on R&D last year, with the meatier glob being spent on computing and communications. Microsoft, for example, spent $6.6 billion last year, IBM and Intel a little over $6 billion each, and Cisco Systems and Hewlett-Packard almost $4 billion each. Most of this went into making incremental improvements and getting new ideas to market fast.

IBM, for instance, has eight laboratories on three continents. At its India Research Laboratory, around 300 scientists concentrate on areas such as automatic speech recognition, relational databases and disk storage.

Till a few years ago, researchers were judged by their patents and papers; they still are, but can now also be found rolling up their shirtsleeves and working alongside the company's consultants to design and deliver services for larger corporate clients. This reflects IBM India's transition into 'services science'.

The services business from India is fast becoming commoditised, just as hardware before it, and IBM realises it must add intellectual property to its India offerings. "Putting researchers on the case is a great way to charge clients a premium," says Ambarish Dasgupta, executive director at research and consulting major PricewaterhouseCooper.

Companies usually invest development dollars to stay ahead with products their customers are using. To some extent this is a matter of protecting the cash cow -- the virtual computer at IBM, printers at Hewlett-Packard, desktop applications at Microsoft.

"If you're a technology innovator, the moment you get ahead of the pack you have to research," says Dr Daniel Dias, director, IBM India Research Lab. "That's to make sure nobody can catch up."

Dasgupta, however, is wary of such R&D spends, typically companies that devote 5-10 per cent of their revenues on 'improving products to keep an advantage over competitors'. "They are doing something to maintain business," he castigates, "And India lends a fertile and cheap bed for this."

That's important to keep in mind if India is to fight for its place as a strategic investment for technology companies. Microsoft India Development Centre started out with only 20 professionals but has now swelled to 1,200 employees.

Koppolu admits, "The first three years, it gave nothing but sleepless nights. Today, the centre has matured into the second-largest outside Redmond and is a platform for taking up key software development works."

This is India's new face.

India inside

Some India-specific solutions

Print what you see (HP Labs)

Want to download broadcast data and print it in real-time? TVPrinCast has run trials for key corporations as well as integrated it into the SatCom-based training programme for gram panchayat members. It can provide reading material, classroom notes, test papers and updates simultaneously during the live broadcast of a television programme.

In other words, the technology augments the TV viewing experience with a printer. HP Labs estimates that it could double up as telemedicine channel, public information dissemination tool, and even provide printable practical guidelines for epidemic prevention in remote areas.

Faster, smaller (Intel)

Intel's Penryn processors are expected to push desktop PCs to run 40 per cent faster than the latest Intel Core 2 chips. Intel's latest processor architecture will allow consumer electronics vendors to make simpler designs for digital home products.

By 2008, it will sell chips as a common foundation that spans products from PCs to consumer electronics, including laptops, televisions, set-top boxes and other networked media players.

All this will be made possible by shrinking its chip features from 60 nanometres to 45 nanometres. Intel India Research Centre's role in producing the next generation silicon wafers and smart chipset processors is undeniably high.

Warana Unwired (MSR)

Microsoft ran an experiment replacing a PC-based system, for helping Warana village, with a mobile phone-based system. The new mobile system replicates almost all PC-based functionality. The project is up and running 24 hours and Microsoft has recorded farmers using data at odd times like 3:30 in the morning. Microsoft now wants to scale up the project to include nearly 54 villages.

Microsoft researchers replaced the client PCs with SMS-enabled phones. A smartphone was attached to the server through a USB port to a PC server, thus creating an SMS gateway. This promises to save nearly $22,000, primarily arising from the maintenance costs of PCs.

Bank on it (IBM IRL)

HDFC Bank receives millions of customer emails. The system requires customer care executives to manually click open each email and respond to it. IBM India Research Lab has developed software that allows the bank to categorise the emails without opening them manually.

The software scans the email for keywords (say "good service", "change of address", "lost ATM card" and so on) and groups the email. It also exploits the context information, so it's easy to find out how many times a customer emails, and whether his problems were addressed or not.

IBM is now planning to market the software to other banks for smart customer information management, targeted marketing, fraud detection and prevention and legal compliance.

Priyanka Joshi in New Delhi
Source: source image