They call it the monkey incident. A couple of months ago, a handful of engineers at Sarnoff Corp.'s lab in Bangalore, India, were conference-calling with colleagues at the research-for-hire outfit's headquarters in Princeton, N.J. They were sitting around a table in a meeting room when they heard loud banging from behind an air conditioner cover on the wall. One of them lifted the cover, and a baby monkey leaped into the room and raced around underfoot.
Two of the engineers were so surprised that they jumped up on the table. Then, "We all fled the room and closed the door," says Kiran Nayak, one of the participants, who recalls the incident with a huge smile.
It all turned out well in the end. In due time, the monkey returned to its mother, out on the building's ledge, and the engineers reclaimed their conference room and resumed talking about data-compression algorithms.
Such are the oddities of global research collaboration.
'Highest expectations'
Sarnoff is one of many Western tech research outfits that have turned to India for its combination of low labor costs, big brains, and English speakers the likes of which are available nowhere else in the world. Notables including Microsoft, Google, and IBM face plenty of challenges, but they're convinced that their investments in Indian research will pay off handsomely in the end. (MSNBC is a Microsoft - NBC joint venture.)
"We have the highest expectations for Indian innovation. There's no question the raw talent exists," says Krishna Bharat, principal scientist at Google, who's starting up the company's new lab in Bangalore.
Sarnoff, a descendant of RCA's original TV-research lab, opened its doors in Bangalore a little more than a year ago and already has 70 employees in two offices. Now it's in the process of consolidating in a larger space to make room for another 80 engineers it plans on hiring within the next 12 months.
Third wave
It's all part of Sarnoff CEO Satyam Cherukuri's master plan for creating a new model for tech research. "We're pioneering global networked R&D on behalf of our customers," says Cherukuri, who came to the U.S. 20 years ago from India for graduate school and has run Sarnoff since 1998.
Cherukuri calls this the "third wave" of tech research. The first wave was in-house R&D in large corporations. The second came with venture capitalists funding innovative startups that eventually grew to maturity or were bought by the big players. "This wave is about harvesting innovations anywhere in the world, with companies using their own employees or third-party researchers like us."
Sarnoff's India operations add to its small army of researchers, who are distributed worldwide. It has 400 engineers and scientists scattered in Princeton, Silicon Valley, Belgium, Japan, and, now, India. The company went through an extensive review of where it should expand next. While it considered 13 countries, it didn't take long to fix on India.
Test bed
Google, Microsoft, and IBM have similar strategies for distributing their research operations around the globe. IBM has long had a research outpost in Delhi, but added a software lab in Bangalore in 2001. Google and Microsoft have opened research labs in Bangalore within the past 18 months.
They can pick up an engineer just out of school for $5,000 to $10,000 a year in salary. But it's not just about the money. It's about the talent, says P. Anandan, managing director of Microsoft Research, India. Also, he says, "India's a test bed for developing technology for emerging economies and rural communities."
Doing research in India isn't without its challenges, however. Tim Mitchell, an Aussie who is Sarnoff's managing director in Bangalore, says it's tough to locate seasoned managers and engineers with the skills in analog-chip design that the company needs.
Personal projects
Google finds recruiting difficult as well. It announced early last year that it hoped to hire 100 researchers before the end of the year, but so far has landed just a couple of dozen. For the first year, the company concentrated on hiring and building a nucleus of senior researchers and managers. "The skill sets we're looking for are hard to come by in senior people," says Bharat, the chief scientist.
As a come-on to the top Indian technologists, the search giant promises them equal status to Google programmers and scientists in the U.S. and at other company outposts. Like all Google researchers and programmers, they're told they can spend 30% of their time on their own projects, in addition to working on assignments from supervisors.
So far, all of the projects the India team is working on are self-contained -- meaning they don't have to do much coordinating with other Google researchers in the Silicon Valley, Tokyo, and Zurich. Eventually, Bharat says, there will be more collaboration and handoffs, taking advantage of the fact that the company has researchers positioned in three time zones to get projects done quickly.
Cricket gear
Sarnoff is further along in the collaboration sphere. Researchers keep in touch via phone, e-mail, and videoconference, and pass tasks off to one another as the day turns to night in one location and engineers arrive at work in another. This is no master-student relationship, however. "A guy in Princeton will ask us to follow up at the end of his day. Or if we're working on something, we ask them to do the same for us," says Prashant Laddha, a lead software engineer in Bangalore.
Sarnoff's Bangalore offices look much like technology offices elsewhere. A bunch of men in their 20s and early 30s labor away in small cubicles. One difference from their American kin: Rather than softball or bowling trophies, these guys keep cricket memorabilia on their desks. Several have motorcycle helmets stowed away, and the parking lot outside is jammed with cycles. The reason: Bangalore's narrow thoroughfares are so crowded with cars and busses that two-wheel transportation is the only way of getting to work in a reasonable amount of time.
The Bangalore researchers are working on a wide variety of projects, but one of them has already met with notable success. It's a good illustration of how networked global research is supposed to work.
Video feed
Since the lab was started, one of the goals was to help create a set of technologies for compressing and transmitting video wirelessly that could be sold, potentially, to a wide variety of customers. Rather than being an assignment from a client, Sarnoff conceived this as technology that it would develop and own itself.
The project is run by Sandip Parikh, Sarnoff's manager for multimedia technologies. He coordinates about 30 people, half in Princeton and half in India, who are working at least part-time on it. The aim was to produce high-quality video transmissions for low-power handheld devices, cell phones, and PDAs, at data speeds as slow as 28 kilobits per second. Sarnoff bases its technology on the latest industry standard for video compression, called MPEG 4, but uses its own proprietary algorithms to produce the best quality video for those demanding conditions.
The breakthrough came during a high-pressure couple of days in April when Parikh and the Bangalore crew teamed up to prove their concept. It was a successful attempt to win over their first customer -- a large European cell-phone operator they wouldn't identify.
Weekend work
This started on the Thursday before Easter. Parikh was in India on a driving vacation with his family when he got a cell-phone call from the customer detailing the video quality levels they wanted to see. While his wife drove, Parikh worked up ideas on his laptop for how to tweak some of the algorithms to squeeze out extra performance.
He called Laddha, and the Bangalore engineers set to work. Over the next three days they took Parikh's ideas, improved on them, and tested the results until they were sure they got it right. By the time the boss returned to the office on the following Monday, the job was done.
To celebrate their victory, Mitchell treated all the employees to a daylong outing at a place on the outskirts of Bangalore called California Resort. They spent the day playing cricket, badminton, and volleyball.
Tech necessity
There could be some long-lasting rewards as well. The Bangalore team has filed five invention-disclosure reports on the work they did on this project -- the first step toward filing U.S. patent applications. "We got some breakthroughs from the Indian engineers," says Parikh.
Over the coming years, that sentence could become a mantra for the many Western tech outfits that are counting on their Indian research operations to give them a competitive advantage. In fact, one day, having a brain trust in India may not be a luxury for the world's tech giants. It may become a necessity.