Inside Microsoft: Innovation still on the menu

If you spent $9bn a year on research and development and employed 900 of the world's top computer scientists to come up with new ideas, what would you expect in return?

More than a new way of playing video games, a cynic might say.

But Microsoft - a company that may well spend more on R&D than any other business - believes its strategy is paying off, and the proof is the XBox Kinect system.

Pick your avatar

On a visit to the company's headquarters, I had a chance to see some of the projects that Microsoft scientists at its laboratories in Redmond, in Beijing and in Cambridge, England, believe will change the way we see computers.

And the striking thing about what Microsoft's research chief Craig Mundie picked to show off to a group of technology journalists was that almost all of them involved Kinect.

The system which turns a player's body into a games controller was developed with the help of seven different research groups at the company's three main labs, some working on voice recognition, others on motion sensors and a range of other technologies.

Now they are looking at what Kinect could do next.

This photo-realistic avatar is designed to be able to be put into a game or social network

We saw a system which would allow two people to see different images on the same screen, their eyes tracked by the Kinect camera.

Other scientists showed off ways that the camera could capture objects and people in 3D, which might have applications in future telepresence systems.

And there was plenty of work on avatars, for use in either games or in video-conferencing. Two Chinese researchers demonstrated a photo-realistic talking head - type in some text and he'll say anything you want, blinking and moving almost like a real person.

Craig Mundie says the success of Kinect, which racked up 8m sales in its first 60 days, is proof that the sheer scale of Microsoft's R&D strategy is paying off.

"Microsoft is at a point where many of the things that we've been researching for twenty years are starting to add up and produce solutions," he says. "You can't rely on two guys in a garage to make all the changes, some of these things require a huge amount of technology and a lot of scale."

'Holy war in search'

But Microsoft desperately needed a hit from its research labs.

Ever since Bill Gates decided 20 years ago that the company would spend big bucks on trying to see into the future, there have been ideas aplenty but few stand out products. A decade ago, for instance, Gates was showing off tablet computers - but it took Apple and its iPad to make them mainstream.

Peter Lee, who runs the Redmond lab, says the research operation has a wide remit, from dealing with instant fixes to current products to blue-sky thinking.

Professor Lee, who joined Microsoft last year after a distinguished academic career and a spell at the US defence agency DARPA, insists the labs are having an impact on a daily basis.

He cites the contribution to what he describes as the "holy war in search", the battle between Google and Microsoft's Bing.

"Hour by hour we have a large group of researchers actively involved in Bing, constantly adding new research advances into the product."

But he thinks the long-term research is equally important.

"Some call it navel-gazing, we call it pushing back the frontiers of human knowledge," he says with a smile, predicting that his lab will one day win a Nobel prize.

Natural user interfaces

His boss Craig Mundie is Microsoft's big thinker, charting the path of its future research. His current obsession is what he calls natural user interfaces, new ways of interacting with computers, of which Kinect is one example.

There is, he says, a shift about to happen from the old graphical user interface to a trend where "the computer is more like us - it sees, it listens, it speaks, it understands, it even seeks to do things on our behalf."

It is an intriguing vision, but here's a sobering fact. All these clever ideas, smart people, and major investment have not stopped Microsoft from being overtaken in the last year in terms of market value by Apple, which seems to focus on the customer experience now, rather than five years down the line.

Rory Cellan-Jones shakes hands remotely, using a 3D projection

Big, sleepy, and dull, I suggested to Mr Mundie, is how many people now perceive Microsoft.

"We don't feel big and sleepy or dull," he responded, "but if people perceive us that way I think looking at the stuff that Kinect brings should change that view."

And he insists that firms that do not have the patience to spend on long-term research will lose out in the end.

"I don't think any company is going to prevail over a long period of time in giving good business returns, if they aren't making these kind of investments. They'll come and go in a generation if they don't have the staying power that's produced by having real mastery of the underlying technologies."

Microsoft, which still generates huge revenues from its core products, Windows and Office, can well afford to keep spending on its blue-sky thinking. But having made such a big bet on science, it will be hoping that the coming years will produce more Kinects, and fewer tablet PCs.