Microsoft hones its research-to-commercial-product pipeline

In any Microsoft Research (MSR) discussion, Microsoft officials emphasize that MSR does both basic and applied research with no pressure applied to researchers to productize the results. Accordingly, there is no guarantee that any given MSR technology will be commercialized at some specific date, in some specific way.

That caveat may soon become unnecessary, given that so many MSR projects have morphed -- and are continuing to morph -- into key components of established Microsoft products, and/or standalone products in their own right.

As Microsoft officials are touting on September 26 – which is the 15th anniversary of the founding of MSR – MSR has become a well-established pipeline for getting new technologies into the market. At the two-day MSR celebration in Redmond for press and analysts (which, alas, I’m covering from afar), Microsoft officials are planning to highlight some of the most successful technology transfers that have occurred between the 700 MSR researchers worldwide and the various Microsoft product groups.

To demonstrate this trend, the MSR team is planning to emphasize heavily researchers’ contributions to Windows Vista and Office 2007. When asked to list the most important MSR contributions to Vista, Senior Vice President of Research Rick Rashid provided this laundry list (in no particular order):

• Static driver verifier, which automates device-driver analysis to check for potential code problems;

• Optimization technologies for speeding up processes in Vista systems;

• Search technique/algorithm elements that will be embedded in Vista’s embedded desktop search engine;

• Photo-handling improvements that are part of the new Windows Media Photo digital-photo format;

• New user experience elements that will become part of the final Vista user interface (thanks to the work of researcher Lili Cheng and her team, who are in the midst of filtering back from the Windows team to research, Rashid said);

• DirectX graphics;

• And last (but definitely not least), security functionality, such as cryptographic enhancements

“I think we do a better job than other companies, in terms of moving our research into products,” Rashid said. However, “we don’t bias the front end. When good results come out, we take advantage of them,” he added.

Microsoft is planning to show off at its research-fest a number of projects, including Windows Live Local mapping, surface computing advances and “virtual worlds” developments. But it also is planning to show off some less-sexy but potentially more far-reaching distributed computing technologies. I’ll have more on that in my next post.