A row of high definition video screens on the wall opposite a narrow table suddenly snap to life. There, across the table and at life size, is Dr Prith Banerjee, newly appointed director of HP Labs, HP's enormous research and development unit. Karlin Lillingtonreports.
Appropriately for an R&D man, he has arrived for an interview in virtual form, using one of HP's more recent R&D triumphs - the ultra high end video-conferencing space co-developed with Steven Spielberg's Dreamworks Studio, known as the Halo Room (nothing to do with the computer game, meaning no alien invaders open fire, thankfully).
He's thousands of miles away in Palo Alto at HP's headquarters, but for all intents and purposes, he is here in Dublin. So pristine is the sound quality of the Halo Room, that when someone knocks gently on the door in Dublin, Dr Banerjee turns to check the door in the duplicate Halo Room on his end.
He launches immediately into a compact biography, not because he relishes his CV, but because he feels his background - unusual for the head of a research organisation - directly relates to his planned overhaul of HP Labs.
In short, he is an academic administrator and professor whose research experience has an entrepreneurial bent. After coming to the US from India and taking his PhD in electrical engineering at the University of Illinois, he worked his way up to his most recent position as dean of the College of Engineering at the University of Chicago.
Along the way, he founded two electronic design automation companies: Accelchip, which he sold to Xilinx in 2006, and Binachip, from which he "hopes to see an exit soon", he says, meaning of course a profitable buyout, not his physical departure.
Of his early research experience, he says: "What frustrated me was I was the creator of all this technology but when I tried to transfer it to companies, they said, 'Thanks very much, but we can't use it'."
He feels most research ends up like this - unappreciated and unused because it is developed in isolation from the places that might commercialise it. The gap between R&D and product is too large. So what he wants to do is duplicate the process he went through in his companies, applied to a large research lab setting.
"Large" is the operative word when talking about HP, a company whose corporate reputation for many decades has been closely tied to that research environment. HP currently employs more than 600 researchers at seven global lab locations. About $3.6 billion (€2.5 billion) of HP's $100 billion revenue goes straight to R&D.
Research currently is done on several timeframes. If it is in line to become a product or service, the timeframe can be as short as six to 18 months.
But HP Labs's remit is to look up to a decade ahead, says Banerjee, and consider areas where HP may have no commercial interests at present, or imagine ways in which existing markets might develop. "My job is to figure out what we should be doing going forward," he says, though he won't yet discuss areas he wishes to explore.
His immediate intent is to transform how research is done. Currently some 150 separate projects are under way, with two to three researchers per project. "I'd like to take a smaller number - say, 30 projects - each with about 15 people working on them; high impact projects with large collaborative groups. I'd like to have a large number of those projects transfer to business units."
It isn't enough to come up with good ideas if they never go anywhere: "HP has some really cool technologies but they are not making it into the hands of customers." Collaboration - between HP teams from other parts of the company, or with university or corporate partners - is the way to get technologies out into the real world, he says, and is a key concept for him.
For each project, if a potentially promising technology results in an initial 18-month period, a third of the group - more commercially-oriented researchers - will carry forward the project, joined by engineers from the business unit. Eventually they will be joined by six people from sales and marketing.
With researchers a notoriously free-thinking bunch, won't it be hard to impose this kind of order? "I've been asking researchers, what is it that excites them? Most researchers want to see what they work on go into products," Banerjee says. Those researchers who don't fancy the idea will return to start on new projects, as only a commercially-bent third will carry any project forward anyway, he notes.
Banerjee's goal is to replace the "random event" nature of current research right now with something more closely aligned to the creativity and excitement of start-up companies. In essence, promising projects will become mini-start-ups within HP Labs.
It's a major shift for the research culture at HP, but Banerjee is optimistic. "I've invited all the researchers to discussion sessions, and they are very excited about the process. I have told them, I want you to research and to write your own future."
