Interview: Mike Carr of BT

Published on:

What does a company such as BT get from its massive spend on research and development? Access to the world's scientists and the ability to form an opinion, the chief scientist tells Alan Burkitt-Gray

Chief science officer Mike Carr: unless something is available globally the economies of the technology won't work

With this issue, Global Telecoms Business is beginning a series looking at the business of innovation in the industry. Which operators invest in innovation, and how do they manage that process? And how do you measure the return on investment in innovation?
Some operators outsource the process of innovation to independent organisations, or rely on the industry's traditional vendors — and later in the series we'll be talking to those in charge of innovation at some of the key suppliers in the business.
Later in the year GTB will be awarding its second annual Innovation Awards at a ceremony in London — see this link for details.
The first feature in this series looks at BT's innovation strategy. In the May/June issue we'll be talking to executives Orange Labs in Paris, and later issues will feature companies such as AT&T, Deutsche Telekom, NTT, Telefónica and Verizon, as well as selected vendors.
Mike Carr is the executive in charge of research and innovation at BT, as the company's chief science officer. He is responsible of an annual budget of £1.1 billion, most of which is spent at BT's own research labs, Adastral Park at Martlesham in eastern England.

World's innovation

And he's clear about the need for a company such as BT to have its own labs. "You need to tap into the world's innovation and you need a research lab to do it," says Carr, who has worked at the group since 1972, almost all of it in the UK — apart from a couple of years around the turn of the millennium which he spent in its offices in Silicon Valley, California.
One company cannot hope to innovate everything, he indicates, because "98% of innovation is somewhere else". That governs his policy on how to run Martlesham: "You have to make labs open, so that they can work with start-ups, scientific institutions and production companies and take their innovations. Our ability to get an edge comes from an ability to recognise what's good."
One of the indicators of effectiveness in innovation is the number of patents, an independent arbitrator of novelty. BT manages 120-130 a year, says Carr. "It's one good way of measuring that we're at the forefront of technology."
In the last five years Carr has seen a tripling in the research and development budget: broadband and BT's 21st Century Network project, as well as the company's growing ambitions to be a corporate IT provider are clearly behind the growth.
"We don't split the budget officially," says Carr, but he's happy to indicate that about a third of that £1.1 billion is spent on 21CN, a third on broadband and a third on ICT — information and computer technology.

Global technology

But the idea is not, he emphasises, to develop solutions that are unique to BT. "Having a technology solely for us won't work," says Carr. "Unless something is available globally the economies of the technology won't work."
So why does BT need to develop, and patent, technology? It's a three-stage process, suggests Carr: "We invent it, get it used, and use it best."
A clue to what he means can be seen elsewhere in Adastral, where Russell Davey is leading a team working on the practicalities of fibre to the home.
Current GPON — gigabit passive optical network — technology uses a single wavelength of light to deliver a maximum of about 2.5 gigabits to maybe 32 homes, and about 1.2 gigabits back into the network. The fibre will carry one wavelength down and one wavelength upstream.
Sounds good today, but an inevitable law of telecoms is that people will need more and more bandwidth. BT and others can predict with confidence that in a few years it will be possible to deliver 10 gigabits or more down a fibre on a second set of wavelengths — and, inevitably, the demand will be there.
But the last thing you want to do — a few years after undertaking such a huge capital investment programme as fibre to the home will be — is reequip everything to expand the bandwidth.
"You could put multiple wavelengths on the fibre," says Davey. But his team recognised that the GPON equipment that manufacturers were proposing today were suitable only for one wavelength in each direction. More precisely, the receivers would pick up all wavelengths, so any attempt to add new ones would just cause mutual interference.

Wavelength filter

"So we thought ahead, and realised we could specify a filter in the current design which would just pass the wavelength used today," says Davey.
That will satisfy the demand for most users over the next few years. As soon as the technology and the demand is there, new customers can be equipped with filters for the wavelength used for 10 gigabits. "We could just give these units to new customers. But it was our ability to think ahead that made us plan how to upgrade. It's much cheaper to replace a few boxes than to replace the fibre."
As a result of the BT team's work, says Davey, the filter has been specified as an essential part of the design.
That's a good example of the detailed innovation that an operator can be responsible for — building in scope for future expansion in capability and avoiding unnecessary extra costs. Longer term, BT has a team of futurologists.
Lesley Gavin is a former architect who has become an enthusiast for virtual worlds, such as Second Life, and is learning how they can help BT understand issues such as remote working.
In such cases BT can have a better understanding of the possibilities that potential customers, she says. "You can push the client into having something they never knew they could have, that will last well into the future. Most people don't think in that way, so just asking the customer what they want doesn't get you there."

Personalised service

Before joining BT she had a company making systems for equity trading: "We wanted to carry the ambient information from a trading floor into electronic trading. It's an iterative process, and a very personalised service. We rented space on a trading floor and had 'crits' [an architect's term for critical discussions] every Tuesday."
Her BT colleague Jonathan Mitchener specialises in devices and their evolution. "We are increasingly a software service company, so we need to know what's coming in the devices space," he says. "We want customers to get the optimum experience."
It means keeping close to the vendors, such as Sony with the PSP and Microsoft with its Xbox, and "looking beyond what we expect to happen", says Mitchener. "It's what device evolution is all about."
Apart from vendors, BT keeps in close contact with 34 universities around the world, says Carr. "MIT is a core collaborator of ours. It helps us see where the world is going — not deep research in science but in its adoption."
Among the other universities that BT is close to are Cambridge — a short drive from the research labs — and, a lot further away, Berkeley and Stanford in California. And both University College London and the University of Essex have researchers on the Adastral site, he adds.

Academic sharing

"We want to open up academic opportunities. The good thing about academics is that they share everything with people. Mainly we pay them for their work rather than pay for a chair [sponsoring a professorship] and hope something happens. It means we can give it some direction."
Another change is that BT — like some other operators — is starting to set up labs in other parts of the world, including one with 20 people in Shanghai and another with 50 in Kuala Lumpur. There's also a "lablet" in India, not directly managed by BT.
Why in those places? "We need to get our early stage research close to our customers," says Carr. "Part of our strategy of open innovation is identifying problems and then finding techniques to solve them, so we need to be close to the end customers."
Each region is "massively different", he notes. It's not a challenge of "just having the technology", he says, but of building "the business case around it".
It's all a long-term process, Carr admits. "Concept to production is about 10 years. We stop loads of things before they're finished," he says. "Half of the things we work on get adopted," but even when they don't the work can be useful.
"Having the best technology in the industry doesn't mean you're successful," he warns. "We could just specify a switch that was for us and was built for us in the UK."
Indeed, that's how it used to happen. One of Martlesham's first projects was System X, the digital switch launched by the ancestor of BT and its selected vendors in 1979. The world of telecoms has changed: operators no longer require their own in-house technology.
"But having a research lab means we've got an opinion," says Carr. "It's a complete change." GTB