Kilby Labs: A birthplace for inventions
An engineer tinkers in a tool-strewn laboratory, coming and going at odd hours. Prototypes are created, tested and revised. Some are discarded – the price of progress.
It was in this kind of work environment that Texas Instruments (TI) engineer Jack Kilby invented the integrated circuit in the summer of 1958, the invention that paved the way for computerization and eventually won Kilby the Nobel Prize in Physics. Too new to have summer vacation, Kilby worked alone in a lab, toiling on borrowed equipment while his colleagues took time off. He worked hard to accomplish his task quickly, because when summer ended he knew he would be pulled back into other projects.
Today, more than half a century later, a new research hub captures the spirit of Kilby's breakthrough. It's a home for dreams to be made into reality, adding a dose of deadline pressure and market focus.
The Dallas-based 7,000-square-foot Kilby Labs, which opened in 2008, hosts 23 researchers-in-residence who are pursuing high-risk, high-reward ideas. Their projects were handpicked, with the goal of creating relevant and profitable products for the next decade.
"We ask ourselves, 'What now?'" said Ajith Amerasekera, director of Kilby Labs. "We look around at opportunities and find a way to make the world better."
Not a thinktank, a "do-tank"
Kilby Labs resemble a tinkerer's garage, with no cubicles, only open workbenches. Engineers from different business units and disciplines rub shoulders.
But the research is in no way haphazard. An advisory board with representatives across all of TI's businesses selected six inaugural projects from 60 proposals.
The goal is to produce not theoretical advances but prototypes.
"Jack Kilby did two things very well," Amerasekera said. "When he had an idea, he went and followed it. Secondly, he built stuff. The integrated circuit was an idea that he built and it worked. Same with the calculator – he put together a team and built it."
Projects at Kilby Labs last no longer than a year. They are reviewed every three months to ensure that they are meeting milestones – design, build, test – and can be terminated if they not progressing.
"This is not an open-ended program," Amerasekera said. "There is a sense of urgency here and a queue of ideas is already in the wings."
Kilby Labs researchers cannot disclose what they are working on. But projects are influenced by TI's collaboration with 40 universities and semiconductor industry leaders. This group looks at where the industry needs to be as far out as 2025.
One of their visions is a future enabled by sensors that positively affect health, safety, energy and the environment.
"We see applications that sense movement, for example, and for that we'll need sensors and controls, microelectricomechanical systems, digital components and circuits," Amerasekera said.
"It will likely be about sensors on the body that create a 'body area network' and transmit data to systems that do the data crunching, tracked remotely," he continued. "For the elderly, it might be a central system that takes action on a seizure or fall."
TI's customers are already looking into these areas, but engineers will first need to develop ultra-low-power devices and analog circuits that are safe for the human body.
Most folks define failure as something that doesn't work – a cake that falls flat, a go-kart engine that putters out. Not so at Kilby Labs, where the only uncertainty is failure.
If a project produces a working prototype, it is a success. If it produces a prototype that does not work, but the project has demonstrated that the underlying concept is not feasible and the reasons why are understood, that is also a success. Only if a project produces no clear answers is it considered a failure.
"To conclude, 'This calls for further study' – that is not a project success," Amerasekera said. "Unless it is a dead end, we see three possible other paths coming out of it."
TI expects that only 10 percent of the projects will go to market, with half terminated midstream because they aren't meeting their targets.
"The intent here is to take on high-risk projects," Amerasekera explained, especially technologies that will improve customers' health, safety and security in the future. "If less than 50 percent are terminated, we won't be doing our job."