Bob Metcalfe — inventor, entrepreneur, pundit, and venture capitalist — is just back from Europe and jet-lagged.
Padding about the immaculate offices of Polaris Venture Partners, his current base of operation, Metcalfe enters a conference room in a white business shirt and sandals and begins expounding on his latest technology crusade: a ”system on a chip,” which will form the backbone of the coming wireless sensor networking revolution.
”The goal is to network everything,” Metcalfe said, glancing around the room and pointing up to rows of recessed ceiling lights and then to an empty clay vase by the window. ”Every light will be a network node. You could put a sensor in a vase to monitor the water level.”
Metcalfe, 58, the venture backer who took over as chairman of mesh networking start-up Ember Corp. in September, evangelized its technology before audiences in England and Spain this month. In one event, at Cambridge University’s Christ College, he spoke to 200 potential customers beneath stately portraits of Darwin and Milton. Metcalfe told them how wireless sensor networking could solve the nettlesome problem of a flashing microwave oven clock by connecting a clock sensor to an Internet site that would automatically reprogram it.
After a power failure, ”you come home and the microwave oven is flashing 12, 12 . . .” Metcalfe recalled grousing, rapidly opening and closing his fist for emphasis. ”Everyone there could relate to that.”
Metcalfe, a Massachusetts Institute of Technology-educated engineer who built his career on networking, is in his element. He may be best known as the father of ethernet, the local area networking standard he invented at the Xerox Palo Alto Research Center in 1973. In 1979, Metcalfe founded 3Com Corp., a company that sold networking gear based on the ethernet standard. He became publisher of International Data Group’s InfoWorld Publishing Co. in 1990 and wrote a weekly Internet column read by 500,000 technologists.
In his latest incarnation as a Polaris general partner, Metcalfe advises and sits on the boards of Ember and three other Polaris-backed companies.
”I’m still new at venture capital,” he said. ”My ambition is to get in on very early-stage investments with high potential.”
Metcalfe counts Ember in that category. He stumbled upon its core technology in the late 1990s when his friend Michael Hawley, a technology pioneer at MIT’s Media Lab, introduced him to a pair of graduate students, Rob Poor and Andy Wheeler, who were writing their dissertation on wireless sensor networks. Poor and Wheeler had scattered tiny sensors, embedded in plastic rocks, up the side of an extinct volcano in Hawaii to measure its different climates, temperatures, and humidity. (Metcalfe himself had spent time on the Hawaiian islands during his Xerox PARC days and said he based ethernet on Aloha Network, a packet radio network built at the University of Hawaii.)
At Hawley’s urging, Metcalfe became an adviser to Poor. His advice included an exhortation that Poor complete his doctoral work before launching a company. Metcalfe later ponied up his own cash to become a seed investor, along with Polaris, in Ember, where Poor now works as chief technology officer and Wheeler as vice president of engineering.
Ember, of course, faces stiff competition from wireless sensor start-ups on both coasts. And at this juncture, it’s difficult to predict which might gain the upper hand. But the Melcalfe imprimatur is clearly viewed as an asset for the three-year-old Ember, people in the industry say.
”He brings reputation, and a contact network,” said John B. Landry, a serial technology entrepreneur who is now chairman and chief technology officer of Adesso Systems in Waltham. ”Being the inventor of ethernet is no small change. He built a technology that did not run out of gas. Now he sees the same opportunities with Ember.”
Metcalfe, in fact, sees a number of parallels between his work on what he calls Embernet, the company’s still-evolving system on a chip, and his work with ethernet a generation ago. Aside from the networking niche, both technologies hinged on the emergence of a standard — in the current case, ZigBee. But in classic venture capital fashion, he is quick to note one difference.
Calculating that 8 billion microprocessors are shipped each year, of which less than 2 percent are networked, Metcalfe said, ”This market opportunity, in units, is 100 times greater than ethernet.”
