Silicon Valley wireless group seeks new builder

The big-name backers of Silicon Valley's proposed wireless network are looking for a new company to build and operate the system after potential financiers rebuffed the original vendor.

Azulstar, a small municipal wireless company based in Michigan, now plans to work on specialized applications on the multifaceted network. The Silicon Valley Metro Connect consortium is seeking another contractor to handle the infrastructure as a whole, said Seth Fearey, vice president and chief operating officer of Joint Venture: Silicon Valley Network, a civic group that spearheaded the project.

It's only the latest setback for an ambitious plan that at one time was expected to start rolling out early this year. Fully deployed, the network would cover 1,500 square miles and offer multiple networks to 40 municipalities as well as businesses and consumers. Cisco Systems plans to donate equipment and IBM plans to serve as system integrator. But Azulstar, their partner in Silicon Valley Metro Connect, wasn't able to raise enough money to build the estimated US$100 million network.

Azulstar approached venture capital firms, traditional investment banks and corporations based in the technology-rich region, but found them skittish amid the upheaval in the municipal wireless industry this year, said Tyler van Houwelingen, the company's founder and CEO. EarthLink's drastic cutbacks in its high-profile muni Wi-Fi business, which sank neighboring San Francisco's citywide coverage plans, rocked the industry, he said.

"Everybody in this business is going through the same thing," Houwelingen said.

Azulstar's size may also have been a factor in the fundraising effort, Fearey believes. The company has built a handful of networks around the U.S. but hasn't been one of the big players in the industry.

The search for a new main contractor has halted, for now, the key steps of building test networks and crafting a model contract to take to the many governments in the area, Fearey said. That contract could be fine-tuned to meet the needs of individual cities. His group and Metro Connect have worked out almost all the details of the model agreement, but their negotiations are now on hold, he said. And the new contractor will have to be on board before the two test networks can be built, he added. In September, an executive of Metro Connect forecast that the contract and testbeds would be finished by year's end.

The concept of citywide wireless networks that support themselves through advertising and residential subscriptions has shattered over the past year, replaced by business models that require governments to buy services for their own applications. The backers of the Silicon Valley network have planned from the beginning to attack the problem with variety, potentially including Wi-Fi, WiMax, public safety, in-car, and other networks in several radio bands. Building them all at once, using the same set of mounting locations such as light poles, will lower the overall cost, he said. Cities and agencies in the region are interested in buying services including meter reading and police communications, he said.

Some analysts think the project's organizers have bit off more than they can chew. They might be in a better position today had they started building the network up from an existing system in one city, or at least gotten a few key municipalities to commit to buying services, said Esme Vos, founder of the industry-watching Web site Muniwireless.com. With firm dollar commitments in hand, it would have been easier for them to get investors interested, she said. The way it was done, it's hard to imagine what partner would step up to build and run the network, Vos said.

A new builder-operator wouldn't have to raise $100 million all at once, Joint Venture Silicon Valley's Fearey pointed out. All it would take to get a commercial network off the ground would be $10 million to $20 million, he estimated.

Yet a commercial deployment would only come after tests on the trial network, which would probably last 60 or 90 days. Full commercial network deployment might take 18 months, Fearey said. And even after the model agreement is finished, each municipality will have to approve it, possibly with its own modifications.

The network is intended to be profitable, Fearey said. As a hub of technology pioneers, Silicon Valley is sensitive to the need for a solid business model that lasts, he added. Many local residents enjoyed the Metricom wireless data network in the 1990s until that company went bankrupt and its outdoor radios were left gathering dust, Fearey said.

"They really don't want that to happen again," Fearey said.

Metro Connect could not immediately be reached for comment.