Founded by a GPON inventor, but powered by cool precision robotics, FiberZone Networks is hoping to carve a business out of automating patch panels -- a topic that's been long discussed without any companies coming up with ideal answers.

FiberZone announced this week it's raised $8 million from investors including Portview Communications Partners LP , Stage One Ventures, and lead investor Novak Biddle Venture Partners . The round follows a $3 million Series A. (See Fiberzone Gets $8M.)

FiberZone was born, indirectly, from the passive optical networking (PON) space. Founder and CTO Yossi Arol was also a founder of FlexLight Networks Inc. , developing GPON in its infancy. But around 2003, he turned his back on the idea. "The company wanted to go in the direction of fiber-to-the-business, and I thought fiber-to-the-home was the right way to go," he says. (He notes that he's since been proven right.)

But while at FlexLight, and in earlier days with ECI Telecom Ltd. , he'd picked up on the problems of the manual patch panel -- an array of holes, essentially, into which fiber-optic cables are plugged. Telecom operators use the patch panel to connect one piece of equipment to another, but the manual process is prone to human screw-ups, given that one bank of panels could connect to hundreds or even thousands of ports.

The industry has long wished for an automated kind of patch panel, and Arol decided that's what his next startup would tackle. FiberZone was created in 2003 and has 17 employees, with plans to bring that figure to 25 or 30, says Sandy Roskes, FiberZone's vice president of marketing.

The company's first product is the AFM-400, a 200x200 patch panel. Using small robotic arms on a pulley system, the system -- a closed box resembling any other type of 23-inch-wide telecom equipment -- strings a fiber from one port to another, like a 2001: A Space Odyssey version of Lily Tomlin's Ernestine. It's built to do its task quickly without going insane and killing the crew.

"The big fiber companies have been looking for something like this for 10-plus years," Roskes says.

The trick at this game is insertion loss, which FiberZone says it can keep to less than 0.4 dB. "If it's above about 1 dB, you're going to have to reconfigure the system," Roskes says. Another key feature is a "passive" latching system that holds the fibers in place, meaning the connection stays up during a power outage.

Fiberzone doesn't appear to have direct competitors, but plenty of other companies have taken a shot at improving patch panels. One potential method is to bypass cables and use an all-optical switch to connect ports. Yes, that would be the optical-mirror technology from Calient Technologies Inc. or Glimmerglass .

"After the bubble was bursting, you heard the big optical crossconnect MEMS guys saying their switches could be used for that application, but the price is super high for that" -- something like $5,000 per port for Calient, compared with "dollars per port" for an old-school ADC (Nasdaq: ADCT) patch panel, says Scott Clavenna, chief analyst at Heavy Reading.

He adds that Siemens AG (NYSE: SI; Frankfurt: SIE) has a smarter-than-average patch panel that uses LEDs to indicate which ports to connect, reducing human error. But it still requires human hands to plug in the cables.

Considering it doesn't do anything else yet, FiberZone's fate depends on whether or not this patch-panel business can sustain a company. "If an operator puts one in every central office to manage their fibers, then you're talking about a big market: 20,000 COs in the U.S.," Clavenna says. That's a big market, but FiberZone is selling into a function that's not necessarily a high priority for carriers and that won't tolerate exorbitant margins. FiberZone has gotten a taste of the market with alpha testing, and the product will go into beta testing late this spring.

— Craig Matsumoto, Senior Editor, Light Reading