(CNNMoney) -- Joe Rizzi first heard the underwater songs of humpback whales a decade ago while scuba diving near Hawaii. Enthralled, he decided to pipe their migration music into his beachfront home.
Here's the difference between Joe Rizzi and your average souvenir seeker: He's a rich venture capitalist. Rizzi's quest to capture the whalesong started with a glass pickle jar, a hydrophone and a kayak. It ends with Liquid Robotics, a Sunnyvale, Calif., company with $22 million from investors, 80 employees and customers like BP Oil and the National Oceanic Atmospheric Administration.
Liquid Robotics operates a fleet of wave-propelled, solar-powered ocean robots. Designed to capture Rizzi's elusive music, they have the potential to do much more: predict tsunamis, track fish, snuff out offshore oil leaks and patrol waters for national security threats. They use no fuel, produce no emissions and can travel up to 2,000 miles using wave power alone.
None of that was part of Rizzi's original plan. The Silicon Valley venture capitalist planned to retire at his home in Kona, Hawaii, after two decades of starting and investing in tech companies like Symantec (SYMC, Fortune 500) and SanDisk Corp. (SNDK, Fortune 500) Then he heard the humpback whales migrating to Alaska.
"It's an enigmatic sound," Rizzi says of the songs that became his obsession. "The more you listen to it, the more captivating it becomes."
On a lark, he convinced his neighbor to help him transmit the sounds to his home using a hydrophone (an underwater microphone), a kayak and a cable strung a hundred feet from the shore. But they couldn't hear the whales.
Rizzi needed to move the hydrophone further into the ocean. His solution: a ham radio transmitter inside a pickle jar and a wireless hydrophone hung off a kayak anchored a few hundred yards off the shore.
The police soon showed up. When they found the empty kayak, they thought someone had drowned. No humans died, but the battery did within two hours.
About that time, Rizzi started a nonprofit, the Jupiter Research Foundation, to study humpback whales. He invited several techie buddies to come stay with him in Hawaii, drink beer, play in the water and help with his whale project.
Together, they upgraded his invention with a waterproof case and a motorcycle battery charged via solar panel. Still, he couldn't hear whales -- all he got was a noise that sounded like frying bacon. The group tweaked and tested the electronics for months before locals explained that the noise was tiny snapping shrimp. Rizzi's device needed to be at least a mile from shore, they told him, to hear the whales.
Rizzi tried launching his contraption out to sea, but 50 m.p.h. winds and 10-foot waves battered the craft.
In spring 2005, three years into his quest, Rizzi had lunch in California with Roger Hine, the former director of robotics at semiconductor equipment maker Asyst Technologies. Hine offered Rizzi a solution: How about a floating surfboard powered by a fin 20 feet below the water?
The fin would act like a whale's tail and propel the craft using the motion of waves for energy. A week later, Hine made a miniature model that floated inside a fish tank, which he presented to Rizzi and several others at the Jupiter Research Foundation.
"Here we were, six grown men sitting around a fish tank for five hours watching it crash into the sides," Rizzi recalls. "Roger had an absolutely unique idea and a design that looked like it might work."
A test run of a life-sized version off the shores of Hawaii's Big Island caught the eye of Coast Guard officials, who said they could use one for the agency. As word began to spread, Rizzi got more phone calls -- and soon, he and Hine knew that this was no longer just a personal project.
Rizzi and Hine launched Liquid Robotics in 2007. For the next two years, they refined their prototype into a dependable craft they could actually sell. It needed electronics that would work consistently in salt water. It needed to survive severe weather, 100 m.p.h. winds and 50-feet waves. And it needed the ability to propel itself when there were few waves or none at all.
"It was deceptively complex," Rizzi says.
Today, solar power runs the robots' computers, which chart a course using GPS and collect environmental data with onboard sensors. A cable dangles an underwater glider that looks like window shutters, flapping to push the craft forward.
"They had inadvertently solved one of the great conundrums -- they found a way to harness wave power for propulsion," says Alan Salzman, CEO of Vantage Point Capital Partners, a prominent Silicon Valley venture firm that invested in Liquid Robotics. "When I saw it, I said, 'Oh my god, this is brilliant.'"
The company has since built 120 robots, called wave gliders, that can be deployed to collect data for a wide variety of customers, including oil and gas companies, fisheries, and defense, meteorology and maritime security firms. Traditionally, those industries have relied on ships, buoys or satellite imagery to collect oceanic information.
None are cheap: a ship could cost $150,000 a day and sending out a buoy can run $1 million to $3 million. Liquid Robotics sells its robots for $200,000 each, or rents them for data collection at $1,500 to $3,000 per day.
"You just put it off the dock, tell it where to go and a month later, it's there," says Dr. Jonathan Berger, senior scientist at Scripps Institution of Oceanography at the University of California San Diego. His agency purchased two $200,000 robots for real-time monitoring of tsunami threats across the globe.
The company's biggest challenge, Berger says, will be convincing more people that the technology will be reliable for the long-term.
As for Rizzi? He is not entirely retired, still serving as chairman of Liquid Robotics. But he can finally listen to whales sing from his living room -- and he even broadcasts the sounds online at JupiterFoundation.org. 
