AI Startup Digs Up Business Opportunity in Aging Water Pipes

(Bloomberg) -- When a fifth of the residents of Wakayama in southwestern Japan faced a three-day water stoppage last week to fix a 60-year-old pipe network, they rushed to get ready, only to learn that the repairs could be made without a shutdown.

Some 3,000 complaints were filed to city officials, who said they had no way of knowing until they dug up the pipes.

Cities across the world are facing similar challenges in dealing with deteriorating infrastructure, because of a lack of precision in where and when to fix aging water pipelines. Global climate change, with an increasing frequency of floods, droughts and warmer weather, is overloading water systems. Now, some cash-strapped cities are embracing new technology to make water repairs more efficient, with the goal of cutting construction costs and lowering utility bills.

“No one can live without tap water,” said Takashi Kato, chief executive officer of Fracta Inc., a software startup based in Redwood City, California, that uses software to identify the weakest points in water networks.

In the U.S., public spending on water utilities totaled $113 billion in 2017, according to the Congressional Budget Office. Fixing and expanding water networks across the U.S. alone will cost more than $1 trillion over the next 25 years, according to the American Water Works Association. The decay of water systems in major metropolitan areas has far outpaced their replacement, according to Kato, a former banker who sold a robotics startup to Google in 2013 and later founded Fracta.

The startup’s technology lets cities, utilities and construction companies pinpoint the decay of water pipes before they break ground, helping them save 30% to 40% of their long-term costs, Kato said.

Fracta combines artificial intelligence with machine learning and measures everything from soil quality to population density of an area, along with historical information on when pipes were installed and what they are made of. The startup says it has already won contracts in 23 states, including cities such as San Francisco and Oakland, which have faced water shortages.

“It’s a captivating technology,” said Takashi Matsumura, a consultant at Ernst & Young. “There’s no doubt that everyone is looking for a solution to this problem,” he said.

Similar startups have popped up to tackle the water infrastructure problem, including VODA.ai. Kato says Fracta can win more clients if he’s able to sign up 150 more U.S. customers.

Boosting water-use efficiency has become a global issue, with the United Nations setting ambitious targets for sustainable access to safe water supplies by 2030. That will create business opportunities in Europe, South America and other places across the globe, Kato said. Last year, Fracta started experimental tests in the U.K.

Fracta is also active in Kato’s home country. In 2018, Kurita Water Industries Ltd. bought a majority stake in Fracta for 4 billion yen ($37 million), with plans to acquire 100% of the startup’s shares within the next four years. Kato is remaining on to run the business. “We wanted to create an innovative mindset by acquiring Fracta,” said Michiya Kadota, president of Kurita.

In Japan, tap water reaches almost 98% of the country, marking one of the highest rates in the world. As Japan’s shrinking population weighs on the budgets of local agencies, cities have struggled to rebuild water systems. In Tokyo and other wealthier cities, inefficiency comes from over-investment, according to Kato. Tokyo’s leakage rate is only 3.5%, according to Tokyo Metropolitan Government, one of the lowest in the world.

“Take water pipes constructions in Tokyo for example — it’s like rebuilding houses everyday so that people don’t have to clean them up,” Kato said. “Clearly, that is inefficient.”

To contact the reporters on this story: Shiho Takezawa in Tokyo at stakezawa2@bloomberg.net;Hiromi Horie in Tokyo at hhorie@bloomberg.net

To contact the editors responsible for this story: Kana Nishizawa at knishizawa5@bloomberg.net, Reed Stevenson, Lisa Du

©2020 Bloomberg L.P.