Many of us take access to clean water for granted. But at least 2 billion people around the world use unsanitary water and 785 million people don’t even have a basic drinking water service, according to the World Health Organization. For developing countries, water sanitation can be an expensive challenge, but one Japanese startup is harnessing the power of artificial intelligence to make it quicker and cheaper.
AnyTech is a Tokyo-based company established in 2015 that has been turning heads for its innovative approach to clean water. It’s using computer vision and deep learning, an approach to AI, to automatically detect contaminants in water. The technology is quicker and cheaper than traditional methods and could be a game-changer for the sanitation industry.
At the Rakuten Technology Conference in November 2019, AnyTech was presented with the Rakuten Technology & Innovation Gold Award for its revolutionary tech and for raising awareness of environmental issues. The accolade came on the heels of AnyTech’s acquisition by major Japanese steelmaker, JFE Engineering. The company says it plans to launch a new business in collaboration with AnyTech.
By 2025, about half of the world’s population is expected to be in water-stressed areas. But water quality has been in the spotlight recently even in developed countries amid reports of lead contamination in drinking water in North America.
AnyTech CEO Yoshiki Shimamoto grew up in Saitama Prefecture, bordering Tokyo. During his youth, dioxin pollution was a major public health concern, affecting water and crops. His initial interest in the problem deepened when he learned about it affecting people overseas. He went on to study physics at university and to begin looking for ways to address to social issues.
Shimamoto began his research into liquid monitoring about 10 years ago, eventually developing a water-monitoring solution called DeepLiquid that is based on a core technology he calls Liquid Texture Mining. The first of its kind in the world, the platform uses cameras to capture images and video of the surface of a pool of water. Computer-vision algorithms then analyze factors such as foam volume, turbidity and anomalies in the water.
“The unique feature of our technology is the use of video analysis specializing in liquid movement, which I majored in, and deep learning AI technology,” says Shimamoto. “We can detect various forms of contamination, including metals such as lead, because each produces a change in the water.”
This contrasts with the traditional method of water-quality monitoring, which consists of chemical sensors and manual sample inspections. AnyTech claims its approach can perform three days’ manual work in one second with practical accuracy.
Plans for expansion to other liquids and industries
DeepLiquid is already being used at some water-treatment plants in Japan. AnyTech plans to expand its business overseas over the mid-to-long-term and is targeting developing countries in various parts of the world. That’s part of its goal of helping society and the environment with cutting-edge technology.
But the know-how can also be applied to other industries. In addition to factories for beverages, cosmetics and biopharmaceuticals, it is also useful for monitoring liquid metals — such as molten iron — as well as food such as melted chocolate and fried chicken.
“We would like to expand our business to various industries because DeepLiquid is a solution that can be applied to many markets,” says Shimamoto.
“I’m very pleased to receive this Rakuten award,” Shimamoto adds, “because there’s a huge demand for environmental solutions but this is not getting enough attention. An award like this can draw public attention to some concrete actions that are being undertaken to solve environmental problems. I’m looking forward to making more social contributions and collaborating with others to make a better future.”
For more about the Rakuten Technology Conference, visit here.