Japan shows the highest number of food waste per capita in Asia: more tan 6 million tonnes of food waste a years costs Japan about 2 trillion yen ($19 billion), government data shows. That why the japanese government enacted a new law to halve such costs from 2000 levels by 2030, pushing companies to find solutions.

One solution introduced the convenience store chain Lawson Inc: the company started using AI from U.S. firm DataRobot, which estimates how much product on shelves may go unsold or fall short of demand. Lawson aims to bring down overstock by 30 percent in places where it has been rolled out, and wants to halve food waste at all of its stores in 2030 compared with 2018. Disposal of food waste is the biggest cost for Lawson’s franchise owners after labour costs.

Drinks maker Suntory Beverage & Food Ltd is experimenting with another AI product from Fujitsu Ltd to try to determine if goods such as bottles of oolong tea and mineral water have been damaged in shipping.

Until now, that’s been a time-consuming human endeavour. With the new AI, Suntory hopes to gauge when a damaged box is just that, or when the contents themselves have been damaged and need to be returned.

Suntory aims to reduce the return of goods by 30-50 percent and cut the cost of food waste and develop a common standard system that can be shared by other food makers and shipping firms.