Why AI is going to reduce corporate food loss and waste

As the use of artificial intelligence expands worldwide, companies are increasingly using it to reduce food loss and waste in the supply chain. Japanese firms are applying the technology with the aim of cutting costs, and larger-scale joint public and private initiatives are also being developed.

As hard as we may try to reduce our food waste at home, the bulk of the problem is in the hands of companies. In Japan, 53% of food loss and waste occurs in the supply chain.

The Japanese government has focused on the problem and in 2019 the Ministry of Agriculture, Forestry and Fisheries enforced the Law Concerning the Promotion of Food Loss Reduction for food companies. The following year the Cabinet approved a basic policy based on that law to promote cooperative actions to reduce food loss as a national movement through coordination among various actors, such as the national and local governments, businesses, and consumers. Let’s look at some of the initiatives being taken in the food industry.

AI is a game changer in supply-chain food loss and waste

AI is being applied in two main ways in the Japanese food industry. The first is via machine learning algorithms to predict demand for food products. This is typically used by food retailers like family restaurant chains and fast food stores. The second method is the application of computer vision techniques to sort and grade food products, which is being utilized by food processing companies.

Here are some examples of AI applied in the food industry.

Freshness monitoring and demand prediction

Sushiro, a conveyor-belt sushi chain company, has used AI to monitor the freshness of sushi on its revolving lanes and to collect data on when the sushi is sold and when it is wasted. It has enabled the company to reduce food waste by 75%. Of course, the company’s aim was to reduce costs and increase sales, and that led to a reduction of food waste.

Using big data to predict demand

Chain stores such as noodle restaurant Ringer Hut reduce waste in the supply chain by using an AI system that forecasts the number of orders and controls inventory, shipment volume and staffing in line with that. Its original system uses data such as past sales, weather information, and regional information, such as local events. Furthermore, the system was modified to reflect changes in sales data for the most recent days in order to be prepared for emergencies, such as natural disasters or a pandemic.

The system was developed to cope with emergencies and to improve the company’s supply chain but has resulted in a significant reduction in food waste because the AI system suggests the optimum amount of ingredients required to meet demand.

Product quality examination system

By studying and memorizing criteria such as the size of cracks, dents and chips in tofu, Shikoku Kakoki’s AI line picking system STI-ALPS has automated visual inspections that had required years of experience to perform. The system prevents products from being delivered to retail shops and then wasted due to a defect.

Demand prediction collaboration by public and private sectors

The Tokyo Metropolitan Government is aiming to halve food waste in the city by 2030 and is leading a group of food manufacturers, retailers and other industries to share and compile information to forecast demand and prevent overproduction.

More specifically, it aims to centralize information, such as sales and inventory, provided by retailers, wholesalers and food manufacturers into a company that specializes in demand forecasting. The forecasting company would then provide demand forecasts for food products by considering a range of factors, including weather and events.

The system is needed to reduce the large amount of food waste generated by small retailers, restaurants and hotels.

Two companies were selected to conduct pilot projects in 2020. DATAFLUCT Inc. forecast demand for vegetables, such as green onions. Sinops’ project managed the ordering and discounting at supermarkets of prepared foods, like rice balls and sandwiches, based on real-time inventory information and demand forecasts for the products, which have high gross margins but also high loss ratios. The project reduced waste by 30~40%.

Opportunity to reduce food waste and loss

The motivation to cut costs has been effective in urging companies to reduce food waste and loss. It will be interesting to see how much key, proprietary information companies will be willing to share to reduce costs in public-private initiatives, which could lead to a great leap forward in food loss reduction. As consumers, we should be supportive of companies that are sincerely trying to reduce their food waste and loss and support society in the long term, as well as continue to reduce our own waste at home.

More about food waste and loss in Japan