An action research study conducted at LUT shows that machine learning can help restaurants reduce food waste and emissions. Decreasing food waste at Kampusravintolat restaurants helps the university cut down on its indirect carbon dioxide emissions.
LUT's researchers have developed a new web application that connects food consumption forecasting with a restaurant’s sales, menu, and food waste monitoring systems. The app forecasts both the size of the lunch crowd and the demand for individual dishes. This enables the kitchen staff to match the production amounts to actual consumption and thereby reduce waste. The new application combines and complements efforts to reduce food waste.
Junior Researcher Jubeen Sharbaf and Research Assistant Sara Mattila from LUT’s Sustainability Change Research Group came up with the solution.
“Overproduction is a typical problem in large buffet restaurants, where it is difficult to anticipate the number of diners. Kampusravintolat already has a history of systematically reducing its waste. The app helped them further reduce their serving waste by one-fifth,” says Sharbaf.
The Curious People newsletter shares our solutions for helping build resilient communities, industry, and businesses while promoting the energy transition and the regenerative use of natural resources.
Emissions expected to keep declining
In the application development, the researchers collaborated closely with Kampusravintolat and its technology partners. That enables the application to use data generated by three software programmes. In the first three months, the serving waste dropped 22 per cent. The project also utilised historical data and machine learning to forecast the numbers of customers more accurately. This, in turn, prevents overproduction.
In buffet restaurants and food service operations, the largest sources of food waste are overproduction, waste from serving, and food left on plates. The next phase of the study will focus on plate waste or uneaten food left on customers’ plates.
“We’re expecting food waste to decrease even more due to continuous improvements to the system. By combining serving waste and plate waste, we're aiming for a comprehensive and scalable model that can be useful to food service actors across Finland,” Sara Mattila assesses.
The study was financed with LUT's own climate funding, which aims to decrease the university's carbon footprint and promote sustainability on campus through practical, research-based solutions. Kampusravintolat has been working steadily to reduce food waste. It has used the Biovaaka food waste tracking system for years, drawing customers’ attention to plate waste. As a result of the study, more leftover food has been donated to charity or sold at a discount at the end of the day.
More information
