Svinnkollen (The Food Waste Experiment)
Can more – and better – information result in less food being wasted? Researchers will be investigating this together with pupils and teachers across the whole of Sweden in the Food Waste Experiment (Svinnkollen). To assist them, they will be using an artificial intelligence (AI) app and the world’s largest food sustainability database.
Our food is responsible for over one-third of human-caused climate emissions. At the same time, almost one-third of food produced globally is thrown away. In the Food Waste Experiment, researchers will be developing and testing a new way to reduce food waste in Swedish schools: by providing more information and individual feedback. During the three weeks of the Food Waste Experiment, teachers and pupils will both help develop, and use, an app to find out how much food they are throwing away. The first week, teachers and pupils will train the AI to identify many different kinds of food. Here they will photograph their lunch plates and tell the AI what kinds of food are on the plates. The two following weeks, they will use the app to photograph their plates before and after eating, which will allow the app to calculate how much food is being wasted. Before lunch, they receive information about the day’s menu being served in the school dining hall, the nutritional content of the various dishes and their climate footprint. The information comes from the world’s largest database on factors relating to food and environmental impact, health and justice, developed by the company Consupedia in collaboration with researchers at Dalarna University and KTH Royal Institute of Technology.
Through the project, researchers want to test whether it is possible to create virtuous circles in how food is dealt with – from wholesalers through to pupils. Making pupils more aware of how their food is affecting the climate, health and the environment, can lead to more thoughtful choices in the dining hall. As school kitchens get better information about what food the pupils like and dislike, menus can be adapted to better suit the needs of the pupils. The kitchens can then make more tailored orders from the wholesalers so that the amount of waste is reduced both in school kitchens and at wholesalers.
To investigate whether it is feasible to calculate food waste and its climate impact with AI in a smartphone app. To explore whether more – and better – individual information and feedback can result in more climate-friendly food choices and less food waste.
How to participate
Students and teachers participate by:
1) training the AI to identify different kinds of food, by scanning their lunch plates before they eat, and informing the AI what is on the plate
2) scanning their lunch plates before and after they eat, to calculate amount of food waste and its climate impact
Mobile phone with Svinnkollen/Consupedia app.