These screenshots reveal how the app automatically analyses the nutritional score of our grocery shopping. | Graphics: Jing Wu / HSG

Just about anyone who buys groceries at a supermarket in Switzerland will have a loyalty card to show at the checkout. Researchers at the University of St. Gallen have now created an app that can offer personalised tips on nutrition by using the data collected by these cards. First it analyses the calorie content and the ingredients of the food purchased. Then it employs an algorithm – developed in collaboration with nutrition experts from their own university and from Inselspital Bern – to make automatic recommendations such as reducing the sugar in your muesli.

“You first have to link it up with your loyalty card, but then there’s nothing more you have to do”, says Jing Wu, a PhD student at St. Gallen. This is important. Other nutrition apps require you to enter the details of every meal you eat, and they’re susceptible to error because people make mistakes and soon tire of the effort involved. But the developer still has one problem to solve: In a household of several people, the app simply assumes that everyone eats roughly the same food.

J. Wu et al.: FoodCoach: Fully Automated Diet Counseling. IEEE Journal of Biomedical and Health Informatics (2025)