Developing food products is a time- and resource-intensive process. Information on environmental impacts, nutritional value, food safety and the quality of foods is either lacking, not easily available, or difficult to predict. Developers must solve a multidimensional optimisation problem of high complexity with many parameters to consider. The project ‘OptiSignFood’ (July 2021–June 2024, coordinated by Agroscope) is developing a tool that optimises the nutrition, sensory properties, safety, healthiness and environmental parameters of a new product.
The model approach is based on scientific data and uses machine learning and mathematical methods to solve the multidimensional optimisation problem. Development of the OptiSignFood software requires harmonisation of the nutritional and environmental databases and testing of the prediction models with real data. OptiSignFood is intended to enable faster product development and safe, environmentally friendly and nutritious foods, in addition to highlighting possible trade-offs.