AI in the food chain - a preliminary study

Important results from the project

One goal has been to develop an AI model that can be used by other authorities´ customer service too. There are also other authorities that develop ideas in the area of expert customer service. It has been possible to make certain comparisons between techniques and methods. The long-term goal is a gradual development in collaboration with other stakeholders. There is a great need for fast, efficient and competition-neutral support for companies that want to start or further develop their business. With joint support, the same service is provided regardless of who you are.

Expected long term effects

The result: - Test model that can also be used by other authorities´ customer service - Increased knowledge within the National Food Administration regarding AI - Inspiration for other players in the food industry - Quantitative evaluation of the test model based on the authority´s needs - Market research / RFI. This is a first step in a long-term investment in collaboration with others stakeholders in the food chain. With the help of AI, we can streamline support and also increase quality and equivalence. The project is the basis for the continued work in the AI journey for the food chain.

Approach and implementation

The project has used a systematized amount of data in the form of questions and expert answers. The data set has been used by RISE in the language model "Bert" and an AI model has been developed. In the long run, the project can create a common, automated tool for national, regional and local authorities that support actors in the food chain. The National Food Administration has put the needs in focus and secondarily focused on technical solutions. The project has largely been about learning in one´s own organization and exchange with other authorities.