AIFOOD - AI for sustainable FOOD chain from farm to fork
| Reference number | |
| Coordinator | Mälardalens Universitet - Akademin för innovation, design och teknik, Västerås |
| Funding from Vinnova | SEK 6 999 111 |
| Project duration | November 2020 - April 2023 |
| Status | Completed |
| Venture | AI - Leading and innovation |
| Call | AI in the service of climate |
Purpose and goal
The project´s overall goal was to show how AI can be used to control critical processes in a larger facility in urban environments with the aim of reducing its total climate impact. In addition, develop digital tools for holistic decision support as well as develop opportunities for civic involvement by creating a circular sharing economy. Our AI-based models show that considerable energy efficiency can be achieved if the incentives for cost efficiency are followed. For project partner Swegreen, this would mean a reduction in energy costs of approx. 70% units.
Expected results and effects
The project has demonstrated that the management of infrastructure, such as indoor vertical farming, is feasible and that the needs of other stakeholders can be met through domain-specific AI algorithms. Decision support has been developed through a digital twin, which also improves communication among different actors. An app has been created for the development of civic circular sharing economy. We show that tangible and quantitative improvements for the environment and climate can be achieved through minimal investments based on AI algorithms.
Planned approach and implementation
The project was based on experiences from the consortium´s previous project Neighborfood within the same theme. This allowed the project to be adapted to a worsening Covid-19 situation. Then lack of CO2 in the DN scraper, due to empty offices, we decided, with Swegreen at the lead, develop completely new closed cultivation units adapted to AI control. At the same time, we looked at different data sources available in the DN scraper and realized that the current tools do not allow integration, which has led to the development of a digital twin.