Exploiting Automated Planning to Compute Carbon Footprint Mitigation Plans for the Agricultural Sector

The agricultural sector has experienced significant benefits at all its stages (planning and preparation, planting, crop maintenance, harvesting, post- harvest, and commercialization) thanks to Artificial Intelligence advances. Combined with the Internet of Things, Artificial Intelligence algorithms pro- vide detailed information on crop health, enabling early detection of diseases or pests. Although the applications of Artificial Intelligence to transform agriculture practices are vast, there are no studies on how to exploit these al- gorithms to mitigate the carbon footprint generated by agricultural activities, which contribute to 14% of global greenhouse gas emissions. In this paper, we present AP4CF, an approach that models the problem of mitigating the carbon footprint generated by agricultural activities using a State Transition System. Our model enabled us to apply the Artificial Intelligence technique known as Automated Planning to generate carbon footprint mitigation plans automatically from a series of mitigation activities collected during a thor- ough literature review. AP4CF was evaluated in terms of planning time, plan length, and estimated cost of computed plans. The results demonstrate the flexibility and scalability of AP4CF to new products, extending beyond the agricultural sector.