AGR-o-RAMA: Precision Agriculture with Autonomous Robots for Active Monitoring
The AGR-o-RAMA project proposes the development of autonomous drone swarms to monitor large agricultural fields. Active monitoring strategies will be developed exploiting online planning and collaboration among the drones. Active monitoring requires that every drone can decide where to move and what to observe, on the basis of an estimation of the relevance of the observed areas. This allows to obtain data at high resolution only in the area where this is actually needed, while less relevant areas can be observed with lower accuracy. Additionally, the possibility to observe the same point of interest from multiple perspectives allows a tridimensional reconstruction. The goal is to provide a map of the field at multiple resolution and in 3D, where the elements of interest can be suitably labelled, for example by differentiating between normal and pathological conditions. The developed techniques will be adapted to a specific case study relevant for the Lazio region: vineyards.