Highlights and preliminary results from a project on effective crop protection using an autonomous tractor fleet.

New technologies are required for safe, site-specific and efficient control of weeds, pathogens and insects in agricultural crops and in forestry. The development and use of autonomous tractors equipped with innovative sensor systems, data processing techniques and actuation tools can be highly beneficial because this technology allows pest control measures to be applied only if, when, and where they are genuinely needed, thus reducing costs, environmental damage and risks to farmers. RHEA (Robotics and associated High-technologies and Equipment for Agriculture) is an EC-funded research project conducted by a consortium composed of 15 research partners from eight European countries. The focus of the project is the design, development and testing of a new generation of automatic and robotic systems for both chemical and physical pest management. A heterogeneous fleet of small, cooperative ground and aerial robots equipped with advanced sensors, enhanced end effectors and improved decision control algorithms will be used. Initially, we are investigating three major scenarios: (a) chemical weed control in winter wheat, (b) thermal weed control (i.e., flaming) in maize and (c) variable applications of pesticides in olive crops. A preliminary system evaluation demonstrated that the intelligent sprayer boom applied the control agent to over 95% of the target area and that the response time, 10 s, of the directinjection system was anticipated in the sprayer system to ensure the accuracy of herbicide spraying. Field trial results showed that the estimated cost for site-specific flame weeding was approximately 24 € ha_1, whereas approximately 52 € ha_1 was needed to perform a conventional broadcast treatment. Thus, the use of VRA (Variable Rate Application) flaming reduces the use of liquid petroleum gas (cost savings of 28 € ha_1). The results also indicated that the control system, mounted on a prototype, air-blast sprayer design, produced a precise system response to variation in the target features, an approximate accuracy of 0.1 m in horizontal resolution and a rapid actuation response of approximately 100 ms. Workshop and field experiments provide convincing evidence that autonomous agricultural vehicles equipped with intelligent implements represent an important step forward for optimizing pest control applications in sustainable row crop, orchard and cereal crop production systems.