A Modular Sensor Fusion Approach for Agricultural Machines

Summary Today’s architectures for sensor data handling in agricultural equipment are static and single vehicle-centric. This means that they assume a static system configuration over time while ignoring data contributed by other sources like implements and other machines that they currently interact with. Therefore, it can be stated that the inability to adapt to machine reconfiguration can be identified as a significant bottleneck. The approach presented in this paper is designed to overcome this shortcoming. For this purpose, a data handling system was designed that combines solutions for management of distributed knowledge and data fusion algorithms especially designed for the demanding application field at hand. Besides supporting the user, this modular approach enables engineers to design future systems in a holistic manner with reduced complexity. Furthermore it can help to reduce hardware costs due to elimination of redundant sensing systems and pushes the limit of applications to open up new opportunities in the future.