Behavior-based trajectory adaptation based on bucket-volume estimation

The paper at hand presents ongoing work within the autonomous ex- cavator project THOR (Terraforming Heavy Outdoor Robot). Long-term goal of this project is to develop a fully autonomous bucket excavator which is capa- ble of performing landscaping tasks like trench excavation or material movement autonomously to create a desired surface shape on a construction site. Determi- nation of the amount of excavated material is needed during excavation to adopt executed trajectories based on occurring torques. This knowledge about the sys- tem can then be used to improve trajectories to lower energy consumption and safety aspects like tip-over prevention. Furthermore, the accumulated volume is a measure for the excavated material, the amount of truck fillage and the produc- tivity of the excavation process.