This project aims at developing concepts for the automation of construction equipment in agriculture and forestry, as well as building construction. In this context, concepts from mobile robotics shall be applied on construction equipment - cooperating with companies working in these domains. In a concrete application, autonomous control of a backhoe loader from John Deere are dealt with. The objective is assisting a bucket excavator's operator by executing complex subtasks autonomously - without the need of manual intervention.
The Backhoe Loader project has the long-term goal to let an 8 ton backhoe loader (John Deere 410J TMC) perform landscaping tasks fully autonomous. Therefore, the real research machine is equipped with electrohydraulic control valves and angle encoders.
As the real excavator creates excavation- and lifting forces of a few tenthousand newton, it has a high hazard potential for human beings and buildings in the environment. For safe and cheap testing of implemented algorithms, the simulation -and test environment V-Rep is used, for that a realistic model of the backhoe loader was created. The machines behavior and its interaction with dynamic objects in the environment is realistically simulated by the integrated physic engine of V-Rep.
- A Hierarchical Approach for Autonomous Planning and Execution of Excavation Tasks.
- Autonomous trenching with hierarchically organized primitives.
Automation in Construction, Vol. 98, S. 214 - 224. (2019)
- Backhoe Bucket Volume and Trench Monitoring using a Time-of-Flight Camera.
7. Fachtagung Baumaschinentechnik 2018, (2018)
- A Behavior-based Architecture for Excavation Tasks.
34th International Symposium on Automation and Robotics in Construction (ISARC), S. 1005 - 1012. (2017)
- Autonomous Backhoe Loader.
6. Fachtagung Baumaschinentechnik 2015, Vol. 49, S. 351 - 366. (2015)
- Application Scenario for an Automated Backhoe Loader.