Entwurf einer adaptiven Manipulationsplanung für Massenbewegungen im Steinbruch

Large, chain-driven mobile excavators are often used in quarries /construction pits to load various materials, such as sand, earth, blasted rock and mixed material. The reason for this is that these types of excavators outmatch wheel loaders in terms of breaking and tearing forces. When loading onto trucks or conveyor belts, the stones are often deposited in a certain order depending on their geometry and material properties (e.g. for cement production) in order to achieve a better mixture in the roll grinder or simply a first presorting.

The project aims to develop an autonomous system that solves the application problem described above. For this purpose, camera-based segmentation and classification approaches should be examined for the different types of rock. On the basis of this segmentation, a planning algorithm is to be designed, taking into account the scene stability, which specifies a sequence of the actions to be carried out by the mobile excavator in order to pick up the stones or the bulk material. In order to validate the overall system, the approach is to be tested both in a simulation and in a quarry on a Volvo mobile excavator. Behavior-based control should be used based on the iB2C control architecture developed at the chair. Such types of controllers are particularly suitable for permanently adapting to changing environmental conditions (e.g. slipping of the stones).