Predictive Maintenance in Warehouse Automation
For industrial systems or installation such as large warehouse it is of particular interest to forecast if a component or subsystem will fail or cause a problem in the near future. In order to avoid downtime caused by failing components in predictive maintenance one estimates based on control and sensor input as well as models of the system when it is best to schedule a maintain of the part. In this master thesis it should be investigated how predictive maintenance can be used in warehouse automation. The thesis will be carried out in cooperation with SSI-Schäfer (https://www.ssi-schaefer.com/de-at) an expert in warehouse automation. Task of the thesis are literature research, selection of a modeling method, modeling the system, instrumentation of the system (i.e. sensors, controller), and implementing a prototype system. The prototype will be evaluated using a real sub-system such as a conveyors.
Urban Navigation with Semantic Maps
In a previous master’s thesis an autonomous robot was developed which is able to navigate in urban environments like campuses or city centers (see video). The navigation is based on a hierarchical map representation and path planning algorithm. In order to improve the navigation skills and the robustness of the robot in this master’s thesis the environment representation should be enriched with semantic information such as information about street crossings, pedestrian crossings or other important aspects. This information should be taken into account during navigation and the robot should automatically adapt its behavior according to that information, e.g. different motion pattern for pedestrian crossings.