Reactive Mission Planning for UAV based crane rail inspection in an automated Container Terminal

摘要

At the automated Container Terminal Altenwerder (CTA) Hamburg, Rail-mounted gantry cranes (RMGC) transfer the containers between the incoming ships and the customer container trucks. Due to the special geomorphological conditions of the soil at CTA and heavy load of the crane movement, the rail tracks move in horizontal and vertical plane. Currently, these rail tracks are bi-annually inspected using traditional surveying methods during which the container storage blocks need to be stopped to allow the human surveyors to operate, leading to significant loss of operational capacity. In the AeroInspekt project we aim to inspect the rails using UAV based autonomous aerial photogrammetry without affecting the crane operations. To avoid the bad effect on the photogrammetry process from the images with crane occluding the rails, several intelligent drone mission planning strategies were designed and tested in a simulation environment in order to minimize the flying time. The optimized approach was implemented in the control station app on an Android tablet and tested in practice.