The work this year has mainly focused on the installation of the instrumentation into the cargo hall which has been finished. The equipment is mounted on the cargo hall ceiling for the purposes of detecting the bags as well as their depth (distance to the ceiling). In addition, we have begun to search different ML techniques to achieve the detection of the bags with the use of computer vision techniques, like edge detectors and we have started to check the detection of the bags with different types of features (blue red handles (blue/red handles, orientation).
Furthermore, we have developed the algorithms that are needed for the optimization time that is needed to lift and place the bags. In addition, we have developed the algorithms for the optimal trajectory generation of the crane. The algorithms are now going to be tested in the system.
Finally, we also worked with the algorithms development for the optimal cargo hold logistics solution. An initial version of the cargo hold logistics optimizer has been developed, this produces a stowage plan with valid placements for all bags as well as a loading sequence to be used as input by the crane. This is currently based on a construction heuristic, and solution quality will be improved by using more sophisticated optimization methods in future work.
The innovation project shall develop a safe, efficient, and robust crane system for autonomous planning and execution of loading and offloading of ‘big bags’ containing fish food. The overall system consists of three main components: Situational awareness, crane autonomy, and cargo hold logistics. Cranes are widely used in industrial applications for load handling. Although their sizes and capabilities vary according to their specific applications, they share a common objective; to move objects accurately and safely from one point to another minimizing the transfer time. In today's industry fast, safe, and accurate operations is an operator skillset; operators use their judgement and experience to interpret the situation and to compensate for any hazards that might arise. In the current project, we will remove the operator and hence develop a robust and safe situational awareness crane system with a certain level of autonomy such that the crane can sense, interpret and act upon any detected situation while moving a bag from A to B. In addition, operators' experience, and best judgment for how to stack the 'big bags' will be replaced by a cargo hold logistics system which will plan the individual bag's placement in the cargo hold. It will also execute the loading, offloading, and re-arranging, by issuing pick-up and drop-off coordinates to the crane autonomy module. The overall innovation is going to be tested on board to verify its principles.