AI-based control architecture for interconnected, flexible material flow systems

Manufacturing companies need to control and organize the material flow in their production systems as efficient as possible to face increasing competition. Beyond that, quantities of product series decrease because of the increasing demand for individualized products. This results in less standardized processes in the production area and indicates the need for more flexible solutions to realize the material flow. In the course of this project, this problem is exemplarily investigated for a material flow system for the supply of a production with material. In this system, material flow and supply is realized both with automated guided vehicles (AGV) and with robots which transport material on a grid which is mounted on the ceiling of the building.

As production areas change dynamically (humans moving around, boxes standing around, ...), the navigation of AGVs has to be able to cover such situations. Therefore, approaches utilizing Artificial Intelligence (AI) are applied to realize navigation based on data generated by laser scanners and cameras of the AGVs. The work at IFL is focused on the routing of the robots on the grid. Using AI-based approaches, a fast route from a start to an end point on the grid has to be determined. Beside the pure distance, aspects like other robots moving on the grid and non-accessible areas on the grid have to be taken into account. Combining these different systems to supply the production with material, an interconnected and flexible material flow system can be realized.