Due to shortening product life cycles, increasing customization of products and the subsequently required changes of production and logistics processes, production and logistics systems have to be more and more flexible, alterable and able to adapt to changing conditions. To execute in-house material flows, more and more autonomous mobile robots (AMR) are used instead of fixed and inflexible systems. They have to operate in a dynamic environment as there are usually persons and obstacles in production and logistics systems as well. In general, AMRs are able to react to such obstacles – e.g. by stopping. Reactions that are adapted to obstacles – e.g. an alert if there is a person or driving around a palette that lies on the floor – and the consideration of information generated locally on fleet level requires sophisticated methods though.

In AdaptiV², methods that enable an adaptive behaviour of single AMRs and a fleet of AMRs are developed. On the one hand, methods to connect and interpret the data generated by different sensors of an AMR are developed. Based on the interpreted data – that allows to e.g. differentiate between different obstacles – an adapted behaviour of an AMR is implemented. On the other hand, information generated locally by single AMRs – e.g. about detected obstacles – is aggregated on fleet level and methods that enable an adaptive control of a fleet of AMRs based on this information – e.g. by taking detected obstacles for routing into account – are developed. Overall, it is intended to increase the performance of AMRs in dynamic production and logistics systems by such adaptive behaviour and to realize a successful interaction of man and machine.