On-the-fly transformation synthesis for information sharing in heterogeneous multi-agent systems

Cooperation in multi-agent teams opens up the possibility of solving large-scale and complex tasks. One of the important aspects in performing cooperation is the ability to share information. In this paper, we propose an on-the-fly synthesis of transformations that enables information sharing in a dynamic group of heterogeneous agents. The synthesis is done by a declarative logic program, which is based on semantic descriptions of representations. Our experiments show that this approach is useful to compensate heterogeneity in dynamic multi-agent teams.