Emergent short-term forecasting through ant colony engineering in coordination and control systems

This manuscript presents a design for the emergent generation of short-term forecasts in multi-agent coordination and control systems. Food foraging behavior in ant colonies constitutes the main inspiration for the design. A key advantage is the limited exposure of the software agents in the coordination and control system. Each agent corresponds to a counterpart in the underlying system and can be developed and maintained exclusively based on knowledge about its counterpart. This approach to make non-local information available without exposing the software agents beyond their local scope is the research contribution and focus of the discussion in this paper. The research team applies this design to multi-agent manufacturing control systems and to supply network coordination systems, but its intrinsic applicability is broader.