A Multi Agent Architecture to Support Self-organizing Material Handling

Emerging market conditions press current shop floors hard. Mass customization implies that manufacturing system have to be extremely dynamic when handling variety and batch size. Hence, the ability to quickly reconfigure the system is paramount. This involves both the stations that carry out the production processes and the transport system. Traditionally system reconfiguration issues have been approached from a optimization point of view. This means allocating a certain batch of work to specific machines/stations in an optimal schedule. Although in a an abstract way these solutions are elegant and sound sometimes the number and nature of their base assumptions are unrealistic. Approaching the problem from a self-organizing perspective offers the advantage of attaining a fair solution in a concrete environment and as a reaction of the current operational conditions. Even if optimality cannot be ensured the solutions attained and the online re-adjustments render the system generally robust. This works extends the IDEAS Agent Development Environment (IADE) developed in the FP7 Instantly Deployable Evolvable Assembly Systems (IDEAS) project which has demonstrated the basic concepts of the proposed approach. The main architectural changes are presented and justified and the prospects for the analysis and self-organizing control are presented.