Agent-Based Generative Simulation of an Intelligent Distributed Scheduling World with Netlogo

Abstract

Unplanned disruptive events and disturbances such as arrivals of rush orders or machine breakdowns must be managed locally to avoid propagating the effects along the value chain. To overcome the traditional separation between task scheduling and manufacturing execution systems the novel idea of emergent synthesis/control of schedules for better handling the dynamics at the shop-floor is proposed. A new interaction mechanism for simultaneous distributed scheduling and execution control is evaluated using a generative simulation model in Netlogo. The interaction mechanism has been designed around the concept of order and resource agents acting as autonomic managers within the artificial society of a dynamic Gantt world. The advantages of generative modelling in agent-based simulation are discussed to emphasize how difficult to predict emerging behaviours and bottom-up macroscopic dynamics in a manufacturing case study can be addressed by proper design of agent interactions. Results obtained for different abnormal scenarios are presented to highlight the benefits of simulating artificial societies of intelligent agents.