An Agent-Based Approach to Self-Organized Production

TL;DR

This paper models a material handling system using agent-based simulation to analyze how local interactions and cooperative behavior among units can reduce blockages and improve production efficiency.

Contribution

It introduces an agent-based model with local and indirect interactions to simulate and analyze self-organized production processes.

Findings

Cooperative behavior among units reduces system blockages.

Simulation results align with real plant processes.

Local interactions influence overall system efficiency.

Abstract

The chapter describes the modeling of a material handling system with the production of individual units in a scheduled order. The units represent the agents in the model and are transported in the system which is abstracted as a directed graph. Since the hindrances of units on their path to the destination can lead to inefficiencies in the production, the blockages of units are to be reduced. Therefore, the units operate in the system by means of local interactions in the conveying elements and indirect interactions based on a measure of possible hindrances. If most of the units behave cooperatively ("socially"), the blockings in the system are reduced. A simulation based on the model shows the collective behavior of the units in the system. The transport processes in the simulation can be compared with the processes in a real plant, which gives conclusions about the consequencies for the production based on the superordinate planning.