Security for a Multi-Agent Cyber-Physical Conveyor System using Machine Learning

Abstract

One main foundation of Industry 4.0 is the connectivity of devices and systems using Internet of Things (IoT) technologies, where Cyber-physical systems (CPS) act as the backbone infrastructure based on distributed and decentralized structures. This approach provides significant benefits, namely improved performance, responsiveness and reconfigurability, but also brings some problems in terms of security, as the devices and systems become vulnerable to cyberattacks. This paper describes the implementation of several mechanisms to increase the security in a self-organized cyber-physical conveyor system, based on multi-agent systems (MAS) and build up with different individual modular and intelligent conveyor modules. For this purpose, the JADE-S add-on is used to enforce more security controls, also an Intrusion Detection System (IDS) is created supported by Machine Learning (ML) techniques that analyses the communication between agents, enabling to monitor and analyse the events that occur in the system, extracting signs of intrusions, together they contribute to mitigate cyberattacks.