The objective of the thesis is to design, develop and test a multi-agent intelligent system in the context of industry 4.0, aiming at improving different aspects of the production process, such as anomaly detection, advise actions and facilitate monitoring of the machines. This system, called NTMAS, wants to provide a partial automation of the process exploiting a Human-in-the-loop model, in this way the human supervisor has a complete control on what is happening on the plant. Each agent action must be validated by a supervisor before it takes effect, this is also a way to improve its decisions. Moreover, it comes with a web dashboard that allows user to interact with the agents and monitor the system.

The NTMAS will be tested in a partially simulated environment, derived from a real dataset. In this way it will be possible to analyse the behaviour of the agents, in particular their ability to communicate and to take decision autonomously. At the end, an evaluation of the NTMAS advantages and limits is performed.