Traditional control approaches are based on centralised and hierarchical structures, following the well-known ANSI/ISA-95 (2010), also known as automation pyramid, which presents good production optimisation, but a weak response to condition change and reconfigurability due to the rigidity, monolithic and centralisation of their control structures (Colombo et al. 2021). These control approaches are not designed or prepared to respond to the current demanding requirements of responsiveness, scalability, reconfigurability and robustness (Leitão 2009a). The advent of Industry 4.0 (Kagermann et al. 2013 ; Leitao et al. 2020), led to the use of distributed control structures, the use of the decentralisation of control nodes and the introduction of intelligence to transform the existing assets into smart processes and machines, and also considering smart products (Barbosa et al. 2016) as important players in this ecosystem. Multi-agent Systems (MAS) and holonic systems are suitable to face these demanding requirements since they offer an alternative way to design and implement such innovative control systems taking advantage of their
capability to decentralise the control over distributed structures towards modularity, scalability, robustness, fault-tolerance, reconfigurability and re-usability (Leitao, Karnouskos, Ribeiro, Lee, Strasser and Colombo 2016 ; Karnouskos and Leitao 2017). The development of Industry 4.0 compliant solutions will constitute a new opportunity
to use MAS and holonics to realise these innovative, and emergent industrial CPS (Colombo et al. 2015 ; Leitão and Karnouskos 2015 ; Leitao, Karnouskos, Ribeiro, Lee, Strasser and Colombo 2016 ; Colombo et al. 2017 ; Karnouskos et al. 2020). This chapter introduces the main conceptual foundations of MAS and holonic systems and presents the framing of industrial agents as an instantiation of such technological paradigms to face industrial requirements. The alignment of industrial agents with RAMI 4.0 (Deutsches Institut fuer Normung (DIN SPEC 91345) 2016) is also
addressed and in particular, it is discussed the use of industrial agents to realise industrial Cyber-Physical Systems (CPS) and concretely Asset Administration Shells (AAS), which play an important role in the development of Industry 4.0 components (Boss et al. 2020). At the end, the chapter discusses some research challenges and directions that current raise from the deployment of industrial agents.
