Trust-based resource sharing mechanism in distributed manufacturing

• Published in: International journal of computer integrated manufacturing Vol. 33; no. 1; pp. 1 - 21
• Main Authors: Szaller, Ádám, Egri, Péter, Kádár, Botond
• Format: Journal Article
• Language: English
• Published: Taylor & Francis 02.01.2020
• Subjects: collaborative manufacturing; Distributed manufacturing control; distributed multi-agent systems; resource sharing; simulation; trust
• ISSN: 0951-192X; 1362-3052
• DOI: 10.1080/0951192X.2019.1699257

Manufacturing has moved from local towards global economy in the past decades enabling new paradigms to come into practice, such as distributed manufacturing which aims at increasing companies' flexibility with a decentralized system consisting of autonomous entities. In the paper, a distributed collaboration framework of manufacturing agents is introduced, where the members with resource shortages can request resources from others, divide requests among multiple agents, reorganize their production to be able to complete a request and cancel an undertaken task if needed. In a collaboration framework, it is essential to have a commitment to the promises: if participants cannot count on these commitments, the framework's efficiency may decrease. So as to motivate agents to keep their promises and to enable differentiating between reliable and non-reliable partners, here agents consider trustfulness during the selection from resource offering agents' proposals and make decisions considering subjective trust and public reputation values, which are computed based on successfulness of task performing and meeting due dates. In the paper, the impact of the proposed mechanism is investigated with multi-agent simulation. It is shown that considering trustfulness improves the overall system performance; and the improvement depends on the number of participants and the federation's load.