Task-service matching problem for platform-driven manufacturing-as-a-service: A one-leader and multi-follower Stackelberg game with multiple objectives

Abstract

Along with the increased use of digitization, platform-driven manufacturing-as-a-service (p-MaaS) is becoming an inevitable trend of the manufacturing industry. End-users openly share their personalized manufacturing tasks, which necessitates platform-based crowdsourcing to conduct manufacturing service collaboration and at last achieve efficient task-service matching (TSM). This crowdsourcing takes into account the autonomy of end-users, platforms, and manufacturing servicers, which challenges previous opinions that distributed manufacturing services must be centralized and controlled by platforms. This paper proposes a novel TSM problem for p-MaaS under the framework of crowdsourcing. The platform plays the role of allocating new emerged tasks and broadcasting to corresponding servicers. All servicers receive the broadcast information and conduct scheduling-based task acceptance (STA) independently. The above manufacturing task allocation (MTA) focuses on maximizing the net revenue of TSM and at the same time enables servicers to accept tasks as many as possible. In terms of the inherent interactive mechanism between MTA and STA, in which MTA generates a decision space for STA and STA feeds task acceptance schemes and the corresponding fulfillment costs back for use in MTA decision-making, a bilevel multi-objective optimization (BMO) is formulated to simultaneously address the two subproblems based on a Stackelberg game. The BMO is a type of multi-objective nonlinear programming, and a nested algorithm is designed to solve it. The better performance of the BMO is verified through a practical case study.