Platform-based task assignment for social manufacturing (PBTA4SM): State-of-the-art review and future directions

Abstract

Abstract

Mass individualization is calling for a more sustainable manufacturing paradigm which can address the paradoxes of diversity, complexity, and affordability. Social Manufacturing (SM) represents a democratized servitization trend trying to reshape the traditional production relationship between consumers and manufacturers. To achieve the SM visions, new operational mechanisms for SM should be constructed to overcome the challenges of information sharing, accuracy, efficiency, security, sovereignty, etc. The survey found that task assignment (TA) is one of the foundational mechanisms for the implementation of regular autonomous manufacturing systems, as well as the role of TA is further amplified for distributed collaborative environments. Therefore, inspired by the relevant research of management science, Platform-based Task Assignment (PBTA) is proposed to distinguish and conceptualize this different research topic. In SM platforms, the diverse capacities and resources can be shared, so that knowing "who can do” and “select whom to do" is more important than knowing "how to do". Furthermore, the studies on TA for SM present a difference from the previous studies on TA in manufacturing. From a perspective of supply-demand mapping, PBTA illustrates the foundational operational mechanism for SM attracting many researchers’ attention from different fields. Meanwhile, research on PBTA is also required for the platform practices in the era of digital, shared, and platform economy. Given the academic importance and practical value, this survey carefully selects 250 valuable research articles relevant to PBTA for SM. A novel workflow model and knowledge framework, namely PBTA4SM, is proposed to identify and organize the critical issues and challenges. This study shows the state-of-the-art research advancement of PBTA including task design considerations, modelling methods, typical engineering problems, algorithms, decision patterns, key activities, and governance mechanisms. Finally, we complete this holistic survey by highlighting eight potential directions for future research in the Generative Artificial Intelligence (GAI) era.