Article Preview
Top1. Introduction
Cloud-based Manufacturing (CM) is a new production paradigm for the intelligent networking of geo-distributed manufacturing capabilities and resources. With the aid of the Internet, CM users can request for any kind of manufacturing services (e.g. design, maintenance, and construction services) as a cloud service (F. Tao, Zhang, Venkatesh, Luo, & Cheng, 2011). Different advanced technologies, such as the internet of things (IoT), cloud computing, virtualization, and service-oriented architecture (SOA) has enabled CM to create temporary and flexible production lines based on the users’ requirements and a combination of distributed and heterogeneous virtual resources (Wu, Greer, Rosen, & Schaefer, 2013).
Although Cloud-based manufacturing offers many benefits to its stakeholders, such as enhancing the efficiency of resource loading (Wu et al., 2013), improving collaboration(X. Xu, 2012), cooperation (T. Li, Gupta, & Metere, 2017), and increasing the intelligence of the system through big data analysis (Psannis, Stergiou, & Gupta, 2018); it also faces various kinds of incidental and intentional threats(Jouini & Rabai, 2016) and changes as a result of the complexity and dynamism of its operating environment and the heterogeneity of resources (Wu, Rosen, & Schaefer, 2014; W. Xu et al., 2012). Accordingly, security (Stergiou, Psannis, Kim, & Gupta, 2018), reliability, and safety (Wu et al., 2014) of the Cloud-based system are regarded as critical issues.
Service composition is a key and complex process of CM (Loskyll et al., 2011; F. Tao et al., 2011; Xiang, Hu, Yu, & Wu, 2014), whose results should be implemented in a completely uncertain and dynamic environment in the presence of a variety of risks (Fei Tao, LaiLi, Xu, & Zhang, 2013a) such as changes in job requirement, cancelation of submitted jobs, variations in quality of service (QoS) attributes, service disruption, entry of new services, and dependencies between services (L. Zhang, Guo, Tao, Luo, & Si, 2010).Therefore, the resilience of composite services at runtime to the disruptive internal and external factors is considered as an important property of the service.
In recent years, although resilience engineering in various areas of management and engineering has attracted much attention, however, the knowledge of analysing, designing and managing resilience is in the early stages (WJ Zhang & Van Luttervelt, 2011). The results of the authors’ review of research papers published between 2010 and 2018 on the service composition in Cloud-based manufacturing revealed that the focus of the researches has been on QoS-aware service composition, mainly considering overall time, cost, energy, reliability, availability, trust, reputation, security, and maintainability of the services (Chen, Dou, Li, & Wu, 2016; Huang, Li, & Tao, 2014; Jin, Yao, & Chen, 2017; Li, Guan, Liu, Ma & Zhang, 2018; Li, Jiang & Ge, 2014; Liu & Zhang, 2017; Lu & Xu, 2017; Que, Zhong, Chen, Chen & Ji, 2018; Tao, Li, Xu & Zhang, 2013b; Xu et al., 2017; Yongxiang, Xifan, Jie & Bin, 2013; Zhang, Yang, Zhang, Yu & Chen, 2017; Zhang, Yang, Zhang, Yu & Li, 2017; Zhou & Yao, 2017a, 2017b, 2017c, 2017d, 2017e). The QoS-aware composition is a multi-objective NP-hard optimization problem which seeks the best execution paths, consisting of several atomic manufacturing related Cloud services.