Event-driven Job-shop Scheduling For Cloud Manufacturing

With the informatization and servicization of manufacturing industry, newmanufacturing models such as Cloud Manufacturing will become main directions of futuremanufacturing. There are many efforts on business application modes, services, resourceawareness, resource description and resource scheduling for Cloud Manufacturing, whichmake the theory of Cloud Manufacturing become mature. Therefore, it is necessary tocombine the theory of Cloud Manufacturing with actually Job-Shop scheduling, and make animplement to the conception of “Manufacturing as Service”. This paper focues on the JobShop scheduling problems for Cloud Manufacturing, and the main contents are as follows:Firstly, according to the characteristics of Job Shop scheduling for Cloud Manufacturingand the problems such as the collaboration of all kinds of equipments and related servicesfaced by Cloud Manufacturing, this paper introduces event driven concept and proposes theEvent-Driven Job-shop Service Architecture for Cloud Manufacturing (EDJSA4CM). For thesake of realization of this architecture, it elaborates the key technologies of EDJSA4CM, andcombines Fosstrak with Esper and Drools to solve the Job Shop scheduling problems in CloudManufacturing.Secondly, in view of various RFID events occuring during production, Esper platform isused to realize the detection and composition of such RFID events, which enables thearchitecture percept dynamically, and provides a foundation to the dynamic scheduling forCloud Manufacturing.Lastly, taking Job-Shop scheduling FT46as an example, the introduction of Esper allowsthe system to gain the information of workshop dynamically, and then Drools engine is usedto match event patterns to optimize the job shop scheduling. Finally, BPEL is used to makethe job shop scheduling as a service.Basically, this paper realized the function of Job-Shop based on Cloud Manufacture, sothat customers can use the workshop process services conveniently through networks in realtime.