A framework for task allocation in IoT-oriented industrial manufacturing systems

Industrial production performance relies on the efficiency of the machines and the effective management of the allocated jobs. Job allocation and workflow management are significant tasks in the industrial environment. This paper introduces a novel job processing framework for the industrial environment aided by Network-in-Box (NIB). The process control layer of the industry environment and NIB architecture is coupled to manage job allocation and process handling. Considering the functional machines? completion time, Q-learning based efficiency assessment is performed for different states of the operating machines. By pre-determining the forms of the machines, the efficiency and job allocation of the machines are determined. This prediction based job processing framework is reliable in offloading allocated jobs to the machines with high efficiency as per the NIB architecture?s decision-controls. This prediction is sent as a control to the process layer to give the jobs or handle sequential jobs in the industrial environment. The active machines? fault tolerance is observed and assessed using their physical attributes and output of the previous session.Knowing the states of the machines is useful in predicting the efficiency of the machines and hence, the machines will be able to complete the allocated and offloaded jobs in time.In the experimental analysis, the machines? processing rate is 1 log per hour, and the efficiency of the machines is monitored over 24 hours of operation in 2 days.

Automated summary

This paper introduces a novel job processing framework for the industrial environment aided by Network-in-Box (NIB), improving the efficiency of industrial production. By predicting the forms and efficiency of the machines, the operating states and completion time of the machines can be forecasted, thus enhancing job efficiency.