Context-aware manufacturing system design using machine learning

With the development of computer, automation and information technology, workers have more challenges to take care of several devices at the same time. Under this situation, context-aware manufacturing system is proposed to help users capture the most relevant information and make the decision timely. Due to the increased demand for small-batch customized products, manufacturing resources and products frequently change, and this leads to variation of context in manufacturing. Traditional rule-based context-aware manufacturing systems need their rules to be modified manually, which is time-consuming and error-prone under the current variability of the market. To create a framework for updating the context-aware logic automatically, this paper presents a novel notion of applying machine learning techniques in the context-aware manufacturing system design. For the proposed context-aware manufacturing system, components comprising a context model for the manufacturing domain, a machine learning based calibration framework and a context extraction module are designed to improve the update efficiency with less costs. Finally, a test manufacturing scenario is simulated to verify the feasibility of applying machine learning algorithms in context awareness.