Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
509248 | Computers in Industry | 2012 | 18 Pages |
Due to the impossibility of predicting with certainty the occurrence of disruptive events, buffers defined to obtain a robust schedule could not absorb all the changes. Then, local modifications of the schedule are usually performed to avoid a new planning task. For this task, obtaining disruptive event information in advance can help to make better decisions. As a result, ability to predict disruptive events that affect the execution of the supply process an order represents is required. With the objective of satisfying this requirement, this work proposes a model driven development approach based on a reference model to automate the generation of the monitoring model of a supply process able to anticipate the occurrence of a disruptive event by monitoring variables that can explain it.The approach proposes both a reference model to represent the monitoring model independently of the implementation platform, and a specific model to represent the monitoring model with the particular language of the implementation platform. An engine based on transformation rules allows automating the generation of a platform dependent monitoring model from an instance of a platform independent metamodel. The monitoring component of a SCEM system has been developed, which implements the transformation engine as a Bayesian Network model, and uses an appropriate tool to execute it. For an empirical validation of the model three case studies are presented.
► A model driven development approach based on a reference model to automate the generation of monitoring models. ► Monitoring model for predicting disruptive events in supply processes. ► Reference model to represent the monitoring model independently of the implementation platform. ► Specific model to represent the monitoring model as a Bayesian Network.