FEIT Research Project Database

Supply Chain Quality Control on the Cloud Using Machine Learning

Project Leader: Kevin Otto
Primary Contact: Kevin Otto (kevin.otto@unimelb.edu.au)
Keywords: industry 4.0; Internet of Things (IoT); machine learning; process mining; process modelling
Disciplines: Mechanical Engineering

This research is to develop and analyze process instrumentation and data collection for in-situ process conditions of equipment and workstations in a production supply chain, including alternative schemes for individual part coding and tracking.  Integration with cloud based time series databases.  Use with process simulation tools and uncertainty quantification studies, including variance based sensitivity analysis and Shapley values. Research alternative methods to identify defect and performance deviations patterns in the assimilated data.  Exploration of machine learning classifiers suitable for datasets with few defects, including decision trees, oversampling techniques and gradient boosting.  Study interactions among factors early in part fabrication of supplier operations and factors later in system assembly and operation.