Intelligent condition monitoring for small induction motors

点击翻译按钮获取中文摘要

Addos Technologies is an Edmonton-area provider of Industrial Automation and System Integration services.They are currently developing a line of business in condition monitoring of small electric motors in industrialfacilities. Only low-cost sensors are economically viable for this problem; Addos believes that inferentialsensing based on machine learning will provide competitive sensitivity and specificity using these sensors, butthey lack the in-house expertise to conduct the necessary research and development.The proposed research will evaluate Addos' planned approach to condition monitoring. Addos has collectedmachine condition data from both laboratory experiments (in which faults are deliberately induced), and fieldobservations at an industrial site. We will use these datasets to train and evaluate inferential sensors based on avariety of traditional and deep learning algorithms. Once the best-performing algorithm is identified, Universityand Addos personnel will work together to productize it. We will hold a three-day workshop with the Addosand University personnel, during which the Addos personnel will first be instructed in intelligent systemsdesign, and then the entire team will work on the productization as an extended example thereof. At theconclusion of the workshop, Addos should be able to design and implement intelligent condition monitoringsolutions for other clients, giving them a new line of business. Assuming this project is successful, Addosexpects to hire 20-30 additional personnel over the next five years to meet the additional demand for theirservices.