AI Summary & Key Takeaways
AI InsightThe experimental results showed that our proposed approach achieved around 98% in the R2 performance metric.
📌 3 Key Takeaways
Abstract
Sustaining productivity with guaranteed machine availability is of the utmost significance while reducing costs. With the rising technology and the collected data in the industry, accomplishing such a goal is not fictional anymore. This paper proposes an artificial intelligence-based model that predicts the remaining useful life (RUL) of the plastic injection molding machines before requiring maintenance. Data collected from machines in production via sensors is preprocessed by performing various techniques, and anomalies in the data are detected and cleaned. Based on the historical data, the RUL of the machine, which is the duration until maintenance is required, is calculated, and the data is labeled with the RULs accordingly. In the proposed method, the labeling step is followed by feature engineering where the useful features are extracted from the raw data, such as entropy, peak to peak, and crest factor. A feature selection method is also applied to determine their contribution to the estimation accuracy of the RULs. As a comparison, we experimented with various regression models along with various evaluation metrics. The experimental results showed that our proposed approach achieved around 98% in the R2 performance metric.