Optimising Pressure Testing in Manufacturing: Machine Learning for Energy Efficiency and Sustainability

This study explores the application of Machine Learning (ML) techniques to optimise pressure drop tests, which are vital for assessing quality and ensuring the integrity of products during manufacturing. Traditional pressure testing methods are energy-intensive and time-consuming, posing industrial efficiency and sustainability challenges. Leveraging a dataset of 1.7 million test records of fuel pumps collected over 15 months, this study developed predictive models that significantly reduced test durations—from an average of 10 seconds to as low as 2 seconds, thereby reducing energy consumption and increasing efficiency. The Random Forest model demonstrated superior performance, achieving an RMSE of 0.0017 and an R² score of 0.9925. This research underscores the potential of ML in transforming manufacturing processes, offering enhanced quality assurance, operational efficiency, and environmental sustainability. These findings provide practical solutions for improving energy efficiency and advancing quality control in manufacturing processes.