Predictive maintenance should work for entire plants
A new system combines predictive maintenance with machine learning and expert knowledge, with the aim of preventing false alarms.
Share
The plant and process monitoring tool ‘Prerecognize’ from Israeli startup Visual Process analyzes a plant’s operating data across an entire year to generate a statistical model for the real-time analysis of future data. The difficulty here is in distinguishing between anomalies, such as planned downtimes or changes to the configuration, and impending malfunctions, and thereby preventing false alarms. For this reason, Prerecognize leverages a model that correlates current sensor values with the relevant values recorded on other days; a further model presents the findings of the statistical analysis as causal correlations. Ideally, each time an anomaly is detected, the system will be fed the causes. Thanks to its machine learning capabilities, it should become more accurate over time.
The Israeli startup that developed Prerecognize was acquired by German valve specialist Samson in June 2018 and is now developing the predictive monitoring and diagnostic tool SAM GUARD here, based on Prerecognize. SAM GUARD has been successfully used in practice to identify a leaking heat exchanger in a gas plant with 2,500 sensors. The analysis was based on the correlation between minor temperature fluctuations at the inflow and outflow pipes and the position of the valves.
Related Exhibitors
Interested in news about exhibitors, top offers and trends in the industry?
Browser Notice
Your web browser is outdated. Update your browser for more security, speed and optimal presentation of this page.
Update Browser