Cyient‘s client, a leading laboratory equipment manufacturing company, collects data from their Ultra Low Temperature (ULT) freezers deployed across the globe using multiple sensors. They faced challenges in managing this massive data and could therefore barely use it to identify critical failures in the freezers so that immediate maintenance action could be taken. In spite of their huge effort, maintenance was always reactive in nature and the company failed to keep the system up all the time.
While the data collected from freezers gave indications of past failures and potential issues, often the delay in extracting these issues from the data resulted in increased gravity of failures. The company thus contended with greater maintenance effort and cost.
The client targeted early detection of issues and the ability to predict critical failures in advance to significantly improve the reliability of these equipment and customer satisfaction.
Cyient identified the opportunity to build an analytics engine that can predict impending failures well in advance, so that proactive maintenance action can be taken to avoid system outages. Analyzing the historical data from numerous freezers, they built specific patterns for critical failures. Cyient created an analytical model that can detect the possibility of such failures in any other freezer from hours to days in advance. This is achieved by running the model using current data from the freezer.
Cyient also built a comprehensive system health dashboard that gives insights into the nuances of freezer behavior from multiple perspectives. Using visual analytics, the dashboard shows the impact of various internal and external factors on the health of the freezer. The factors could be power fluctuations, door openings (user behaviour) and ambient temperature. The system enables early issue and failure detection by uncovering system abuse, and providing quick issue categorization.
Find out more about how Cyient used predictive analytics to help solve their client’s problem. Click through to access the complete case study.