Transcript
With hundreds or even thousands of sensors in our facilities, operators often get fatigued by the sheer number of alerts they receive. Alarms are typically lagging indicators of a problem. We need better ways to predict real issues earlier with fewer nuisance alarms.
Acting early is critical to avoid facility downtime, loss batches and expensive equipment fixes. This real-life demonstration shows how SmartFactory Rx predicts adverse events more accurately and earlier than ever before. Here we’ve used engineer knowledge of the sterilization cycle on a bioreactor based on past data to build a picture of the normal rate of heat transfer.
SmartFactory Rx then monitors this part of the SAP recipe each time it runs. The HMI here is showing the sterilization loop on the bottom left. We’ve simulated a fouled heat exchanger by manually throttling the flow through the steam valve.
The bioreactor will still reach temperature, so none of the traditional alarms will be triggered, but it will take longer. SmartFactory Rx is going to see this issue and notify us of the problem via email. And you can see here a new email has just been generated to the maintenance team.
It’s showing that the heat exchanger is only 18% effective based on that heat transfer equation. There’s also a prescriptive suggested set of actions and a history of the last work order done on this vessel. From there we can click into a simple web interface to understand key contributors behind this issue and over time the system can learn behavior and alert you to action earlier and earlier.
SmartFactory Rx is making predictive maintenance simple through web-based tools that give clear information to operators, allowing them to act more decisively.
