Amazon Web Services (AWS), an Amazon.com company, announced it is working with Pfizer to create innovative, cloud-based solutions.
AWS is helping Pfizer enhance its clinical manufacturing processes by incorporating predictive maintenance capabilities built with AWS machine learning services like Amazon Lookout for Equipment (AWS’s service for detecting abnormal equipment behavior by analyzing sensor data).
Pfizer can maximize uptime for equipment such as centrifuges, agitators, pulverizers, coaters, and air handlers used in clinical drug manufacturing. The overall focus of this collaboration is to support Pfizer in more rapidly and reliably producing new drugs and evaluating their potential health benefit for patients.
AWS is working with Pfizer to develop a prototype solution for detecting abnormal data points in its drug product continuous clinical manufacturing platform for solid, oral-dose medicines. The prototype solution uses Amazon SageMaker (AWS’s service for building, training, and deploying machine learning models quickly in the cloud and at the edge), Amazon Lookout for Equipment, Amazon Lookout for Metrics (AWS’s service for automatically detecting anomalies in metrics and identifying their root cause), and Amazon QuickSight (AWS’s scalable machine learning-powered business intelligence service for the cloud).
The machine learning models used in the prototype were able to provide early warnings for alarms with minimal false positives and direct users to the relevant signals. As a result, Pfizer can process data from the equipment and sensors involved in Portable Continuous Miniature and Modular (PCMM) manufacturing to detect anomalies as they occur, predict maintenance needs, and reduce potential equipment downtime.