How it works - Amazon Monitron

How it works

Amazon Monitron is a machine learning end-to-end condition monitoring solution system that detects developing faults within machinery, enabling you to implement a predictive maintenance program and reduce lost productivity from unplanned machine downtime. Amazon Monitron includes purpose-built sensors to capture vibration and temperature data, gateways to automatically transfer data to the AWS Cloud, and an application for system set up, analytics, and notification when tracking equipment condition. Reliability managers can quickly deploy Amazon Monitron to track machine health of industrial equipment, such as bearings, motors, gearboxes, and pumps, without any development work or specialized training.

The Amazon Monitron workflow

The following diagram shows the basic workflow of Amazon Monitron.

  1. An Amazon Monitron sensor captures temperature and vibration data from the equipment (the asset) and transmits it to the gateway.

  2. An Amazon Monitron gateway transmits the data to the AWS Cloud using the factory's internet connection.

  3. The Amazon Monitron ML-based service in the AWS Cloud analyzes the sensor data.

    1. Amazon Monitron looks for abnormalities in the data that could indicate developing faults.

    2. If Amazon Monitron finds potential failures, it notifies reliability managers and technicians through the Amazon Monitron app so they can take appropriate action.

    3. Technicians investigate based on the alerts, and resolve the developing fault. They enter feedback on the accuracy of the alerts, and report the failure mode, cause, and action taken in the app. Amazon Monitron learns from this feedback and continually improves.

  4. The app displays current and past temperature and vibration data in charts that are easy to understand and can be used while investigating an issue.