Is it possible to bring cloud capabilities closer to local devices? Using Edge computing, data can be processed close to the source where data is generated.
Edge Computing addresses two critical scenarios, one where the devices are in remote locations with limited connectivity and second when the customers do not want to send their data to cloud for security reasons. In both the scenarios customer would like to have all the cloud capabilities implemented locally close to the devices.
With the help of AWS Greengrass, PROLIM was able to build a resilient system for predictive and prescriptive maintenance of Windmill and Solar Farms. Edge computing enables you to implement all the cloud capabilities close to the devices, which is very critical for remote locations with limited connectivity.
Sensor data (temperature, humidity and velocity) is collected and published to local Greengrass Core Device using MQTT protocol. A Lambda function which is running on Greengrass Core subscribes to that topic payload (Sensors Data) and stores it on local file storage. Sensor data is monitored continuously, and alerts are triggered based on various rules. In critical situations, device shutdown command will be transmitted to the device. As data is being collected, it is analysed for future Predictive Maintenance, so complete breakdown of the Windmill breakdown completely.
With the help of AWS Greengrass, we were able to build a resilient system for predictive and prescriptive maintenance of Wind Mill and Ore Mines. AWS Greengrass can really help you to build a complete eco system on the top of Edge Devices where you can perform different event-driven serverless operations and also you can write Machine Learning inference to prepare Local ML models to make the Edge device more powerful for continuous monitoring of basic equipment’s required for workers to work on such hazardous environment.
These services grant us the super power to act upon anomalies and critical situations instantly with a fast response time thereby avoiding casualties or further damage in the system while reducing the maintenance cost with almost zero downtime.
