AWS and Klika Tech Introduce New Model for IIoT Anomaly Detection
April 15, 2021
Klika Tech and AWS during embedded world 2021 demonstrated a development kit for deploying predictive maintenance in Industrial Internet of Things ecosystems (IIoT).
Built on AWS and featuring STMicroelectronics, the kit introduced a machine learning model for automated response to industrial equipment anomalies.
The March 1 demonstration was one of five live presentations AWS hosted under the umbrella of bringing cloud connectivity and intelligence to the IoT edge during embedded world 2021 The event was presented live by Anton Shmagin, Partner Solutions Architect at Amazon Web Services specializing in IoT, and Konstantin Meshcheryakov, Solution Architect, Klika Tech.
"A crown jewel of this demo is the latest release of AWS Greengrass, Version 2, which offers a secure messaging bus, local processing, streamlined Machine Learning models for deployment to an entire fleet, Over-the-Air updates, and an identity broker," Shmagin said during the presentation. "But the biggest benefit is the flexibility it provides."
The kit was designed to increase the speed of equipment-level data processing and reduce sensor-to-cloud latency and bandwidth. Per the companies, it helps industrial operations eliminate unplanned downtime and monitor operations and equipment maintenance. The demonstration simulated response to industrial equipment anomalies and illustrated wireless connectivity for data collection and transfer, in addition to enabling edge device capabilities including local compute, messaging, data caching, sync, and machine learning inferences.
The kit features data collection from sensors connected to STMicroelectronics' P-NUCLEO-WB55 multi-protocol, ultra-low-power devices for Bluetooth Low Energy (BLE) integration. That enables wireless transmission of data to STMicroelectronics' STM32MP1 Series Microprocessing Unit (MPU) and embedded evaluation boards running Amazon Greengrass Version 2. The MPU sends data to AWS IoT Core and a user dashboard, enabling sensor-to-cloud real-time monitoring of equipment performance, advanced system fault diagnosis and predictive analytics.
Presenters said the kit enables operators to attach a range of sensors to industrial equipment such as motors, gearboxes, fans, or pumps to measure actions such as vibration, volume, temperature, humidity, and more. The kit is capable of providing operators with real-time assessment of industrial equipment anomalies and providing competitive speed combined with ease of use.
The all-virtual event was the fourth-annual embedded world pairing of AWS and Klika Tech, an IoT and cloud product and solutions development company headquartered in the U.S. Klika Tech has published many of the demos within its portfolio of IoT and cloud solutions at https://klika-tech.com/portfolio.