IoT Voice and Edge

Message: all-ST end-to-end solution and reference design for Smart Home, including Voice frontend, Cloud connectivity and Edge computing. The current implementation works with Amazon Voice Services (AVS) and Amazon Web Services (AWS).
Description: Smart Home Voice control with BlueCoin or SensorTile via BlueVoice link to Amazon AVS Gateway on DiscoveryF769; BLE IoT devices managed by local Gateway by means of Linux version of BlueST SDK. Voice interactions with the IoT nodes through Amazon Alexa using custom Skills and Lambda functions.


Predictive Maintainance:

Predictive Maintenance Platform (PMP) is an application for monitoring industrial equipment which involves sensing vibration data from motors (e.g. pumps), collecting data on a gateway, and sending them to the cloud for visualization on a dashboard.All manufacturing equipment with moving parts are subject to degradation which require servicing or component replacement, but traditional maintenance approaches based on set schedules ignore actual equipment condition. In Condition-based Monitoring, maintenance is instead scheduled according to the estimated condition of the machine from inspection or from sensor data. Predictive Maintenance makes a step forward: it implies feeding sensor data into dynamic predictive models, in an attempt to foresee maintenance intervention in the future before a real failure happen. This can translate into more efficient maintenance planning, less machine down time and longer operating life.
The architecture here proposed is based on an IO-Link capable STEVAL-IDP004V1 master board and up to four IO-Link STEVAL-BFA001V1B smart sensor nodes, which collect vibration data from motors and send them to the Amazon AWS cloud through an STM32MP1 edge gateway. The gateway comes with an OpenSTLinux distribution that runs the AWS Greengrass edge computing service, which enables the execution of local logic directly on the gateway. Moreover, a web-based dashboard allows an operator to manage device provisioning, setting thresholds for anomaly detection, and investigate the equipment conditions.