Solution Description
This solution is based on a smart sensor node built around an ultra-low-power STM32WB5MMG 2.4GHz wireless module combining MEMS sensors and embedding AI library to detect anomalies and perform classification in monitored equipment and sending the results to a mobile app or a PC terminal console.
The anomaly detection algorithm running on the STM32WB5MMG module processes the raw data from the 6-axis ISM330DHCX iNEMO inertial module with Machine Learning Core, Finite State Machine and digital output or the low-noise, 3-axis IIS3DWB digital vibration sensor with an ultra-wide bandwidth up to 6 kHz to detect anomalies in early stages.
The Embedded AI is a machine learning library generated by the NanoEdge AI Studio software tool, starting from sensor raw data. The library, running locally on the STM32WB5MMG, embeds preprocessing and on-device learning capabilities in order to learn the normal behavior of the equipment and then proceed with the anomaly detection.
To ensure the detection of anomalies remains accessible, keeping the user aware of the status of each machine remotely, the smart sensor node is connected via Bluetooth LE to the STBLE Sensor mobile app (available for Android and iOS). You can then set sensor and library parameters, control learning and detecting phases, and monitor equipment fault states in a very user-friendly view. Moreover, in Advertise mode, the app displays a complete overview of the status of all the smart sensors throughout the factory at all times. This way, nodes can be connected only when needed, in the event of an anomaly, further saving power and extending battery life.
-
Key Product Benefits
STM32WB5MMG - Ultra-low-power wireless microcontroller
Ultra-low-power, small form factor, certified 2.4GHz wireless module with a dual core (Arm Cortex-M4 and Cortex-M0+)
IIS3DWB - Ultar-wideband vibration sensor
Ultra-wide bandwidth (up to 6 kHz), low-noise, 3-axis digital vibration sensor highly suited for vibration analysis for fault detection up to 6 kHz, both in terms of performance and cost
ISM330DHCX - high-perfromance 3D gyroscope
Inertial module featuring a high-performance 3D digital accelerometer and 3D digital gyroscope tailored for Industry 4.0
STBC02 - Li-Ion linear battery charger
Li-Ion linear battery charger with LDO, load switches and reset generator
ST1PS02 - step-down converter
400 mA nano-quiescent synchronous step-down converter with digital voltage selection, Power Good and AUX switch
NanoEdgeAIStudio - NanoEdge AI Studio
NanoEdge AI Studio can generate four types of libraries: anomaly detection, outlier detection, classification, and regression libraries. Does not require data science or AI skills.
-
All Features
- Firmware package to develop a sensor node for predictive maintenance applications, featuring motion sensors and process monitoring
- Compatible with NanoEdge™ AI Studio solution, to enable AI-based solution
- Application for datalogging in binary format
- Compliant with Bluetooth® Low Energy SIG v5.2, ZigBee 3.0, and IEEE 802.15.4-2011 specifications
- Compatible with ST BLE Sensor App for Android & iOS, for the following actions:
- Sensor settings (FS and ODR for ISM330DHCX)
- NEAI Library settings (Sensitivity, Anomaly Threshold, Timer, Signals)
- Start - Stop Learning and Detecting phase for NEAI Library
- Battery Status monitor
- Over-the-air firmware updates (FUOTA)
