Developers are exploring ways to incorporate Artificial Intelligence into their products. Thanks to ST and Blues, deploying intelligent devices that can learn and process data on the edge is now accessible to any developer.

This webinar demonstrates how to create complete Edge ML solutions using STMicroelectronics' NanoEdge AI Studio and Blues' connectivity offerings, building machine learning models without requiring Data Science expertise.

Why Edge ML Matters

Edge ML enables devices to process data locally rather than relying on cloud-based analysis. This enhances privacy by keeping sensitive data on-device while reducing latency and bandwidth requirements for real-time decisions.

ST's NanoEdge AI Studio democratizes Edge ML deployment with an intuitive UI that any developer can use to create ML solutions including anomaly detection, outlier detection, classification, and regression capabilities.

Complete Edge ML Workflow

Creating Your First Model

Watch experts demonstrate building an Edge ML solution from scratch using NanoEdge AI Studio with sensor inputs, specifically a triple-axis accelerometer for detecting vibration patterns and anomalies.

The hardware this demonstration uses:

• Triple-axis accelerometer (Adafruit LIS3DH) for data collection
• STM32-based Swan microcontroller for Edge ML model deployment
• Notecard Cell+WiFi for secure cloud connectivity

Learn the four ML models supported by NanoEdge AI Studio:

• Anomaly Detection: Identifying unusual sensor data patterns
• Outlier Detection: Spotting deviations from normal behavior
• Classification Models: Categorizing inputs into predefined groups
• Extrapolation Models: Predicting future values from historical data

Connecting to the cloud:

Creating an Edge ML model is just the beginning. Real value comes from securely syncing generated inferences with cloud services. This webinar shows how to bridge that gap using connectivity offerings from Blues.

Generate real-time alerts:

Configure your Edge ML system to generate intelligent alerts when anomalous behavior is detected, including setting up Notehub to monitor data and trigger alerts based on predefined thresholds.

30-Minute Build Process

This hands-on recording demonstrates the complete development process:

• Set up NanoEdge AI Studio for Edge ML development
• Collect and prepare training data from sensors
• Train anomaly detection models optimized for microcontrollers
• Deploy compiled libraries to STM32-based devices
• Integrate cloud connectivity for remote monitoring

Ready to start building? Watch this on-demand webinar to master the complete workflow from model creation to cloud-connected deployment.