STMicroelectronics Launches STM32Cube.AI – Empowering Edge Devices with Embedded Neural Networks

STMicroelectronics expands its STM32 microcontroller family and the STM32CubeMX ecosystem with advanced AI capabilities, enabling developers to harness trained neural networks for real‑time data classification from motion, vibration, environmental, microphone, and image sensors—outperforming traditional handcrafted signal processing.

STM32Cube.AI converts pre‑trained neural networks into highly efficient C code that calls optimized libraries on STM32 MCUs. The toolbox ships with ready‑to‑use function packs, including human activity recognition and audio scene classification examples, all immediately deployable on the ST SensorTile reference board and controlled via the ST BLE Sensor mobile app.

Additional engineering support is available through qualified partners in the ST Partner Program and the dedicated AI & Machine Learning STM32 online community, ensuring professional guidance from concept to production.

The FP‑AI‑SENSING1 software function pack delivers end‑to‑end motion (human‑activity recognition) and audio (audio‑scene classification) workflows. It leverages the SensorTile board for data capture, labeling, and neural‑network training, then runs inference on the same hardware. The BLE Sensor app serves as a remote controller and visual display.

With the STM32Cube.AI mapping tool, concise application examples, compact battery‑powered SensorTile hardware, partner program, and community support, developers can rapidly and confidently implement neural‑network solutions on STM32 devices.