The growing potential of artificial intelligence (AI) and machine learning (ML) in embedded systems is driving new application solutions and products, but developing AI-based systems can be challenging.
STMicroelectronics' NanoEdge AI Studio is an ML technology that accelerates innovation, allowing you to create the perfect ML library for your AI project based on minimal data, which can also help you develop your AI solution. In this video, we collect current and vibration meter information to train an ML model to detect anomalies in the system.
NanoEdge AI Studio is a PC-based, push-button development studio for developers that runs on Windows, Linux, or Ubuntu. The platform allows you to create four optimized tinyML libraries for anomaly detection, outlier detection, classification, and regression, without requiring advanced data science skills.
These libraries can be combined and chained together to create a complete edge AI solution with anomaly or outlier detection to detect equipment issues, classification to identify the cause of the problem, and regression to extrapolate information and provide real insights.
Inputs include vibration, pressure, sound, magnetic field, time-of-flight, and more, and multiple signals from multiple sensors can be combined in a single library or across multiple libraries simultaneously. Training and inference are performed by the microcontroller using the NanoEdge AI on-device learning library, streamlining the edge AI process and significantly reducing development efforts.
