Asset Summary:
Edge AI resources are most useful when they show how intelligence can move closer to the signal, sensor, and system decision. Avnet’s new edge AI resources bring together solution papers, market insights, and supplier partner content to help design engineers evaluate real constraints in production-ready designs.
One featured resource from STMicroelectronics explores the Intelligent Sensor Processing Unit, or ISPU, and how AI can run directly in the sensor for ultra-low power edge applications.
In the paper, engineers will find:
- How MEMS sensors have evolved from connected data sources to standalone decision points
- Why in-sensor AI can reduce dependency on high-bandwidth, always-on connectivity
- How ISPU integrates sensing and programmable processing in the same package
- Power and performance considerations for tinyML and sensor-level inference
- Development options using C, graphical tools, NanoEdge AI Studio and ISPU toolchains
- Example use cases including gesture recognition, presence detection, anomaly detection and sensor
fusion
For teams designing battery-powered, always-on or distributed sensing systems, the paper shows how local processing can reduce power, latency and data movement while keeping intelligence close to the sensor.