Successfully implementing AI on edge devices and equipment is far more challenging than simply porting a machine learning model to your hardware platform. Many edge AI use cases have the need for specialized data collection for model training and optimization. Designs at the edge are constrained by limited power budget, thermal concerns, small size, and cost sensitivity. Implementing AI at the edge is a demanding endeavor which requires foresight of the many potential perils. This talk will share experiences of successful AI implementation on the edge with a variety of use cases. It will also showcase solutions that you will need, from software development platforms to silicon. This includes our most recent technology acquisition, Hailo.
Field Applications Engineer at Microchip Technology Inc
