Running AI at the Edge: NXP on Tiny Models, Big Challenges & the Future of Human–Machine Interaction

In this Electropages podcast, host Robin Mitchell is joined by Ali Ors, Global Director of AI and ML Strategy and Technologies for Edge Processing at NXP Semiconductors, to discuss the technical and design challenges of running advanced processing at the network edge. They examine how to implement demanding workloads on microcontrollers and embedded processors with tight constraints on power, memory, and compute, and how techniques such as model optimisation and quantisation are enabling practical deployment in automotive, industrial, and IoT systems. The discussion also covers the growing importance of natural, conversational interfaces, the balance between versatile CPUs and dedicated accelerators, and how hardware flexibility supports long-term product viability. Ali shares insights into securing devices in the field, protecting against evolving threats, and ensuring reliable operation over extended lifecycles. This episode offers an in-depth look at the engineering strategies and hardware considerations shaping the future of intelligent edge devices.