Autonomy is advancing rapidly across multiple industries—including passenger and commercial vehicles, drones, robotics, and industrial systems—and this progress depends fundamentally on intelligent perception: the ability to transform raw sensor data into timely, reliable understanding. In this keynote, we will consider the development of human mobility and the parallels it offers to the evolution of vehicle architectures. Specifically, we will trace the progression from distributed, loosely coordinated control to increasingly centralized compute, and ultimately to hybrid approaches that combine high-performance central processing with intelligent sensing at the edge. We will then examine what this transition enables, including faster response times, scalable software-defined capabilities, improved safety observability, and reduced power consumption, and discuss the implications for building the next generation of autonomous systems.
