Intel exec to outline open infrastructure for deploying physical AI in robots
Intel's Nagesh Puppala will speak at RoboBusiness 2026 about the infrastructure required to move physical AI from prototypes to production in robotics. He will emphasize open, edge-first architectures that support perception, sensor fusion, motion planning, and real-time AI across diverse hardware. The talk argues that scaling robotics innovation depends on practical deployment foundations, not just larger models or better simulations.
Puppala brings more than two decades of Intel experience, including taking the company's media business from a nascent effort to a fully scaled operation. His education spans an MBA from NYU Stern, a master's in computer science, and a bachelor's in mechanical engineering, combining business and technical expertise.
The RoboBusiness 2026 event celebrates the show's 20th anniversary, running October 20-21 in Santa Clara, with Puppala's session scheduled for the opening day. Intel's Robotics AI Suite, an open development platform built for Intel Core and Core Ultra processors, anchors the company's physical AI strategy. The talk addresses developers working across manufacturing, logistics, field robotics, humanoids, and autonomous systems.
The shift from model-centric AI to deployable physical systems could reshape how robots are built and adopted across industries. If Intel's open, edge-first approach gains traction, smaller robotics developers may find it easier to bring products to market without proprietary lock-in. This could accelerate automation in manufacturing and logistics, potentially affecting employment patterns and workplace safety. However, the emphasis on real-time reliability and safety suggests that widespread deployment may still face significant hurdles before robots become commonplace in everyday environments.