2026 Agenda

As Physical AI systems move from controlled environments into the real world, the limitations of data center-bound intelligence become increasingly clear. This presentation explores how the “agentic edge” — distributed, autonomous decision-making at the point of interaction — is emerging as a critical paradigm for scaling physical AI. By embedding intelligence directly into edge devices such as robots, sensors, and industrial systems, organizations can achieve lower latency, improved resilience, and greater operational autonomy. We will examine the architectural shifts required to enable agentic behavior at the edge, including advances in on-device models, real-time data processing, and decentralized coordination frameworks. The session will also highlight key challenges around orchestration, safety, and continuous learning in distributed environments.