Let Edge Devices Think and Decide!

The exponential growth of edge devices —ranging from sensors to autonomous micro-robots—demands a paradigm shift from centralized cloud-based intelligence to learning and inference at the edge. This keynote argues that edge devices should not remain simply data producers but evolve into context-aware agents, capable of localized intelligence and collaborative decision-making. Leveraging advances in on-device acceleration that targets federated learning, event-driven computing, and edge AI algorithms and models, we can achieve low-latency, privacy-preserving, and energy-efficient intelligence close to data sources. The keynote will highlight the role of intelligent and data-driven dynamicity in edge classification and regression systems,  in enabling real-time learning across heterogeneous devices, paving the way for adaptive, resilient, and autonomous IoT ecosystems.