Track 08 of 09
Edge AI, AIoT and Distributed Intelligence
Making autonomy run where the physics is: efficient on-device inference, edge–cloud orchestration, federated and privacy-preserving learning, and sensing infrastructure.
Scope
Making autonomy run where the physics is: efficient on-device inference, edge–cloud orchestration, federated and privacy-preserving learning, and sensing infrastructure.
Research topics
Topics include, but are not limited to:
- Model compression, quantisation, pruning, distillation
- TinyML and MCU-class inference
- Neuromorphic and event-based computing
- Accelerator-aware model design
- Edge–cloud task partitioning and offloading
- Federated and split learning
- On-device continual learning
- Low-power sensing and energy harvesting
- Deterministic and TSN networking
- Edge orchestration and lifecycle management
Emerging challenges
- Energy budget vs accuracy
- Heterogeneous hardware fragmentation
- OTA update and rollback safety
- Statistical heterogeneity in federated settings
- Security of distributed inference
- Observability at the edge
Industry applications
- Smart cameras and inspection
- Wearables and health monitoring
- Agricultural and environmental sensing
- Predictive maintenance sensors
- Retail analytics
- Edge-based drone autonomy
Keywords
Every submission must connect to physical action, embodiment or real-world deployment, use the IEEE conference format (A4, up to 6 pages including references) and include a Deployment Risk and Limitations statement (max. 200 words). Submission guidelines →