Adaptive Edge Intelligence (Core)

Objective: To realize intelligent systems that operate directly on devices and adapt to constraints such as computation, and communication.

Novelty: Establish design principles that dynamically optimize models, representations, and computation partitioning based on resource and network conditions.

Methods: Model compression, edge-cloud collaboration, federated learning, and uncertainty-aware inference and control.

Results: Prototypes satisfying real-time performance, computational efficiency, and energy constraints, validated via simulation and small-scale demonstrations.

Next: Extend toward context-adaptive AI that integrates communication and application requirements.

Keywords

Edge AI, On-device inference, Model compression, Edge-cloud collaboration, Federated learning, Privacy-aware AI

Objective: To design AI systems that optimize computation and representation based on dynamic network conditions.

Novelty: Jointly consider communication, inference, and control to enable resilient operation under failures, congestion, and variability.

Methods: Distributed optimization, reinforcement learning, graph neural networks (GNN), and service function chain (SFC) placement.

Results: Quantitative evaluation of latency, reliability, and cost under network constraints, along with systematic design of adaptive policies.

Next: Develop unified frameworks that dynamically optimize computation, communication, and information representation.

Keywords

Network-aware AI, Distributed systems, Resilience, QoS/QoE, Distributed optimization, Traffic engineering, SFC placement

Objective: To build practical intelligent systems for real-world environments such as AgeTech and IoT under realistic constraints.

Novelty: Map heterogeneous sensor data (e.g., LiDAR, microwave, pressure sensors) into shared semantic representations robust to missing data, non-IID conditions, and privacy constraints.

Methods: Semantic representation learning, privacy-preserving learning, real-time edge inference, and adaptive operation based on network and energy conditions.

Results: Validation through public datasets, simulation, and small-scale prototyping toward real deployment.

Next: Extend toward resilient systems that remain functional under disruptions such as disasters or network congestion.

Keywords

AgeTech, Assistive technologies, IoT sensing, Activity recognition, Fall detection, Semantic representation, Privacy-preserving learning