Edge Video Analytics Systems (EVOS) 

Developing edge-assisted video intelligence frameworks for real-time perception, adaptive streaming, and interactive analytics under bandwidth/latency constraints. Includes ROI-aware live analytics and mobile video processing pipelines that reduce communication and energy cost while maintaining accuracy.

[Paper 1] [Paper 2] [Paper 3]

Edge Resource Orchestration (ERO) 

Design scheduling, offloading, and deployment strategies that coordinate compute across devices, edge servers, and mobile, UAV edge nodes under heterogeneous resources and dynamic networks. Spans collaborative offloading mechanisms, optimal edge configuration, and UAV-enabled task scheduling and trajectory planning.

[Paper 1] [Paper 2] 

Data-Quality Framework for Multi-source & Multi-modal Data Analytics 

The reliability and safety of AI systems, particularly in autonomous driving, depend fundamentally on the quality of multisource and multimodal data. Advancing from model-centric approaches toward a task-oriented framework that links data quality with system performance and decision accuracy is essential for building trustworthy and efficient intelligent systems.

[Paper 1] [Paper 2] 

Secure Intelligent Systems 

Research in system security investigates integrity verification, authentication protocols, secure data sharing, and resilient distributed architectures across cloud, edge, and federated environments. Emphasis is placed on cryptographic protocol design, formal security guarantees, adversarial robustness, and trust enforcement in large-scale collaborative systems.

[Paper 1] [Paper 2]

Federated Learning (FL) in Connected and Autonomous Vehicles (CAVs) 

Developing distributed, privacy-preserving learning methods that connect sensing, communication, and onboard computation, enabling continuous model improvement under realistic deployment constraints. By emphasizing heterogeneity aware training, bandwidth/energy efficiency, and safety-relevant evaluation to support reliable autonomy at scale.

[Paper 1] [Paper 2] [Paper 3]

Trusted Permissioned Blockchain Systems 

Builds permissioned blockchain solutions for consortium settings requiring fine-grained confidentiality, immutable provenance, and verifiable integrity. Focuses on policy-based access control and smart-contract validation for structured assets, plus governance frameworks that ensure accountability and privacy.

[Paper 1] [Paper 2] [Paper 3]

Smart Health & Wearable Intelligence

Developing sensing-and-analytics methods for continuous health monitoring using wearable and physiological signals for activity understanding, exercise evaluation, and patient-centric applications. Explores AR/MR-assisted interventions and training systems to support therapy and safety education in real-world settings.

[Paper 1] [Paper 2] 

Agentic AI and Multi-Agent Systems

AI agents and agentic systems that decompose complex, real-world problems into coordinated, role-specialized agents operating under structured orchestration. Emphasis is placed on inter-agent communication protocols, memory integration, adaptive task routing, and iterative validation to improve robustness, scalability, and error containment in dynamic environments.

[Paper 1]