Two Research Engines

RISER Hub Physics Lab x Engineering-AI Lab

RISER Hub Physics Lab anchors resilient intelligent systems for energy and infrastructure. Engineering-AI Lab pushes AI engineering toward open 3D world models, embodied intelligence, and deployable agent systems.

AI for Energy

Resilient intelligence for clean energy systems

Sensing + PHM

Advanced sensing and data analysis for energy system PHM

Machine learning pipelines for prognostics and health management across energy systems, linking data sensing, degradation patterns, and maintenance intelligence.

  • Asset state estimation
  • Failure prediction and maintenance planning
  • Deployable analytics for operating systems
Renewable Grids

Robust operation of renewable-dominated power grids

AI-supported reliability and cybersecurity methods for power grids with high renewable penetration, distributed assets, and changing operating conditions.

  • Cyber-physical grid resilience
  • Extreme-event preparedness
  • Recovery-aware operation
Microgrids + Storage

Microgrids, storage, and carbon-neutral energy planning

Data-driven planning and analysis for microgrids, storage systems, and low-carbon energy infrastructure under uncertain supply, demand, and policy constraints.

  • Storage system assessment
  • Carbon-aware operation
  • Distributed energy coordination
Transport + Assets

Sustainable transportation and lifelong asset management

Reinforcement learning and quantum-inspired learning ideas for sequential maintenance, electrified transportation, adaptive policy design, and long-horizon asset decisions.

  • Policy learning under uncertainty
  • Lifecycle decision-making
  • Energy-transport coupling

AI Engineering

3D World Model research with Engineering-AI Lab

Engineering-AI Lab
Open 3D Worlds

Foundation 3D world model systems

Engineering pipelines for representing, generating, and simulating interactive 3D worlds as reusable AI infrastructure.

  • World state representation
  • Scene-level generation
  • Reusable simulation assets
Representation

3D reconstruction and controllable asset modelling

Methods for turning visual observations into structured 3D assets that can be edited, composed, and used inside intelligent systems.

  • Single and multi-view reconstruction
  • Part-aware representations
  • Editable 3D assets
Spatial Synthesis

Compositional scene understanding and synthesis

AI systems that reason about layout, physical relations, semantics, and temporal consistency across complex 3D scenes.

  • Scene graph reasoning
  • Spatial layout generation
  • Consistent world editing
Embodied Agents

Decision intelligence in 3D environments

Agent systems that connect language, perception, action, feedback, and planning inside dynamic 3D worlds.

  • Language-guided planning
  • Interaction dynamics
  • Action-feedback learning