Ziheng Zhang

His work has three lines, grouped by the kind of AI.

Validated on a live site

On-site quality inspection

AI for perception

He builds mixed reality systems that measure rebar cages and building elements with vision foundation models. The inspecting engineer sees pass or fail results on site and confirms each one.

How it worksPapers (5)

Illustrative photo: Matt Brown, CC BY 2.0

Maintenance and production

AI for decision-making

He builds open benchmarks and human-in-the-loop methods for maintenance dispatching, and deep reinforcement learning for production scheduling in precast module factories. The benchmarks compare learned policies with classical priority rules, and optimality bounds guide training.

How it worksPapers (8)

Maintenance and BIM

AI for generation

CertiGATE checks each work-order schedule from a large language model agent before release. VeriBIM trains a locally deployable language model to edit IFC building models, with a verifier that scores each edit.

How it worksPapers (2)

Looking ahead, his goal is an inspection-to-maintenance system: findings from vehicles, drones, mobile devices or mixed reality headsets update an asset's digital twin across inspection cycles and become maintenance actions that engineers can check.

1. [1]

   Integrating Mixed Reality with Key Digital Technologies for Elevating On-site Construction Efficiency: An Integrative Review and Framework

   Zhang, Z., Ghosh Mondal, T., Yang, X., and Fu, Y.*

   ASCE Journal of Construction Engineering and Management, 152(10), 03126011 (2026). DOI

2. [2]

   Automated Noncontact Geometric Quality Assessment of Building Elements Integrating Mixed Reality, Depth Sensing, and Vision Foundation Model

   Zhang, Z., Ghosh Mondal, T., Yang, X., and Fu, Y.*

   ASCE Journal of Computing in Civil Engineering, 40(1), 04025130 (2026). DOI

3. [3]

   Gaze-Driven Mixed Reality for Dimensional Quality Assessment of Timber Components Using a Few-Shot RGB-D Vision Foundation Model

   Zhang, Z., Yang, X., Fu, Y.*, and Ghosh Mondal, T.

   Advanced Engineering Informatics, 77, 105186 (2027). DOI

4. [4]

   Human-in-the-Loop Rebar Inspection: Integrating Vision Foundation Model with Mixed Reality for Construction Quality Assurance

   Zhang, Z.*, Ghosh Mondal, T., and Yang, X.

   2026 IEEE International Conference on Human-Machine Systems (ICHMS), Singapore, 452–457 (2026). DOI

• Peer review

  Reviewer for six journals since 2024

  • Engineering Structures
  • Engineering Applications of AI
  • Scientific Reports
  • IEEE JSTARS
  • ASCE JCEM
  • Journal of Building Engineering

• Professional membership

  • Member, American Society of Civil Engineers (ASCE)
  • Student Member, IEEE

• Student leadership

  • Vice President (Welfare), NTU Graduate Students' Association, 2023–2024. He led welfare policy and university-level events for more than 5,000 postgraduate students.
  • Member, NTU Board of Discipline, 2023–2024.