HKU Develops AI Platform OpenBIM Doctor to Enhance Construction Management Efficiency
AM730 · 1 SOURCESabout 9 hours ago2 MIN

Summary
The University of Hong Kong's Faculty of Architecture has developed OpenBIM Doctor, an AI-powered information repair platform that combines large language models with openBIM standards and evidence-gated verification mechanisms. The platform helps the construction industry systematically identify, verify, and address potential information issues in building information models. The research recently won the 2026 buildingSMART International openBIM Student Research Group Award, gaining international recognition.
Key Points
- Developed by Dr. Gao Wenjun under the supervision of Professor Lu Weisheng and Professor Chen Junjie at HKU's iLab
- The platform analyzed four construction cases, identifying 4,660 information issues across five categories: information loss, geometry, classification, relationships, and semantic degradation
- OpenBIM Doctor completed 684 repairs across all cases; the largest case processed approximately 1.8 million data records and completed 448 repairs including 437 material-related repairs
- Unlike AI systems that directly modify data, OpenBIM Doctor uses an "Evidence-Gated" mechanism that only writes corrections when suggestions pass verification checks
- In comparative testing involving 60 components, the evidence-gated platform wrote zero unfounded repair suggestions compared to 37 from a baseline model
Why It Matters
Hong Kong's construction industry relies heavily on international supply chains due to limited land and natural resources. Information errors in building information models can significantly impact construction schedules, material procurement, and project costs. This platform provides a more reliable foundation for smart construction by ensuring AI-assisted decisions are based on verifiable evidence rather than speculation .
Hong Kong's construction industry relies heavily on international supply chains due to limited land and natural resources. Information errors in building information models can significantly impact construction schedules, material procurement, and project costs. This platform provides a more reliable foundation for smart construction by ensuring AI-assisted decisions are based on verifiable evidence rather than speculation .