Hanning Liu 刘函宁

PhD Student, Tsinghua University
Institute of Future Human Habitats
Shenzhen International Graduate School

My research explores how computational methods can support automated interior design in existing housing to accommodate changing occupant needs. I currently focus on multimodal spatial knowledge representation and knowledge base construction, linking floor plans, interior images, and textual descriptions to support design case retrieval, furniture layout generation, and interior finish material recommendations. I also investigate how collaboration between AI agents and human reviewers can improve the efficiency and quality of knowledge base construction.

I am advised by Peter John Russell. I received my M.Arch. from Tongji University in 2025, advised by Philip F. Yuan, and my B.Arch. from Xi’an University of Architecture and Technology in 2022. My earlier work explored mixed reality in architectural design and assembly.

Portrait of Hanning Liu

News

Current Research

I focus on the furniture configurations and interior finishes that help existing homes adapt to changing living needs. The following projects are ongoing doctoral research.

Axonometric view of a residential interior reconstructed in Rhino

Multimodal knowledge representation for residential interior cases

I am developing a representation that links floor plans, photographs, renderings and text through four levels: rooms, functional zones, furniture groups and furniture instances. A separate branch records wall, floor and ceiling finishes. Source evidence and uncertainty accompany the records.

The aim is to support queries across these levels, such as finding a furniture arrangement for a particular room and tracing it back to its original evidence.

Illustration: an exploratory reconstruction of the TO:DAY interior case.

Workflow for progressive automation of residential interior knowledge base construction through iterative human–AI collaboration and reviewed experience reuse

Progressive Automation of Multimodal Knowledge Base Construction for Residential Interiors

I explore workflows in which agents organize source material and use modeling tools, while people inspect omissions, conflicts and uncertain judgments. Reviewed feedback becomes reusable experience with explicit conditions of applicability.

Current prototypes investigate image interpretation, scene reconstruction and component reuse. Evaluation will compare construction quality and human effort on new cases.

Workflow schematic: multimodal inputs, iterative human–AI collaboration and interlinked outputs. Automation gains remain to be evaluated; the 3D model is a reference illustration. Click the image to view it at full size.

Publications

* Corresponding author. Full details are available in my CV.

Motion capture workflow synchronizing a Rhino and Unity digital model with physical camera pose tracking for AR-assisted assembly

Leveraging Motion Capture System for High Accuracy AR-Assisted Assembly

Hanning Liu, Xingjie Xie, Yujiao Li, Xiaofan Gao, Honglei Wu, Yao Zhang, Philip F. Yuan*

Computational Design and Robotic Fabrication (CDRF), 2024 Best Paper Award

Project video

Shanghai Xuhui West Coast tourist center

Spatial Implementation of Behavioral Performance: Tourist Center of Shanghai Xuhui West Coast

Philip F. Yuan*, Hanning Liu

Chinese and Overseas Architecture, 01, pp. 8–13, 2024

PDF (Chinese)

Photographs of printed concrete components and construction of the experimental prestressed composite bridge

Prestressed 3D Printed Reinforced Concrete Composite Structure Design and Construction: A Case Study of Experimental Bridge Building

J. Wang, Hanning Liu, L. Qian, H. Wu, X. Xie, M. Yuan, Philip F. Yuan*

Computational Design Symposium, Architectural Society of China, 2023 Outstanding Paper Award

Architectural image study using generative design tools

Prompt Writing Approach in GAI Tools Aided Architectural Design: Taking Urban Camp Center Design as an Example

Hanning Liu, Honglei Wu, Xingjie Xie, M. Yuan, Philip F. Yuan*

National Architectural Academy Department of Architectural Digital Technology Teaching and Research Academic Symposium, pp. 433–436, 2023

PDF (Chinese)

Software

Grasshopper plugin for marker-based augmented reality assembly

Fiducial marker-based AR-assisted assembly plugin for Grasshopper

A Grasshopper plugin for pose estimation and AR-assisted assembly. China Software Copyright, 2024.

Food4Rhino Source code

Education

Tsinghua University Sep 2025–present

PhD student, Institute of Future Human Habitats, Shenzhen International Graduate School
Advisor: Peter John Russell

Tongji University Sep 2022–Jun 2025

Master of Architecture, College of Architecture and Urban Planning
Advisor: Philip F. Yuan
GPA: 4.81/5   Ranking: 1st

Thesis: Research on Mixed Reality-Assisted Design and Assembly towards Discrete Free-form Components
University-level Outstanding Master’s Thesis Award

Xi’an University of Architecture and Technology Sep 2017–Jul 2022

Bachelor of Architecture
GPA: 3.93/5   Ranking: 3/192

Teaching

Selected Honors and Awards