AI‑Guided Vision‑Based Digital Building Modeling: From Reconstruction to Engineering‑Oriented Model Enrichment

Call for Papers – AI‑Guided Vision‑Based Digital Building Modeling: From Reconstruction to Engineering‑Oriented Model Enrichment

Journal: Journal of Building Engineering
Publisher: Elsevier (ScienceDirect)
Submission window: 10 February 2026 – 31 December 2026

This special issue focuses on AI‑guided, vision‑based digital building modeling for physics‑informed engineering analysis, aiming to bridge computer vision, artificial intelligence, and structural/building‑physics domains. In many practical settings—such as existing‑building assessment, heritage documentation, and post‑disaster evaluation—complete design documentation is missing, and digital models must instead be built from images, point clouds, RGB‑D scans, or mobile‑sensing data. This issue seeks work that goes beyond surface‑level reconstruction to enrich sparse or partial visual data with latent structural, material, and performance‑related attributes needed for engineering decision‑making.

Why this issue matters

• Existing‑building digital models derived purely from observable data often lack interior layouts, concealed structural elements, and material/physics‑based attributes critical for energy, HVAC, structural, damage, and resilience analyses.

• AI‑based inference, data‑generation techniques, and physics‑aware constraints can help estimate and enrich missing features (e.g., wall types, structural systems, load paths, thermal properties) from limited or partial visual information.

• With the rise of digital twins, smart sensing, and extended reality (XR), there is a growing need for robust, validated, and reproducible modeling workflows that link vision‑based data with physics‑based simulations for real‑world engineering practice.

Key topic areas

The SI invites original research articles, review papers, and high‑quality case studies that combine vision‑based data acquisition with AI and physics‑informed engineering modeling. Topics of interest include, but are not limited to:

• AI‑guided modeling of hidden or latent building features

  • Inferring interior layouts, structural systems, and material properties from incomplete or limited‑observability data (images, point clouds, RGB‑D, mobile sensing).

  • Integrating learning‑based methods with domain knowledge and physics‑aware constraints to build engineering‑oriented digital models.

• Semantic enrichment and multi‑level abstraction

  • Semantic labeling and graph‑ or BIM‑like abstractions of building models to support performance‑based analysis and decision‑making.

  • Multi‑resolution representations for energy, HVAC, structural, and resilience simulations.

• Deep analysis and feature extraction from vision‑based data

  • Use of deep learning and feature‑extraction methods on images, point clouds, and RGB‑D data to support physics‑informed modeling.

  • Learning‑based reconstruction and completion of partial or occluded building geometries.

• Data‑driven and multi‑source data fusion

  • Fusion of vision‑based data with other sensor data (e.g., thermal imagery, IoT, SHM sensors) for digital‑twin and performance‑oriented modeling.

  • Workflow designs that connect site‑level sensing with building‑physics and structural‑analysis tools.

• Physics‑informed engineering applications

  • Energy and HVAC performance assessment, structural evaluation, damage and resilience analysis, and performance‑based optimization using AI‑guided digital building models.

  • AI‑enhanced workflows for post‑disaster assessment, retrofit design, and resilience planning.

• Validation, uncertainty, and reproducibility

  • Uncertainty quantification, benchmark datasets, and comparative studies of AI‑guided modeling approaches.

  • Reproducible modeling pipelines and open‑source‑style workflows for engineering decision support.

Guest editors

• Prof. Zhao Xu, Southeast University, Nanjing, China

  • AI‑guided modeling, digital twins, building‑performance simulation, resilience simulation.

• Prof. Yang Zou, Loughborough University, UK

  • Digital twins, AI, extended reality (VR/AR/MR), smart sensing, visualization.

• Assoc. Prof. Ker‑Wei Yeoh, National University of Singapore (NUS), Singapore

  • Applied AI for construction management, BIM, computer‑aided constructability, system dynamics.

• Prof. Decheng Feng, Southeast University, Nanjing, China

  • Damage analysis, performance assessment, uncertainty analysis, machine learning, risk and resilience.

Inquiries about topic fit or suitability can be directed to the Leading Guest Editor, Prof. Zhao Xu:
xuzhao@seu.edu.cn

Submission information

• Submission portal:
  https://www.editorialmanager.com/jbe/default.aspx

• When submitting, select the article type:
  “VSI: JOBE_AI‑Guided Building Modeling”

• Submission period: 10 February 2026 – 31 December 2026

• All manuscripts will be peer‑reviewed based on originality, significance, technical quality, and clarity.

• Accepted papers will be published in a regular issue of Journal of Building Engineering and collected in the online special issue.

For formatting and structural requirements, authors should follow the journal’s Guide for Authors:
https://www.sciencedirect.com/journal/journal-of-building-engineering/publish/guide-for-authors

ServiceSetu Academics — Premier Platform for Academic Opportunities & Research Collaboration

Visit official website of the publisher