Learning and Simulating Building Evacuation Patterns for Enhanced Safety Design Using Generative Models

2025-10-22

AI Summary

Traditional building evacuation simulation relies on high-parameter, fine-grained agent-based modeling, hindering rapid iteration in early design stages. To address this, we propose DiffEvac—a conditional diffusion model leveraging decoupled feature representation and image-based prompting. Trained on a novel dataset comprising 399 functional layouts with corresponding evacuation heatmaps, DiffEvac enables physics-informed, low-threshold generative simulation. It encodes physical attributes—such as architectural layout and pedestrian density—directly into image prompts, eliminating the need for complex agent modeling. Experiments demonstrate that DiffEvac outperforms Conditional GANs by +37.6% in SSIM and +142% in PSNR, while reducing per-simulation time to just two minutes—a 16× speedup. This efficiency enables scalable what-if analysis and facilitates concurrent multi-objective safety optimization in architectural design.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Sampling/Simulation-based Search
• Cognitive Modeling & Cognitive Systems: Agent Architectures

Application Category

• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Evacuation simulation is essential for building safety design, ensuring properly planned evacuation routes. However, traditional evacuation simulation relies heavily on refined modeling with extensive parameters, making it challenging to adopt such methods in a rapid iteration process in early design stages. Thus, this study proposes DiffEvac, a novel method to learn building evacuation patterns based on Generative Models (GMs), for efficient evacuation simulation and enhanced safety design. Initially, a dataset of 399 diverse functional layouts and corresponding evacuation heatmaps of buildings was established. Then, a decoupled feature representation is proposed to embed physical features like layouts and occupant density for GMs. Finally, a diffusion model based on image prompts is proposed to learn evacuation patterns from simulated evacuation heatmaps. Compared to existing research using Conditional GANs with RGB representation, DiffEvac achieves up to a 37.6% improvement in SSIM, 142% in PSNR, and delivers results 16 times faster, thereby cutting simulation time to 2 minutes. Case studies further demonstrate that the proposed method not only significantly enhances the rapid design iteration and adjustment process with efficient evacuation simulation but also offers new insights and technical pathways for future safety optimization in intelligent building design. The research implication is that the approach lowers the modeling burden, enables large-scale what-if exploration, and facilitates coupling with multi-objective design tools.

Problem

• Simulating building evacuation efficiently for rapid safety design iteration
• Learning evacuation patterns from layouts using generative diffusion models
• Reducing simulation time from hours to minutes while improving accuracy

Innovation

• Uses generative models to learn evacuation patterns
• Proposes diffusion model with image prompts
• Achieves faster simulation with improved accuracy