Automated 2D Blueprint-to-3D Gamified Tactical Training Simulator for Security Forces Using YOLOv8

Abstract

Security forces frequently utilize 2D architectural blueprint diagrams to plan operations within complex indoor environments. Nevertheless, understanding spatial information within a 2D blueprint under time constraints limits operational efficiency. This paper proposes an automated system that converts 2D blueprint diagrams into interactive 3D tactical training environments using a combination of deep learning and procedural generation techniques. The system employs a custom-trained model based on YOLOv8 to detect critical structural elements such as walls, doors, windows, staircases, and balconies within blueprint diagrams. The detected elements are then structured and integrated with a Unity-based engine to create an interactive 3D simulation environment in real time. The system also includes gamified mission scenarios, user interaction mechanisms, and performance analytics such as path efficiency, time efficiency, and task efficiency metrics. The entire system is implemented as an offline pipeline, ensuring efficient execution on conventional computing platforms. Experimental results demonstrate efficient detection performance and generation efficiency, thereby reducing time and cost compared to conventional physical model-based approaches. The proposed system presents a scalable solution for efficient tactical training environments.

Introduction

The text proposes a system for security force training that converts 2D architectural blueprints into interactive 3D tactical simulation environments.

It highlights a major problem in current training methods used by forces like NSG, ATS, RAF, and SWAT: they rely on 2D blueprints or expensive physical mock-ups, which are time-consuming, costly, hard to modify, and require heavy mental effort to visualize in 3D. Existing simulation and gamification tools improve training but are limited because they cannot automatically transform real scanned blueprints into usable 3D environments, especially under real-world constraints like poor image quality, offline use, and low hardware resources.

The proposed solution is a fully offline AI system that:

• Uses a YOLOv8-based model to detect structural elements (walls, doors, windows, etc.) from scanned blueprints (JPG/PNG/PDF)
• Converts detected 2D structures into procedural 3D environments using Unity
• Generates interactive tactical simulations with features like movement, threat zones, and breach scenarios
• Provides performance evaluation (path efficiency, time efficiency, mission analysis)

The system is designed for real-world deployment on regular laptops without internet or expensive equipment.

Conclusion

This paper has presented a system that can automatically convert a 2D architectural blueprint into a gamified 3D tactical simulation in under 5 seconds. A 0.902 mAP50 is achieved by a YOLOv8 model in door detection; a Flask API bridges the ML component and Unity 3D; a modular procedural generation pipeline in C# creates a 3D environment using solid geometry; and a web-based command dashboard serves all evaluator and trainee requirements. This system directly addresses SIH Problem Statement PS1773 (MHA/NSG) and eliminates the cost and time associated with physical mockup construction. Preliminary user testing has shown a significant improvement in path efficiency and door passage rate in three iterations of training, validating the pedagogical effectiveness of game-based rehearsal in security force training. Overall, the proposed system demonstrates a scalable, cost-effective, and real-time solution for modern tactical training using AI-driven simulation.

References

[1] E. S. Sinogina et al., “Potentsial simulyatsionnykh i igrovykh tekhnologiy v obuchenii osnovam bezopasnosti i zashchity Rodiny,” Nauchno-pedagogicheskoe obozrenie, vol. 6, no. 64, pp. 45–55, 2025. [2] A. Zulick et al., “Considerations for implementing simulation-based instructional training for police cadets: insights from practitioners,” Policing: A Journal of Policy and Practice, 2025, doi: 10.1093/police/paae033. [3] A. G. Alshehhi et al., “Exploring the influence of a simulation-based serious game learning approach for developing an understanding of strategic thinking for policing,” Policing: A Journal of Policy and Practice, vol. 18, 2024, doi: 10.1093/police/paae083. [4] K. Kubola et al., “Towards 3D Serious Game Simulation for Military Training,” in 2024 21st International Joint Conference on Computer Science and Software Engineering (JCSSE), 2024, pp. 656–661, doi: 10.1109/JCSSE61278.2024.10613743. [5] E. S. Sinogina et al., “Potentsial simulyatsionnykh i igrovykh tekhnologiy v obuchenii osnovam bezopasnosti i zashchity Rodiny,” Nauchno-pedagogicheskoe obozrenie, vol. 6, no. 64, pp. 45–55, 2025. [6] L. Bojor, “Game-based learning in modern military education: digital wargaming as a tool for tactical training,” International Conference KNOWLEDGE-BASED ORGANIZATION, vol. XXX, no. 2, 2024, doi: 10.2478/kbo-2024-0060. [7] T. Ivanjko et al., “Gamification in Support of Decision Making in Military Higher Education,” in MIPRO 2024, 2024, pp. 352–357, doi: 10.1109/MIPRO60963.2024.10569635. [8] Z. Batzos et al., “Gamification and Serious Games for Cybersecurity Awareness and First Responders Training: An overview,” 2023. [9] G. A. Rakhmat and R. Fauzian, “Design and Build a Holomatrix-Based Tactical Combat Simulation Application,” Journal of Sisfotek Global, vol. 12, no. 1, pp. 38–49, 2022. [10] L. Steven et al., “Empowering Military in Tactical and Warfare Area with Virtual Reality Technology: A Systematic Literature Review,” Procedia Computer Science, vol. 227, pp. 892–901, 2023, doi: 10.1016/j.procs.2023.10.306. [11] A. C. Barreiro et al., “Automatic Reconstruction of Semantic 3D Models from 2D Floor Plans,” arXiv:2306.01642, 2023. [12] S. Feist et al., “Automatic Reconstruction of 3D Models from 2D Drawings: A State-of-the-Art Review,” Eng, vol. 5, no. 2, pp. 784–800, 2024, doi: 10.3390/eng50202042. [13] P. Schuegraf et al., “PLANES4LOD2: Reconstruction of LoD-2 building models using a depth attention-based fully convolutional neural network,” ISPRS Journal of Photogrammetry and Remote Sensing, vol. 211, pp. 425–437, 2024. [14] W. Cai et al., “3D reconstruction of building interiors based on scan-to-BIM and generative design for as-built building,” Engineering, Construction and Architectural Management, 2024, doi: 10.1108/ECAM-07-2024-0980. [15] P. Deshmukh et al., “2D to 3D Floor plan Modeling using Image Processing and Augmented Reality,” in 2023 International Conference on Recent Advances in Electrical, Electronics & Digital Healthcare Technologies (REEDCON), 2023, pp. 682–687, doi: 10.1109/REEDCON57544.2023.10151165. [16] S. Nikoohemat et al., “Point Cloud Based 3D Models for Agent Based Simulations in Social Distancing and Evacuation,” ISPRS Annals, vol. IV-4/W5, pp. 113–120, 2021. [17] G. Yeti?, “Auto-Conversion from 2D Drawing to 3D Model with Deep Learning,” M.Sc. thesis, Middle East Technical University, Sep. 2019. [18] F. Tanaka et al., “Recognition of House Structures from Complicated Electrical Plan Images,” Information, vol. 15, no. 3, 2024, doi: 10.3390/info15030147. [19] N. Fathima et al., “Conversion of 2D Blueprints into 3D Models,” IJIRT, vol. 11, no. 8, pp. 566–571, 2025. [20] P. K. Vinodkumar et al., “Deep Learning for 3D Reconstruction, Augmentation, and Registration: A Review Paper,” Entropy, vol. 26, no. 3, p. 235, 2024, doi: 10.3390/e26030235. [21] Dr. B.J.Job Karuna Sagar, \"Harnessing Machine Vision Algorithms to Direct Car -T Cell Navigation Across Complex Tumor Landscapes in Next Generation Immunotherapy \",Trends In Immunotherapy, volume 09,issue 04,2025 [22] Dr. B.J.Job Karuna Sagar,\"Internet of Things Based Autonomous Robot System Architecture for Home Automation and Health care services\", \"Indonesian Journal of Electrical Engineering and Computer Science\". ISSN: 2502-4752. [23] Dr. B.J.Job Karuna Sagar,\"Cloud computing environment based hierarchical anomaly Intrusion Detection system using artificial neural network\". \"International Journal of Computer Engineering\" (IJECE). ISSN: 2088-8708, DOI: 10.11591/IJECE. V15IL.PP1209-1217, VOL15, NO 1, FEB 2025.1209-1217. [24] Dr. B.J.Job Karuna Sagar,\"Thermodynamic Analysis of a Novel Concentrated Solar Power plant with integrated thermal energy storage\" International journal of thermal Science and engineering progress. TSEP-0-24-02980R1.

Copyright

Copyright © 2026 Mahin Fareeha, Abdul Qadir Syed Quadri, Mohammed Shafi Omair, B. J. Job Karuna Sagar. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.