Intelligent Industrial Risk Assessment using Computer Vision and Hierarchical Fuzzy Logic
DOI:
https://doi.org/10.61467/2007.1558.2027.v18i1.1483Keywords:
Fuzzy logic, YOLOv5, industrial safety, Sugeno inference, Lógica difusa, seguridad industrialAbstract
Industrial safety monitoring using computer vision typically focuses on detecting personal protective equipment (PPE); however, operational risk is gradual, spatial, and context-dependent. This work proposes a hierarchical Sugeno-Takagi fuzzy inference system integrated with YOLOv11, regions of interest (ROI), and homography to estimate continuous risk in a university manufacturing laboratory. The model calculates a protection deficit, hazard exposure, and operational context to generate an instantaneous, exponentially smoothed index on a scale of [0, 100]. The evaluation used a proprietary dataset of 1,611 PPE images, geometric activation of ROIs, and six controlled scenarios with five 90-second videos per scenario. YOLOv11 achieved mAP@0.5 = 0.886. Homography reduced spatial activation errors by up to 93.1% in actual ROI entries and 87.7% in aisle crossings. The system produced low, moderate, high, and critical risk levels consistent with experimental conditions.
Spanish-language metadata / Metadatos en español
Título en español:
Evaluación inteligente del riesgo industrial mediante visión por computador y lógica difusa jerárquica
Resumen:
La monitorización de la seguridad industrial mediante visión por computador suele centrarse en la detección de equipos de protección personal (EPP); sin embargo, el riesgo operacional es gradual, espacial y dependiente del contexto. Este trabajo propone un sistema jerárquico de inferencia difusa Sugeno-Takagi integrado con YOLOv11, regiones de interés (ROI) y homografía para estimar de forma continua el riesgo en un laboratorio universitario de manufactura.
El modelo calcula un déficit de protección, la exposición a peligros y el contexto operacional para generar un índice instantáneo, suavizado exponencialmente, en una escala de [0, 100]. La evaluación utilizó un conjunto de datos propio de 1.611 imágenes de EPP, la activación geométrica de las ROI y seis escenarios controlados, con cinco vídeos de 90 segundos por escenario. YOLOv11 alcanzó un mAP@0,5 = 0,886. La homografía redujo los errores de activación espacial hasta en un 93,1 % en entradas reales a las ROI y en un 87,7 % en cruces de pasillos. El sistema produjo niveles de riesgo bajo, moderado, alto y crítico, coherentes con las condiciones experimentales.
Palabras Claves:
Lógica difusa, ROI, YOLO, seguridad industrial, inferencia de Sugeno, evaluación de riesgos.
Smart citations:
References
Ahmed, T., Hoque, A. S. M., Karmaker, C. L., & Ahmed, S. (2023). Integrated approach for occupational health and safety (OHS) risk assessment: An empirical (case) study in small enterprises. Safety Science, 164, Article 106143. https://doi.org/10.1016/j.ssci.2023.106143
Al-Azani, S., Luqman, H., Alfarraj, M., Sidig, A. A. I., Khan, A. H., & Al-Hamed, D. (2024). Real-time monitoring of personal protective equipment compliance in surveillance cameras. IEEE Access, 12, 121882–121895. https://doi.org/10.1109/ACCESS.2024.3451117
Ayoubi, M., & Arashpour, M. (2026). Enhancing workplace safety through assistive computer vision: Real-time hazard recognition using the Workplace Hazards Dataset (WHD). Computer Vision and Image Understanding, 265, Article 104681. https://doi.org/10.1016/j.cviu.2026.104681
Aziz, A., Suzon, M. M., & Hasan, R. (2025). A fuzzy logic-based risk evaluation and precaution level estimation of explosive, flammable, and toxic chemicals for preventing damages. Heliyon, 11(1), e41216. https://doi.org/10.1016/j.heliyon.2024.e41216
Baraza, X., & Torrent-Sellens, J. (2026). Artificial intelligence and emerging risks in occupational safety and health. Encyclopedia, 6(1), Article 25. https://doi.org/10.3390/encyclopedia6010025
Dodoo, J. E., Al-Samarraie, H., Alzahrani, A. I., Lonsdale, M., & Alalwan, N. (2024). Digital innovations for occupational safety: Empowering workers in hazardous environments. Workplace Health & Safety, 72(3), 84–95. https://doi.org/10.1177/21650799231215811
Gürcanli, G. E., & Müngen, U. (2009). An occupational safety risk analysis method at construction sites using fuzzy sets. International Journal of Industrial Ergonomics, 39(2), 371–387. https://doi.org/10.1016/j.ergon.2008.10.006
Huber, J., Anzengruber-Tanase, B., Schobesberger, M., Haslgrübler, M., Fischer-Schwarz, R., & Ferscha, A. (2025). Evaluating user safety aspects of AI-based systems in industrial occupational safety: A critical review of research literature. International Journal of Environmental Research and Public Health, 22(5), Article 705. https://doi.org/10.3390/ijerph22050705
Iannizzotto, G., Lo Bello, L., & Patti, G. (2021). Personal protection equipment detection system for embedded devices based on DNN and fuzzy logic. Expert Systems with Applications, 184, Article 115447. https://doi.org/10.1016/j.eswa.2021.115447
Ilbahar, E., Karaşan, A., Cebi, S., & Kahraman, C. (2018). A novel approach to risk assessment for occupational health and safety using Pythagorean fuzzy AHP & fuzzy inference system. Safety Science, 103, 124–136. https://doi.org/10.1016/j.ssci.2017.10.025
Kim, H., Kim, K., & Kim, H. (2016). Vision-based object-centric safety assessment using fuzzy inference: Monitoring struck-by accidents with moving objects. Journal of Computing in Civil Engineering, 30(4), Article 04015075. https://doi.org/10.1061/(ASCE)CP.1943-5487.0000562
Li, L., Huang, Z., Wang, J., Du, B., & Dai, L. (2026). Automated construction monitoring based on computer vision: A comprehensive review. Developments in the Built Environment, 25, Article 100832. https://doi.org/10.1016/j.dibe.2025.100832
López, L., Suárez-Ramírez, J., Alemán-Flores, M., & Monzón, N. (2025). Automated PPE compliance monitoring in industrial environments using deep learning-based detection and pose estimation. Automation in Construction, 176, Article 106231. https://doi.org/10.1016/j.autcon.2025.106231
López-Cifuentes, A., Escudero-Viñolo, M., Bescós, J., & Carballeira, P. (2022). Semantic-driven multi-camera pedestrian detection. Knowledge and Information Systems, 64(5), 1211–1237. https://doi.org/10.1007/s10115-022-01673-w
Martinez, C. F., Galvan, F. S., Santos, H. B., Estrada, O. V., De La Cruz, D. I. M., & Ponce del Angel, F. G. (2026). Real-time ergonomic assessment with fuzzy logic and RGB-D sensors. Engineering, Technology & Applied Science Research, 16(1), 31609–31617. https://doi.org/10.48084/etasr.14044
Rasouli, S., Alipouri, Y., & Chamanzad, S. (2024). Smart personal protective equipment (PPE) for construction safety: A literature review. Safety Science, 170, Article 106368. https://doi.org/10.1016/j.ssci.2023.106368
Reglamento (UE) 2024/1689 del Parlamento Europeo y del Consejo, de 13 de junio de 2024, por el que se establecen normas armonizadas en materia de inteligencia artificial y por el que se modifican los Reglamentos (CE) n.º 300/2008, (UE) n.º 167/2013, (UE) n.º 168/2013, (UE) 2018/858, (UE) 2018/1139 y (UE) 2019/2144 y las Directivas 2014/90/UE, (UE) 2016/797 y (UE) 2020/1828 (Reglamento de Inteligencia Artificial). (2024). Diario Oficial de la Unión Europea, L, 2024/1689. EUR-Lex official text
Sadeghi, H., Mohandes, S. R., Hosseini, M. R., Banihashemi, S., Mahdiyar, A., & Abdullah, A. (2020). Developing an ensemble predictive safety risk assessment model: Case of Malaysian construction projects. International Journal of Environmental Research and Public Health, 17(22), Article 8395. https://doi.org/10.3390/ijerph17228395
Salhab, D., Sabek, M., Pourrahimian, E., Gonzalez, V. A., & Hamzeh, F. (2026). Integrated framework of computer vision and fuzzy systems for lean workspace management in construction. Automation in Construction, 181, Article 106618. https://doi.org/10.1016/j.autcon.2025.106618
Secretaría del Trabajo y Previsión Social. (2025, March 28). Norma Oficial Mexicana NOM-017-STPS-2024, Equipo de protección personal-Selección, uso y manejo en los centros de trabajo. Diario Oficial de la Federación. Official DOF record
Secretaría del Trabajo y Previsión Social. (n.d.). Riesgos de trabajo registrados en el IMSS [Interactive data portal]. Retrieved August 29, 2026, from STPS occupational-risk statistics portal
Shin, Y.-S., & Kim, J. (2022). A vision-based collision monitoring system for proximity of construction workers to trucks enhanced by posture-dependent perception and truck bodies’ occupied space. Sustainability, 14(13), Article 7934. https://doi.org/10.3390/su14137934
Takagi, T., & Sugeno, M. (1985). Fuzzy identification of systems and its applications to modeling and control. IEEE Transactions on Systems, Man, and Cybernetics, SMC-15(1), 116–132. https://doi.org/10.1109/TSMC.1985.6313399
Toth, K., Singhal, M., Chauhan, D., Polra, J., Zhou, C., Page, G., & Fisher, C. (2026). A dual-model approach to industrial safety: Computer vision for PPE compliance and hazard-zone monitoring in steel production. Integrating Materials and Manufacturing Innovation, 15, 326–340. https://doi.org/10.1007/s40192-026-00461-6
Tran, S. V.-T., Pham, H. C., Le, Q. T., & Lee, U.-K. (2025). Digital technologies for monitoring hazardous area entry in construction sites. Automation in Construction, 177, Article 106357. https://doi.org/10.1016/j.autcon.2025.106357
Vukicevic, A. M., Petrovic, M., Milosevic, P., Peulic, A., Jovanovic, K., & Novakovic, A. (2024). A systematic review of computer vision-based personal protective equipment compliance in industry practice: Advancements, challenges and future directions. Artificial Intelligence Review, 57, Article 319. https://doi.org/10.1007/s10462-024-10978-x
Zhang, M., Cao, Z., Yang, Z., & Zhao, X. (2020). Utilizing computer vision and fuzzy inference to evaluate level of collision safety for workers and equipment in a dynamic environment. Journal of Construction Engineering and Management, 146(6), Article 04020051. https://doi.org/10.1061/(ASCE)CO.1943-7862.0001802
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 International Journal of Combinatorial Optimization Problems and Informatics
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
