An Artificial Intelligence -Based Approach to Calculating A Risk Assessment Matrix Supported by Safety Management in High-Risk Environments

The calculation of the Risk Assessment Matrix (RAM) is fundamental to recognizing diverse possible incidents, managing risks, and promoting safety in various high-risk sectors. However, the traditional methods of calculating a RAM value rely on manual evaluation and expert opinion which is often time-consuming and introduces room for variability. This research sets forth a new methodology to improve risk assessment using Large Language Models (LLMs) to “read” incidents, conduct historical analysis, and compute RAM values. The approach refers to the application of LLMs for the understanding and interpretation of complex multi-structured texts for predicting possible consequences, estimating their effects, and their probabilities using historical data. Concerning computation and LLM assessment of risks, the proposed framework was more efficient, accurate, and reliable than conventional models. It was noted that this method is practically useful for enhancing risk management practices in construction and other high-risk industrial environments.

Risk Assessment Matrix, Risk evaluation, Safety Analysis, Artificial Intelligence, Large Language Models

M. Smetana, L. Salles de Salles, I. Sukharev, and L. Khazanovich, "Highway construction safety analysis using large language models," Applied Sciences, vol. 14, no. 4, p. 1352, 2024.

P. X. Zou, G. Zhang, and J. Wang, "Understanding the key risks in construction projects in China," International Journal of project management, vol. 25, no. 6, pp. 601-614, 2007.

F. Caffaro, M. Roccato, M. Micheletti Cremasco, and E. Cavallo, "Falls from agricultural machinery: risk factors related to work experience, worked hours, and operators’ behavior," Human factors, vol. 60, no. 1, pp. 20-30, 2018.

Y. Sun, D. Fang, S. Wang, M. Dai, and X. Lv, "Safety risk identification and assessment for Beijing Olympic venues construction," Journal of Management in Engineering, vol. 24, no. 1, pp. 40-47, 2008.

N. G. Leveson, Engineering a safer world: Systems thinking applied to safety. The MIT Press, 2016.

B. J. Ale, "Risk assessment practices in The Netherlands," Safety Science, vol. 40, no. 1-4, pp. 105-126, 2002.

M. Salvatore and R. Stefano, "Smart operators: How Industry 4.0 is affecting the worker’s performance in manufacturing contexts," Procedia Computer Science, vol. 180, pp. 958-967, 2021.

H.-H. Kwon and Y.-I. Moon, "Improvement of overtopping risk evaluations using probabilistic concepts for existing dams," Stochastic Environmental Research Risk Assessment, vol. 20, pp. 223-237, 2006.

S. Andersen and B. A. Mostue, "Risk analysis and risk management approaches applied to the petroleum industry and their applicability to IO concepts," Safety Science, vol. 50, no. 10, pp. 2010-2019, 2012.

M. Eskandari, M. K. V. Indukuri, S. M. Lukin, and C. Matuszek, "LLM-Supported Safety Annotation in High-Risk Environments," in HRI 2025 Workshop VAM-HRI.

F. Afzal, S. Yunfei, M. Nazir, and S. M. Bhatti, "A review of artificial intelligence-based risk assessment methods for capturing complexity-risk interdependencies: Cost overrun in construction projects," International Journal of Managing Projects in Business, vol. 14, no. 2, pp. 300-328, 2021.

B. A. Demiss and W. A. Elsaigh, "Application of novel hybrid deep learning architectures combining Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN): construction duration estimates prediction considering preconstruction uncertainties," Engineering Research Express, vol. 6, no. 3, p. 032102, 2024.

P. Fratczak, Y. M. Goh, P. Kinnell, A. Soltoggio, and L. Justham, "Understanding human behavior in industrial human-robot interaction by means of virtual reality," in Proceedings of the Halfway to the Future Symposium 2019, 2019, pp. 1-7.

Ismael, Bahaa Muneer, et al. "Non-dominated sorting genetic algorithm for channel assignment in multiple radio interfaces with multiple channels." AIP Conference Proceedings. Vol. 3393. No. 1. AIP Publishing LLC, 2026.‏

Ismael, Bahaa Muneer, et al. "Multi-Agent Reinforcement Learning for User-Router Assignment in Multi-Radio Multi-Channel Wireless Mesh Networks." International Journal of Intelligent Engineering & Systems 18.8 (2025).‏

Abdullah, Aws Mahmood, Ali Mohsin Kaittan, and Mustafa Sabah Taha. "Evaluation of the stability enhancement of the conventional sliding mode controller using whale optimization algorithm." Indonesian Journal of Electrical Engineering and Computer Science 21.2 (2021): 744-756.‏