Article Open Access

Deep Learning-Based Worker Posture Classification for Ergonomic Risk Evaluation in Manufacturing

Rahmadwati Rahmadwati, Farrel Rafif Ferdian, Yeni Sumantri, Dhaffin Rayhzan Ferdian

Abstract


Continuous ergonomic monitoring in manufacturing remains challenging because conventional posture assessment methods rely on manual observation, making evaluations time-consuming, subjective, and unsuitable for continuous industrial applications. This study proposes an automated ergonomic risk assessment framework that integrates Media Pipe Pose with a Convolutional Neural Network (CNN) to classify worker postures into low-, medium-, and high-risk ergonomic categories. The framework extracts 33 anatomical body landmarks from RGB images and video frames to generate marker less posture representations for deep learning-based classification. A dataset consisting of 4,500 posture samples collected from assembly, packaging, and welding workstations was expanded to 12,000 samples through data augmentation techniques, including rotation, scaling, horizontal flipping, and brightness adjustment, to improve model robustness and generalization. The CNN model was trained and evaluated using an independent test dataset, achieving an overall classification accuracy of 94.2%, with precision, recall, and F1-score consistently exceeding 94% across all ergonomic risk categories. Comparative evaluation against a conventional REBA/RULA-based rule-driven assessment demonstrated that the proposed framework improved classification accuracy by 7.5 percentage points while eliminating the need for manual posture scoring and reducing observer subjectivity. Furthermore, computational performance analysis showed that the complete inference pipeline operated at an average of 14 ms per frame (approximately 28.5 FPS) on a standard Intel Core i7 CPU with 16 GB RAM, without requiring GPU acceleration, indicating its suitability for real-time deployment in manufacturing environments. The proposed Media Pipe–CNN framework provides an efficient, accurate, and marker less solution for automated ergonomic risk assessment, supporting intelligent occupational safety management, continuous workplace monitoring, and the implementation of smart manufacturing systems aligned with Industry 4.0 initiatives

Keywords


Deep Learning, Convolutional Neural Network, Media pipe Pose, Posture Classification, Ergonomic Risk Assessment

References


Park, J., Kim, Y., & Han, B. (2018). Work sectors with high risk for work-related musculoskeletal disorders in Korean men and women. Safety and Health at Work, 9(1), 75–78. https://doi.org/10.1016/j.shaw.2017.06.005

Bao, S., Howard, N., & Lin, J. H. (2020). Are work-related musculoskeletal disorders claims related to risk factors in workplaces of the manufacturing industry? Annals of Work Exposures and Health, 64(2), 152–164. https://doi.org/10.1093/annweh/wxz084

Fritzsche, L., Wegge, J., Schmauder, M., Kliegel, M., & Schmidt, K. H. (2014). Good ergonomics and team diversity reduce absenteeism and errors in car manufacturing. Ergonomics, 57(2), 148–161. https://doi.org/10.1080/00140139.2013.875597

Hembecker, P. K., Reis, D. C., Konrath, A. C., Gontijo, L. A., & Merino, E. A. (2017). Investigation of musculoskeletal symptoms in a manufacturing company in Brazil: A cross-sectional study. Brazilian Journal of Physical Therapy, 21(3), 175–183. https://doi.org/10.1016/j.bjpt.2017.03.014

Punnett, L. (2014). Musculoskeletal disorders and occupational exposures: How should we judge the evidence concerning the causal association? Scandinavian Journal of Public Health, 42(13_suppl), 49–58. https://doi.org/10.1177/1403494813517324

Dong, Y., Jiang, P., Jin, X., Jiang, N., Huang, W., Peng, Y., Shen, Y., He, L., Forsman, M., & Yang, L. (2022). Association between long-term static postures exposure and musculoskeletal disorders among university employees: A viewpoint of inflammatory pathways. Frontiers in Public Health, 10, 1055374. https://doi.org/10.3389/fpubh.2022.1055374

Caicoya, M., & Delclos, G. L. (2010). Work demands and musculoskeletal disorders from the Spanish National Survey. Occupational Medicine, 60(6), 447–450. https://doi.org/10.1093/occmed/kqp191

Charles, L. E., Ma, C. C., Burchfiel, C. M., & Dong, R. G. (2018). Vibration and ergonomic exposures associated with musculoskeletal disorders of the shoulder and neck. Safety and Health at Work, 9(2), 125–132. https://doi.org/10.1016/j.shaw.2017.10.003

Kee, D. (2021). Comparison of OWAS, RULA and REBA for assessing potential work-related musculoskeletal disorders. International Journal of Industrial Ergonomics, 83, 103140. https://doi.org/10.1016/j.ergon.2021.103140

Kee, D. (2022). Systematic comparison of OWAS, RULA, and REBA based on a literature review. International Journal of Environmental Research and Public Health, 19(1), 595. https://doi.org/10.3390/ijerph19010595

Lin, P.-C., Chen, Y.-J., Chen, W.-S., & Lee, Y.-J. (2022). Automatic real-time occupational posture evaluation and select corresponding ergonomic assessments. Scientific Reports, 12, 2139. https://doi.org/10.1038/s41598-022-05812-9

Lee, Y.-C., & Lee, C.-H. (2022). SEE: A proactive strategy-centric and deep learning-based ergonomic risk assessment system for risky posture recognition. Advanced Engineering Informatics, 53, 101717. https://doi.org/10.1016/j.aei.2022.101717

Lan, G., Wu, Y., Hu, F., & Hao, Q. (2022). Vision-based human pose estimation via deep learning: A survey. IEEE Transactions on Human-Machine Systems, 53(1), 253–268. https://doi.org/10.1109/THMS.2022.3219242

Kulikajevas, A., Maskeliunas, R., & Damaševi?ius, R. (2021). Detection of sitting posture using hierarchical image composition and deep learning. PeerJ Computer Science, 7, e442. https://doi.org/10.7717/peerj-cs.442

Roggio, F., Trovato, B., Sortino, M., & Musumeci, G. (2024). A comprehensive analysis of the machine learning pose estimation models used in human movement and posture analyses: A narrative review. Heliyon, 10(21), e39977. https://doi.org/10.1016/j.heliyon.2024.e39977

Deshpande, U. U., Araujo, S. D. C. S., Deshpande, S., Kangralkar, V., Patil, R., Chate, R. A. A., et al. (2025). A review of machine learning techniques for ergonomic risk assessment based on human pose estimation. Discover Artificial Intelligence, 5(1), 287. https://doi.org/10.1007/s44163-025-00566-5

Swain, D., Satapathy, S., Acharya, B., Shukla, M., Gerogiannis, V. C., Kanavos, A., & Giakovis, D. (2022). Deep Learning Models for Yoga Pose Monitoring. Algorithms, 15(11), 403. https://doi.org/10.3390/a15110403

Malik, N. u. R., Abu-Bakar, S. A. R., Sheikh, U. U., Channa, A., & Popescu, N. (2023). Cascading Pose Features with CNN-LSTM for Multiview Human Action Recognition. Signals, 4(1), 40–55. https://doi.org/10.3390/signals4010002

Yu, Y., Li, H., Yang, X., Kong, L., Luo, X., & Wong, A. Y. (2019). An automatic and non-invasive physical fatigue assessment method for construction workers. Automation in Construction, 103, 1–12. https://doi.org/10.1016/j.autcon.2019.02.020

Seo, J., & Lee, S. (2021). Automated postural ergonomic risk assessment using vision-based posture classification. Automation in Construction, 128, 103725. https://doi.org/10.1016/j.autcon.2021.103725

Li, L., & Xu, X. (2019). A deep learning-based RULA method for working posture assessment. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 63(1), 1095–1099. https://doi.org/10.1177/1071181319631174

Tang, Q., Chen, W., Fang, Y., et al. (2021). Development of a fully automated RULA assessment system based on computer vision. International Journal of Industrial Ergonomics, 86, 103218. https://doi.org/10.1016/j.ergon.2021.103218

Plantard, P., Shum, H. P. H., Le Pierres, A. S., & Multon, F. (2017). Validation of an ergonomic assessment method using Kinect data in real workplace conditions. Applied Ergonomics, 65, 136–144. https://doi.org/10.1016/j.apergo.2017.06.004

Paudel, P., Kwon, Y. J., Kim, D. H., & Choi, K. H. (2022). Industrial Ergonomics Risk Analysis Based on 3D-Human Pose Estimation. Electronics, 11(20), 3403. https://doi.org/10.3390/electronics11203403

S. V., S. B., Parasuraman, T., Akila, S., Ramakrishnan, R., Shivaraju, K., Krishnakumar, S., Sreedev, S. A., & Vijayalakshmi, C. (2024). Yoga poses identification and classification based on convolutional neural network and transfer learning with MediaPipe. International Research Journal of Multidisciplinary Technovation, 6(2), 156–167. https://doi.org/10.54392/irjmt24212

Lemley, J., Bazrafkan, S., & Corcoran, P. (2017). Smart augmentation learning an optimal data augmentation strategy. IEEE Access, 5, 5858–5869. https://doi.org/10.1109/ACCESS.2017.2696121

Mai, J., Gao, C., & Bao, J. (2025). Domain Generalization Through Data Augmentation: A Survey of Methods, Applications, and Challenges. Mathematics, 13(5), 824. https://doi.org/10.3390/math13050824

Zhang, H., Yan, X., & Li, H. (2018). Ergonomic posture recognition using 3D view-invariant features from single ordinary camera. Automation in Construction, 94, 1–10. https://doi.org/10.1016/j.autcon.2018.05.033

Parsa, B., Samani, E. U., Hendrix, R., Devine, C., Singh, S. M., Devasia, S., & Banerjee, A. G. (2019). Toward ergonomic risk prediction via segmentation of indoor object manipulation actions using spatiotemporal convolutional networks. IEEE Robotics and Automation Letters, 4(4), 3153–3160. https://doi.org/10.1109/LRA.2019.2925305

Ding, Z., Li, W., Ogunbona, P., & Qin, L. (2019). A real-time webcam-based method for assessing upper-body postures. Machine Vision and Applications, 30(5), 833–850. https://doi.org/10.1007/s00138-019-01033-9

Dong, H., Liu, R., & Tham, A. W. (2024). Accuracy comparison between five machine learning algorithms for financial risk evaluation. Journal of Risk and Financial Management, 17(2), 50. https://doi.org/10.3390/jrfm17020050

Cabot, J. H., & Ross, E. G. (2023). Evaluating prediction model performance. Surgery, 174(3), 723–726. https://doi.org/10.1016/j.surg.2023.05.023

Mathis, A., Mamidanna, P., Cury, K. M., Abe, T., Murthy, V. N., Mathis, M. W., & Bethge, M. (2018). DeepLabCut: markerless pose estimation of user-defined body parts with deep learning. Nature Neuroscience, 21(9), 1281–1289. https://doi.org/10.1038/s41593-018-0209-y

Xin, C., Kim, S., Cho, Y., & Park, K. S. (2024). Enhancing Human Action Recognition with 3D Skeleton Data: A Comprehensive Study of Deep Learning and Data Augmentation. Electronics, 13(4), 747. https://doi.org/10.3390/electronics13040747

Zeng, W. (2024). Image data augmentation techniques based on deep learning: A survey. Mathematical Biosciences and Engineering, 21(6), 6190–6224. https://doi.org/10.3934/mbe.2024272

Micheletti Cremasco, M., Giustetto, A., Caffaro, F., Colantoni, A., Cavallo, E., & Grigolato, S. (2019). Risk Assessment for Musculoskeletal Disorders in Forestry: A Comparison between RULA and REBA in the Manual Feeding of a Wood-Chipper. International Journal of Environmental Research and Public Health, 16(5), 793. https://doi.org/10.3390/ijerph16050793

Wibowo, A. H., & Mawadati, A. (2021). The Analysis of Employees’ Work Posture by using Rapid Entire Body Assessment (REBA) and Rapid Upper Limb Assessment (RULA). IOP Conference Series: Earth and Environmental Science, 704(1), 012022. https://doi.org/10.1088/1755-1315/704/1/012022

Zhu, J., Li, Y., Lewicke, J., & Li, X. (2026). Improvements for adjustment factors of REBA and RULA based on muscle activation analysis of body segmental movement. International Journal of Industrial Ergonomics, 114, 103944. https://doi.org/10.1016/j.ergon.2026.103944

Zhang, T., Lian, J., Wen, J., & Chen, C. P. (2023). Multi-person pose estimation in the wild: Using adversarial method to train a top-down pose estimation network. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 53(7), 3919–3929. https://doi.org/10.1109/TSMC.2023.3234611

Shao, H., Zhang, X., & Li, L. (2025). Advancing human pose estimation through interdisciplinary physics-inspired deep learning models. Frontiers in Physics, 13, 1558325. https://doi.org/10.3389/fphy.2025.1558325

Ray, E. L., Sasaki, J. E., Freedson, P. S., & Staudenmayer, J. (2018). Physical activity classification with dynamic discriminative methods. Biometrics, 74(4), 1502–1511. https://doi.org/10.1111/biom.12892

Hosseinian, S. M., Zhu, Y., Mehta, R. K., Erraguntla, M., & Lawley, M. A. (2019). Static and dynamic work activity classification from a single accelerometer: Implications for ergonomic assessment of manual handling tasks. IISE Transactions on Occupational Ergonomics and Human Factors, 7(1), 59–68. https://doi.org/10.1080/24725838.2019.1608873

Shi, W., Cao, J., Zhang, Q., Li, Y., & Xu, L. (2016). Edge computing: Vision and challenges. IEEE Internet of Things Journal, 3(5), 637–646. https://doi.org/10.1109/JIOT.2016.2579198

Cao, K., Liu, Y., Meng, G., & Sun, Q. (2020). An overview on edge computing research. IEEE Access, 8, 85714–85728. https://doi.org/10.1109/ACCESS.2020.2991734

Greco, A., Caterino, M., Fera, M., & Gerbino, S. (2020). Digital Twin for Monitoring Ergonomics during Manufacturing Production. Applied Sciences, 10(21), 7758. https://doi.org/10.3390/app10217758

Nioata, A., ??pirdea, A., Chivu, O. R., Feier, A., Enache, I. C., Gheorghe, M., & Borda, C. (2025). Workplace Safety in Industry 4.0 and Beyond: A Case Study on Risk Reduction Through Smart Manufacturing Systems in the Automotive Sector. Safety, 11(2), 50. https://doi.org/10.3390/safety11020050




DOI: https://doi.org/10.52088/ijesty.v6i3.1878

Refbacks

  • There are currently no refbacks.


Copyright (c) 2026 Rahmadwati, Farrel Rafif Ferdian, Yeni Sumantri, Dhaffin Rayhzan Ferdian

International Journal of Engineering, Science, and Information Technology (IJESTY) eISSN 2775-2674