An automatic ergonomic evaluation based on human movement recognition from motion capture data with random forest classifiers

Abstract

Aim: This study aims to develop an automated system for classifying industrial operator movements using motion capture (MoCap) data and a random forest (RF) classifier, with the ultimate goal of supporting ergonomic risk assessment and identifying physically demanding tasks suitable for automation in human–robot collaborative environments.

Methods: MoCap data were acquired using the Xsens system from nine participants (four men and five women) performing three movement classes: walking, standing, and bending. A labelled dataset was constructed and used to train an RF classifier via a sliding-window approach with statistical and frequency-domain feature extraction. The model was evaluated on a held-out test set (80/20 split) and through leave-one-subject-out cross-validation. The trained classifier was subsequently applied to a real industrial use case at a logistics company. MoCap data were additionally imported into IPS IMMA to generate a human digital twin and perform frame-by-frame Rapid Entire Body Assessment (REBA) trunk position scoring during identified bending movements.

Results: The RF classifier achieved 98.44% accuracy on the test set, with F1-scores of 0.98, 0.97, and 1.00 for bending, standing, and walking, respectively. Leave-one-subject-out cross-validation consistently exceeded 91% accuracy across all participants. In the industrial use case, bending accounted for 6.7% of the total task time (127.5 s). During bending, the operator spent 45.98% of bending time in REBA trunk position zone 3 (20°–60° flexion) and 28.79% in zone 4 (> 60° flexion).

Conclusions: The integration of MoCap data with RF-based movement classification and automated REBA trunk position extraction provides an objective, reproducible approach to ergonomic risk assessment in industrial settings. The results demonstrate that bending movements expose operators to elevated musculoskeletal risk, supporting the case for targeted automation of these tasks. This work constitutes a first step toward a broader framework for automated ergonomic evaluation and human–robot collaborative workstation design.

Keywords

human motion classification / MoCap / random forest / ergonomic evaluation / REBA

References

[1] Golovianko M, Terziyan V, Branytskyi V, Malyk D. Industry 4.0 vs. Industry 5.0: Co-existence, Transition, or a Hybrid. Procedia Comput Sci. 2023; 217: 102-13.

[2] Adel A. Future of industry 5.0 in society: human-centric solutions, challenges and prospective research areas. J Cloud Comput. 2022; 11: 40.

[3] Villani V, Pini F, Leali F, Secchi C. Survey on human-robot collaboration in industrial settings: Safety, intuitive interfaces and applications. Mechatronics. 2018; 55: 248-66.

[4] Siciliano B, Khatib O. Robotics and the Handbook. In: Siciliano B, Khatib O, editors. Springer Handbook of Robotics. Cham: Springer International Publishing; 2016. pp. 1-6.

[5] Cibrario V, Vargas Gonzalez M, Fantuzzi C, Pia Cavatorta M, Bosani E, Dengel D, et al. Assessing Ergonomics on Cobot for an Optimized Integrated Solution in Early Phase of Product and Process Design. AHFE Int. 2023: 63-9.

[6] Bertoli A, Nini M, Cibrario V, Vargas M, Perona P, Rossi L, et al. Leveraging Digital Twin Technology with a Human-Centered Approach to Automate a Workstation in the Logistics Sector of Made in Italy: CHIMAR Use Case. Machines. 2025; 13: 303.

[7] Vargas M, Cibrario V, Tumiotto D, Bertoli A, Fantuzzi C. Human-Centered optimization through Digital Twins, and Motion Capture Technologies of a manual activity in the logistics sector. AHFE Int. 2026: 22-31.

[8] Bortolini M, Botti L, Galizia FG, Mora C. Ergonomic Design of an Adaptive Automation Assembly System. Machines. 2023; 11: 898.

[9] Mali SC, Vyavahare RT. An Ergonomic Evaluation of an Industrial Workstation: A Review. Int J Curr Eng Technol. 2015; 5: 1820-6.

[10] IPS IMMA [Internet]. flexstructures GmbH; [cited 2026 Mar 10]. Available from: https://flexstructures.com/products/ips-imma/

[11] Sluchak TJ. Ergonomics. AAOHN J. 1992; 40: 105-12.

[12] International Ergonomics Association [Internet]. IEA; [cited 2026 Mar 10]. Available from: https://iea.cc/

[13] Irimie SI, Irimie S. ERGONOMY AND THE ERGONOMIST. HISTORICAL AND CURRENT REFERENCES. Appl Math Mech Eng. 2021; 2: 171-73.

[14] Sabino I, Fernandes MDC, Cepeda C, Quaresma C, Gamboa H, Nunes IL, et al. Application of wearable technology for the ergonomic risk assessment of healthcare professionals: A systematic literature review. Int J Ind Ergon. 2024; 100: 103570.

[15] Berguer R. The Application of Ergonomics to the General Surgeons' Working Environment. Rev Environ Health. 1997; 12: 99-106.

[16] Naeini HS, Karuppiah K, Tamrin SB, Dalal K. Ergonomics in agriculture: An Approach in Prevention of Work-related Musculoskeletal Disorders (WMSDs). J Agric Environ Sci. 2014; 3: 33-51.

[17] Loos MJ, Merino E, Rodriguez CMT. Mapping the state of the art of ergonomics within logistics. Scientometrics. 2016; 109: 85-101.

[18] Butlewski M, Czernecka W, Zembik M. Integrating Motion Capture Technology into Ergonomics Design. Teh glas. 2025; 19: 49-54.

[19] Hignett S, McAtamney L. Rapid Entire Body Assessment (REBA). Appl Ergon. 2000; 31: 201-5.

[20] Bortolini M, Gamberi M, Pilati F, Regattieri A. Automatic assessment of the ergonomic risk for manual manufacturing and assembly activities through optical motion capture technology. Procedia CIRP. 2018; 72: 81-6.

[21] Menolotto M, Komaris DS, Tedesco S, O'Flynn B, Walsh M. Motion Capture Technology in Industrial Applications: A Systematic Review. Sensors. 2020; 20: 5687.

[22] Berti N, Finco S, Guidolin M, Battini D. Towards Human Digital Twins to enhance workers’ safety and production system resilience. IFAC-Pap. 2023; 56: 11062-7.

[23] Salisu S, Ruhaiyem NIR, Eisa TAE, Nasser M, Saeed F, Younis HA. Motion Capture Technologies for Ergonomics: A Systematic Literature Review. Diagnostics. 2023; 13: 2593.

[24] Rybnikár F, Kačerová I, Hořejší P, Šimon M. Ergonomics Evaluation Using Motion Capture Technology-Literature Review. Appl Sci. 2022; 13: 162.

[25] Petrosyan T, Dunoyan A, Mkrtchyan H. Application Of Motion Capture Systems In Ergonomic Analysis. Armen J Spec Educ. 2020; 4: 107-17.

[26] Bortolini M, Faccio M, Gamberi M, Pilati F. Motion Analysis System (MAS) for production and ergonomics assessment in the manufacturing processes. Comput Ind Eng. 2020; 139: 105485.

[27] Battini D, Berti N, Finco S, Guidolin M, Reggiani M, Tagliapietra L. WEM-Platform: A real-time platform for full-body ergonomic assessment and feedback in manufacturing and logistics systems. Comput Ind Eng. 2022; 164: 107881.

[28] Berti N, Finco S, Guidolin M, Reggiani M, Battini D. Real-time postural training effects on single and multi-person ergonomic risk scores. IFAC-Pap. 2022; 55: 163-8.

[29] Keshvarparast A, Berti N, Chand S, Guidolin M, Lu Y, Battaia O, et al. Ergonomic design of Human-Robot collaborative workstation in the Era of Industry 5.0. Comput Ind Eng. 2024; 198: 110729.

[30] Kačerová I, Kubr J, Hořejší P, Kleinová J. Ergonomic Design of a Workplace Using Virtual Reality and a Motion Capture Suit. Appl Sci. 2022; 12: 2150.

[31] Brosche J, Wackerle H, Augat P, Lödding H. Individualized workplace ergonomics using motion capture. Appl Ergon. 2024; 114: 104140.

[32] Ceylan D, Li W, Mitra NJ, Agrawala M, Pauly M. Designing and fabricating mechanical automata from mocap sequences. ACM Trans Graph. 2013; 32: 1-11.

[33] Simonetto M, Arena S, Peron M. A methodological framework to integrate motion capture system and virtual reality for assembly system 4.0 workplace design. Saf Sci. 2022; 146: 105561.

[34] Vantigodi S, Venkatesh Babu R. Real-time human action recognition from motion capture data. In: Proceedings of 2013 Fourth National Conference on Computer Vision, Pattern Recognition, Image Processing and Graphics (NCVPRIPG); 2013 Dec 18-21; Jodhpur, India. IEEE; 2013. pp. 1-4.

[35] Ijjina EP, Mohan CK. Human Action Recognition Based on MOCAP Information Using Convolution Neural Networks. In: Proceedings of 13th International Conference on Machine Learning and Applications; 2014 Dec 3-6; Detroit, USA. IEEE; 2014. pp. 159-64.

[36] Cho K, Chen X. Classifying and Visualizing Motion Capture Sequences using Deep Neural Networks. In: VISAPP 2014 - Proceedings of the 9th International Conference on Computer Vision Theory and Applications; 2014 Jun; Lisbon, Portugal. SciTePress; 2014. pp. 1-9.

[37] Cai C, Yang C, Lu S, Gao G, Na J. Human motion pattern recognition based on the fused random forest algorithm. Measurement. 2023; 222: 113540.

[38] Iriondo Pascual A, Eklund M, Högberg D. Towards Automated Hand Force Predictions: Use of Random Forest to Classify Hand Postures. In: Proceedings of the 22nd Congress of the International Ergonomics Association; 2025 Oct 25-29. Singapore: Springer; 2025. pp. 201-6.

[39] Jaén-Vargas M, Reyes Leiva K, Fernandes F, Gonçalves S, Tavares Silva M, Lopes D, et al. A Deep Learning Approach to Recognize Human Activity Using Inertial Sensors and Motion Capture Systems. In: Frontiers in Artificial Intelligence and Applications. IOS Press BV; 2021. pp. 250-6.

[40] Igelmo V, Syberfeldt A, Högberg D, García Rivera F, Pérez Luque E. Aiding Observational Ergonomic Evaluation Methods Using MOCAP Systems Supported by AI-Based Posture Recognition. In: Advances in Transdisciplinary Engineering. IOS Press BV; 2020. pp. 419-29.

[41] Barnachon M, Bouakaz S, Boufama B, Guillou E. Ongoing human action recognition with motion capture. Pattern Recognit. 2014; 47: 238-47.

[42] Kadu H, Kuo CJ. Automatic Human Mocap Data Classification. IEEE Trans Multimed. 2014; 16: 2191-202.

[43] Li L, Martin T, Xu X. A novel vision-based real-time method for evaluating postural risk factors associated with musculoskeletal disorders. Appl Ergon. 2020; 87: 103138.

[44] Wang J, Chen D, Zhu M, Sun Y. Risk assessment for musculoskeletal disorders based on the characteristics of work posture. Autom Constr. 2021; 131: 103921.

[45] Han X, Nishida N, Morita M, Sakai T, Jiang Z. Compensation Method for Missing and Misidentified Skeletons in Nursing Care Action Assessment by Improving Spatial Temporal Graph Convolutional Networks. Bioengineering. 2024; 11: 127.

[46] Rodrigues PB, Xiao Y, Fukumura YE, Awada M, Aryal A, Becerik-Gerber B, et al. Ergonomic assessment of office worker postures using 3D automated joint angle assessment. Adv Eng Inform. 2022; 52: 101596.

[47] Jiang X, Hu Z, Wang S, Zhang Y. A Survey on Artificial Intelligence in Posture Recognition. Comput Model Eng Sci. 2023; 137: 35-82.

[48] Qureshi TS, Shahid MH, Farhan AA, Alamri S. A systematic literature review on human activity recognition using smart devices: advances, challenges, and future directions. Artif Intell Rev. 2025; 58: e58.

[49] Gorce P, Jacquier-Bret J. Current Trends in Artificial Intelligence for Recognizing Work Postures to Prevent Work-Related Musculoskeletal Disorders: Systematic Review and Meta-Analysis by Occupational Activity. Bioengineering. 2026; 13: 298.

[50] CHIMAR [Internet]. Chimar Spa; [cited 2026 Mar 10]. Available from: https://chimar.eu/

[51] Salman HA, Kalakech A, Steiti A. Random Forest Algorithm Overview. Babylon J Mach Learn. 2024; 2024: 69-79.

[52] Navada A, Ansari AN, Patil S, Sonkamble BA. Overview of Use of Decision Tree algorithms in Machine Learning. In: IEEE Control and System Graduate Research Colloquium; 2011 Jun 27-28; Pune, India. IEEE; 2011. pp. 37-42.

[53] Nurwulan NR, Selamaj G. Random Forest for Human Daily Activity Recognition. J Phys: Conf Ser. 2020; 1655: 012087.

[54] Cutler A, Cutler DR, Stevens JR. Random Forests. In: Ensemble Machine Learning. New York, NY: Springer New York; 2011. pp. 157-75.

[55] Bargellesi N, Carletti M, Cenedese A, Susto GA, Terzi M. A Random Forest-based Approach for Hand Gesture Recognition with Wireless Wearable Motion Capture Sensors. IFAC-Pap. 2019; 52: 128-33.

[56] Xsens [Internet]. Xsens Technologies; c2026 [cited 2026 Mar 10]. Available from: https://www.xsens.com/movella

[57] Schepers M, Giuberti M, Bellusci G. Xsens MVN: Consistent Tracking of Human Motion Using Inertial Sensing. Xsens; 2018.

[58] Guidolin M, Budau Petrea RA, Roberto O, Reggiani M, Menegatti E, Tagliapietra L. On the Accuracy of IMUs for Human Motion Tracking: a Comparative Evaluation. In: Proceedings of 2021 IEEE International Conference on Mechatronics (ICM); 2021 Mar 7-9; Kashiwa, Japan. IEEE; 2021. pp. 1-6.

[59] scikit-learn: Machine Learning in Python [Internet]. scikit-learn; [cited 2026 Mar 10]. Available from: https://scikit-learn.org/stable/

[60] Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, et al. Scikit-learn: Machine Learning in Python. J Mach Learn Res. 2011; 12: 2825-30.

[61] Testas A. Random Forest Classification with Scikit-Learn and PySpark. In: Distributed Machine Learning with PySpark. Berkeley: Apress; 2023. pp. 243-58.

[62] Smit-Russcher R, Princelle D, Bilal I. MVN white paper: Enhancing motion tracking accuracy with novel gender-specific models. Xsens; 2025.

[63] Brolinab E, Högberga D, Hansonbc L. Design of a Digital Human Modelling Module for Consideration of Anthropometric Diversity. AHFE Int. 2020.