Article
The purpose of the research was to develop and scientifically substantiate the architecture of a personalized powerlifting training system based on artificial intelligence and to experimentally prove the effectiveness of its implementation.
Methods and organization of research. The study was conducted at two universities with 48 male powerlifters aged 18-23 years of varying athletic ability, divided into experimental and control groups. The following research methods were used in the study: theoretical literature review, structural and functional modeling, a pedagogical experiment, expert assessment, and mathematical statistics. For intelligent support, the ChatGPT service with access to GPT-5 Plus and computer vision capabilities was used to analyze video recordings of squats, bench presses, and deadlifts and generate reports.
The research results and their discussion. A closed-loop model was developed (data collection – intelligent analysis in ChatGPT – interpretable report and recommendations – training cycle – repeated adjustments), including target, technological, content, and performance blocks. Intelligent reports provided a phased analysis of technique and error triggers (bar trajectory, core and pelvic stability, phasing, tempo, and signs of fatigue) and linked the identified errors to probable causes and corrective measures. Based on the results of the experiment, the experimental group demonstrated a higher increase in the total score for the three events compared to the control group; the differences between the groups were statistically significant (χ², p<0.05). The greatest effect was observed in athletes of the first and second categories, indicating the effectiveness of AI support during the stage of active development of technique and strength.
Conclusion. Integrating AI support into the powerlifting training process improves the effectiveness of training and is supported by the experimental data obtained in the study.
1. Aksyonov, M. O. Management of the training process in powerlifting based on modern information technologies: Dis. ... Cand. Ped. Sciences. 13.00.04 / M. O. Aksyonov. – Ulan-Ude, 2006. – 206 p.
2. Bindusov, E. E. Experience of using digital technologies in educational and training activities of physical culture / E. E. Bindusov // Physical culture and health. – 2022. – No. 3 (83). – P. 58-62.
3. Bogomolov, G. V. Digitalization of the provision of statistical data in the sphere of physical culture and sports / G. V. Bogomolov, S. B. Eroshkina, V. A. Furaev // Theory and practice of physical culture. – 2021. – No. 1. – P. 14-16.
4. Bondarenko, M. P. Modeling the training process of powerlifters taking into account overcoming disrupting factors / M. P. Bondarenko, A. A. Ilchenko, A. S. Malygin, A. S. Semenov // Scientific notes of the P. F. Lesgaft University. – 2022. – No. 1 (203). – P. 34-37.
5. Borisova, A. N., Volnov, S. A. Using artificial intelligence and computer vision technologies to automate the control of fitness exercises using the FORA VISION online platform as an example / A. N. Borisova, S. A. Volnov // Russian journal of information technology in sports. – 2025. – No. 2. – P. 3-14.
6. Lukyanov, A. B. Management of the training process in powerlifting using information technologies / A. B. Lukyanov // Bulletin of VESU. – 2016. – No. 5 (85). – P. 176-181.
7. Nagovitsyn, R. S. Pedagogical potential of GPT chat in the digital transformation of physical education and sports / R. S. Nagovitsyn, R. Sh. Alimov // Scientific notes of the P. F. Lesgaft University. – 2024. – No. 5 (231). – P. 271-274.
8. Panchenko, L. P. Artificial intelligence in providing first aid / L. P. Panchenko // Theory and practice of physical education. – 2021. – No. 5. – P. 85-87.
9. Albert, J. A. Persist: A Multimodal Dataset for the Prediction of Rating of Perceived Exertion and Heart Rate During Resistance Training / J. A. Albert, A. Herdick, C. M. Brahms, U. Granacher, B. Arnrich // Data. – 2022. – 8(1). – 9.
10. Aleksić, J. Computer vision solutions for range of motion assessment / J. Aleksić // Southeastern European Medical Journalю – 2023. – 7(1). – P. 55-66.
11. Bae, K. Concurrent validity and test reliability of the deep learning markerless motion capture system during the overhead squat / K. Bae, S. Lee, S. Y. Bak, H. S. Kim, Y. Ha, J. H. You // Scientific Reports. – 2024. – 14. – Art. 29462.
12. Balsalobre-Fernández, C. Validity of a smartphone app using artificial intelligence for the real time measurement of barbell velocity in the bench press exercise / C. Balsalobre-Fernández, J. Xu, P. Jarvis, S. Thompson, K. Tannion, C. Bishop // Journal of Strength and Conditioning Research. – 2023. – 37(12). – P. e640-e645.
13. Chariar, M. AI Trainer: Autoencoder Based Approach for Squat Analysis and Correction / M. Chariar, S. Rao, A. Irani, S. Suresh, C. S. Asha // IEEE Access. – 2023. – Vol. 11. – P. 107135-107149.
14. Chen, C. Quantitative analysis and evaluation of research on the application of computer vision in sports since the 21st century / C. Chen, J. Xue, W. Gou, M. Xie, X. Yao // Front Sports Act Living. – 2025. – 7. – 604232.
15. Khanal, S. R. A review on computer vision technology for physical exercise monitoring / S. R. Khanal, D. Paulino, J. Sampaio, J. Barroso, A. Reis, V. Filipe // Algorithms. – 2022. – 15(12). – 444.
16. Nagovitsyn, R. S. Artificial Intelligence Program for Predicting Wrestlers’ Sports Performances / R. S. Nagovitsyn, R. A. Valeeva, L. A. Latypova // Sports. – 2023. – 11(10). – 196.
17. Roggio, F. A comprehensive analysis of the machine learning pose estimation models used in human movement and posture analyses: a narrative review / F. Roggio, B. Trovato, M. Sortino, G. Musumeci // Heliyon. – 2024. – 10(21). – e39977.
