İnşaatta Sıva Aktivitelerinin Rastgele Orman Algoritması ile Sınıflandırılması

Abstract

İnşaat sektöründe çalışanları kontrol etmek ve yönetmek, şantiye yönetiminin etkinliği için çok önemlidir. Geleneksel olarak şantiyelerde bu durum oldukça zordur. Ancak teknolojinin gelişmesiyle birlikte şantiyeler daha etkin bir şekilde kontrol edilebilmektedir. Bu çalışma, rastgele orman (RF) algoritması ile gerçek inşaat ortamındaki sıva işinin faaliyetlerini tahmin etmeyi amaçlamaktadır. Kara sıva inşaat ustasından koluna takılan bir sensör yardımıyla ivmeölçer, jiroskop ve manyetometre verileri toplanmıştır. Model için hazır hale getirilen veriler %80-20 oranında eğitim ve test verisi olarak ikiye ayrılmıştır. Eğitim verileri RF algoritması ile analiz edilmiştir. Modelin tahmini sonucunda elde edilen tahmin değerleri test verileri ile karşılaştırılarak modelin tahmin doğruluğu belirlenmiştir. Analizi sonucunda %88,86 genel tahmin doğruluğu elde edilmiştir.

Keywords

• İnşaat yönetimi
• Sıva aktivitesi
• Aktivite tanıma
• Rastgele orman algoritması
• İnşaat işçisi

Classification of Construction Roughcasting Activities by Random Forest Algorithm

Abstract

Effective monitoring and management of construction-site workers is crucial for optimal site management. While traditionally challenging, modern technological advancements have enabled more efficient site control methods. This study employs a machine learning approach using the Random Forest (RF) algorithm to predict roughcasting work activities in a real construction environment. Data was collected using sensors (accelerometer, gyroscope, and magnetometer) attached to a roughcast master's arm. The methodology involved data preprocessing, including missing data control and standardization, followed by task-based labeling. The data was segmented into windows of 100 data points with 50% overlap, and statistical features were extracted. Using an 80-20% train-test split, the RF model achieved an overall prediction accuracy of 88.86% across approximately 234,000 data points representing various activities: waiting (90%), roughcasting (96%), material preparation (86%), and lining (72%). The study, conducted in a real construction environment, focused specifically on roughcasting activities. This approach, utilizing worker-attached sensors and artificial intelligence, demonstrates potential for broader application across construction activities and represents a step toward technological adaptation in construction site management.

Keywords

• Construction Management
• Roughcasting Activity
• Activity Recognition
• Random Forest Algorithm
• Construction Labor

References

1. Akhavian, R., & Behzadan, A. H. (2016). Smartphone-based construction workers' activity recognition and classification. Automation in Construction, 71, 198–209. https://doi.org/10.1016/j.autcon.2016.08.015
2. Al Jassmi, H., Al Ahmad, M., & Ahmed, S. (2021). Automatic recognition of labor activity: A machine learning approach to capture activity physiological patterns using wearable sensors. Construction Innovation, 21(4), 555–575. https://doi.org/10.1108/CI-02-2020-0018
3. Alemayoh, T. T., Lee, J. H., & Okamoto, S. (2021). New sensor data structuring for deeper feature extraction in human activity recognition. Sensors, 21(8), Article 2814. https://doi.org/10.3390/s21082814
4. Altheimer, J., & Schneider, J. (2024). Smart-watch-based construction worker activity recognition with hand-held power tools. Automation in Construction, 167, Article 105684. https://doi.org/10.1016/j.autcon.2024.105684
5. Ampomah, E. K., Qin, Z., & Nyame, G. (2020). Evaluation of tree-based ensemble machine learning models in predicting stock price direction of movement. Information, 11(6), Article 332. https://doi.org/10.3390/info11060332
6. Antwi-Afari, M. F., Li, H., Umer, W., Yu, Y., & Xing, X. (2020). Construction activity recognition and ergonomic risk assessment using a wearable ınsole pressure system. Journal of Construction Engineering and Management, 146(7), Article 4020077. https://doi.org/10.1061/(ASCE)CO.1943-7862.0001849
7. Antwi-Afari, M. F., Li, H., Seo, J., & Wong, A. Y. L. (2018). Automated detection and classification of construction workers' loss of balance events using wearable insole pressure sensors. Automation in Construction, 96, 189–199. https://doi.org/10.1016/j.autcon.2018.09.010
8. Bangaru, S. S., Wang, C., & Aghazadeh, F. (2022). Automated and continuous fatigue monitoring in construction workers using forearm EMG and IMU wearable sensors and recurrent neural network. Sensors, 22(24), Article 9729. https://doi.org/10.3390/s22249729

9. Bangaru, S. S., Wang, C., Busam, S. A., & Aghazadeh, F. (2021). ANN-based automated scaffold builder activity recognition through wearable EMG and IMU sensors. Automation in Construction, 126, Article 103653. https://doi.org/10.1016/j.autcon.2021.103653
10. Balli, S., Sağbaş, E. A., & Peker, M. (2019). Human activity recognition from smart watch sensor data using a hybrid of principal component analysis and random forest algorithm. Measurement and Control, 52(1-2), 37–45. https://doi.org/10.1177/0020294018813692
11. Batool, S., Khan, M. H., & Farid, M. S. (2024). An ensemble deep learning model for human activity analysis using wearable sensory data. Applied Soft Computing, 159, Article 111599. https://doi.org/10.1016/j.asoc.2024.111599
12. Boyes, H., Hallaq, B., Cunningham, J., & Watson, T. (2018). The industrial internet of things (IIoT): An analysis framework. Computers in Industry, 101, 1–12. https://doi.org/10.1016/j.compind.2018.04.015
13. Forcael, E., Ferrari, I., Opazo-Vega, A., & Pulido-Arcas, J. A. (2020). Construction 4.0: A literature review. Sustainability, 12(22), Article 9755. https://doi.org/10.3390/su12229755
14. Hastie, T., Tibshirani, R., & Friedman, J. (2009). The elements of statistical learning: Data mining, inference, and prediction (2nd ed.). Springer.
15. Jacobsen, E. L., Teizer, J., & Wandahl, S. (2023). Work estimation of construction workers for productivity monitoring using kinematic data and deep learning. Automation in Construction, 152, Article 104932. https://doi.org/10.1016/j.autcon.2023.104932
16. Joshua, L., & Varghese, K. (2011). Accelerometer-based activity recognition in construction. Journal of Computing in Civil Engineering, 25(5), 370–379. https://doi.org/10.1061/(ASCE)CP.1943-5487.0000106
17. Karatas, I., & Budak, A. (2021). Prediction of labor activity recognition in construction with machine learning algorithms. Icontech International Journal, 5(3), 38–47. https://doi.org/10.46291/ICONTECHvol5iss3pp38-47
18. Karatas, I., & Budak, A. (2024a). Development and comparative of a new meta-ensemble machine learning model in predicting construction labor productivity. Engineering, Construction and Architectural Management, 31(3), 1123–1144. https://doi.org/10.1108/ECAM-08-2021-0692
19. Karatas, I., & Budak, A. (2024b). Deep learning-based recognition of construction activities in real construction site environment. Engineering, Construction and Architectural Management, https://doi.org/10.1108/ECAM-08-2024-1036.
20. Kim, K., & Cho, Y. K. (2020). Effective inertial sensor quantity and locations on a body for deep learning-based worker's motion recognition. Automation in Construction, 113, Article 103126. https://doi.org/10.1016/j.autcon.2020.103126
21. Meng, Q., Wang, S., & Zhu, S. (2023). Semi-supervised deep learning for recognizing construction activity types from vibration monitoring data. Automation in Construction, 152, Article 104910. https://doi.org/10.1016/j.autcon.2023.104910
22. Nath, N. D. (2017). Construction ergonomic risk and productivity assessment using mobile technology and machine learning. Master of Science, Missouri State University.
23. Nematallah, H., & Rajan, S. (2024). Adaptive hierarchical classification for human activity recognition using ınertial measurement unit (IMU) time-series data. IEEE Access, 12, 52127–52149. https://doi.org/10.1109/ACCESS.2024.3386351
24. Roberts, D., & Golparvar-Fard, M. (2019). End-to-end vision-based detection, tracking and activity analysis of earthmoving equipment filmed at ground level. Automation in Construction, 105, Article 102811. https://doi.org/10.1016/j.autcon.2019.04.006
25. Roland Berger. (2016). Digitization in the construction industry: Building Europe’s road to Construction 4.0. Roland Berger GmbH.
26. Ryu, J., Seo, J., Jebelli, H., & Lee, S. (2019). Automated action recognition using an accelerometer-embedded wristband-type activity tracker. Journal of Construction Engineering and Management, 145(1), Article 4018114. https://doi.org/10.1061/(ASCE)CO.1943-7862.0001579
27. Ryu, J., Seo, J., Liu, M., Lee, S., & Haas, C. T. (2016). Action recognition using a wristband-type activity tracker: Case study of masonry work. In J.L. Perdomo-Rivera, A. Gonzales-Quevedo, C.L. del Puerto, F. Maldonado-Fortunet, O.I. Molina-Bas (Eds.), Construction Research Congress (pp. 790–799), ASCE. https://doi.org/10.1061/9780784479827.080
28. Sanhudo, L., Calvetti, D., & Martins, J. P. (2021). Activity classification using accelerometers and machine learning for complex construction worker activities. Journal of Building Engineering, 35, Article 102001. https://doi.org/10.1016/j.jobe.2020.102001
29. Taşar, B. (2022). Comparison analysis of machine learning algorithms for steel plate fault detection. Duzce University Journal of Science and Technology, 10(3), 1578–1588. https://doi.org/10.29130/dubited.1058467
30. Valero, E., Sivanathan, A., Bosché, F., & Abdel-Wahab, M. (2017). Analysis of construction trade worker body motions using a wearable and wireless motion sensor network. Automation in Construction, 83, 48–55. https://doi.org/10.1016/j.autcon.2017.08.001
31. Zhang, M., Chen, S., Zhao, X., & Yang, Z. (2018). Research on construction workers' activity recognition based on smartphone. Sensors, 18(8), Article 2667. https://doi.org/10.3390/s18082667