TY  - CONF
TI  - Use of Computer Vision and Object Detection to Measure Construction Productivity Using Crew Balance Charts
C1  - Osaka and Kyoto, Japan
C3  - Proceedings of the 33rd Annual Conference of the International Group for Lean Construction (IGLC 33)
SP  - 810
EP  - 821
PY  - 2025
DO  - 10.24928/2025/0184
AU  - Toledo, Mauricio J.
AU  - Lorca, Macarena
AU  - Mora, Miguel
AD  - Assistant Professor, Head of Civil Engineering Department, Universidad Andres Bello, Chile, mauricio.toledo@unab.cl, orcid.org/0000-0002-3903-7260
AD  - Civil Engineer, Universidad Andres Bello, Chile, m.lorcamadriaza@uandresbello.edu
AD  - Associated Professional, Instituto Profesional IACC, Santiago, Chile, mimora@ing.uchile.cl, orcid.org/0009-0003-4285-8670
ED  - Seppänen, Olli 
ED  - Koskela, Lauri 
ED  - Murata , Koichi 
AB  - Experts in the construction industry identify artificial intelligence (AI) technologies as a key strategy for improving productivity. In Chile, construction productivity has stagnated over the past two decades. This study explores the use of computer vision and a machine learning (ML) algorithm to measure productivity reliably, aiming to improve processes and support data-driven decision-making.
This research uses the YOLOv5 algorithm to detect workers' body postures from video and image data. Body postures are categorized as Productive or Contributory Work based on a predefined taxonomy. The algorithm was trained using 1,500 images extracted from 74 360-degree videos captured using a GoPro camera, representing over five hours of slab formwork installation. Experimental results achieved a mean average precision (mAP 0.5) exceeding 85%.
For productivity measurement, fixed-camera recordings captured images at five-second intervals. YOLOv5 detected postures for key tasks, including: installing perimeter taping (IPT), installing plumbed props (IPP), installing supporting beams (ISB), and installing formwork panels (IFP). Results were visualized through Crew Balance Charts, comparing YOLOv5-based and manually constructed analyses. IFP exhibited the best performance results and most of detected images corresponded to Productive Work.
KW  - Computer vision
KW  - productivity
KW  - crew balance chart
KW  - object detection
KW  - AI.
PB  - 
T2  - Proceedings of the 33rd Annual Conference of the International Group for Lean Construction (IGLC 33)
DA  - 2025/06/02
CY  - Osaka and Kyoto, Japan
L1  - http://iglc.net/Papers/Details/2360/pdf
L2  - http://iglc.net/Papers/Details/2360
N1  - Export Date: 02 June 2025
DB  - IGLC.net
DP  - IGLC
LA  - English
ER  -