Timely and overall knowledge of the states and resource allocation of diverse activities on construction sites is critical to resource leveling, progress tracking, and productivity analysis. Despite its importance, this task is still performed manually. Previous studies have taken a significant step forward in introducing computer vision technologies, although they have been oriented toward limited classes of objects or limited types of activities. Furthermore, they especially focus on single activity recognition, where an image contains only the execution of an activity by one or a few objects. This paper introduces a two-step method for recognizing diverse construction activities in still site images. It detects 22 classes of construction-related objects using convolutional neural networks. With objects detected, semantic relevance representing the likelihood of the cooperation or coexistence between two objects in a construction activity, spatial relevance representing the two-dimensional pixel proximity in the image coordinates, and activity patterns are defined to recognize 17 types of construction activities. The advantage of the proposed method is its potential to recognize diverse concurrent construction activities in a fully automatic way. Therefore, it is possible to save managers’ valuable time in manual data collection and concentrate their attention on solving problems that necessarily demand their expertise.

Automated worker action recognition helps to understand the state of workers’ actions, enabling effective management of work performance in terms of productivity, safety, and health issues. A wristband equipped with an accelerometer (e.g., activity tracker) allows to collect the data related to workers’ hand activities without interfering with their ongoing work. Considering that many construction activities involve unique hand movements, the use of acceleration data from a wristband has great potential for action recognition of construction activities. In this context, the authors examine the feasibility of the wrist-worn accelerometer-embedded activity tracker for automated action recognition. Specifically, masonry work was conducted to collect acceleration data in a laboratory. The classification accuracy of four classifiers—the k-nearest neighbor, multilayer perceptron, decision tree, and multiclass support vector machine—was analyzed with different window sizes to investigate classification performance. It was found that the multiclass support vector machine with a 4-s window size showed the best accuracy (88.1%) to classify four different subtasks of masonry work. The present study makes noteworthy contributions to the current body of knowledge. First, the study allows for automatic construction action recognition using a single wrist-worn sensor without interfering with workers’ ongoing work, which can be widely deployed to construction sites. The use of a single sensor also greatly reduces the burden to carry multiple sensors while also reducing computational cost and memory. Second, influences associated with the variability of movement between subject and experience group were examined; thus, a consideration of data acquisition that reflects the characteristics of workers’ actions is suggested.

Work-related musculoskeletal disorders (WMSDs) are reported to be the most common category of nonfatal occupational injuries that result in days away from work and are also a leading cause of temporary and permanent disability. One of the most effective approaches to preventing WMSDs is to evaluate ergonomics considerations early in the design and construction planning stage before the worker encounters the unsafe conditions. However, a lack of tools for identifying potential ergonomic risks in a proposed workplace design has led to difficulties in integrating safety and health into workplace design practice. In an effort to address this issue, this study explores a motion data-driven framework for ergonomic analysis that automates and visualizes the evaluation process in a virtual workplace. This is accomplished by coupling the ergonomic analysis with three-dimensional (3D) virtual visualization of the work environment. The proposed approach uses motion data from the 3D model of the jobsite to evaluate the risk factors that can produce excessive physical loads on the human body through a biomechanical analysis. A global risk assessment of musculoskeletal disorders is performed on worker motions first, and a biomechanical simulation is then used to further analyze unsafe motions by estimating internal loads on each selected body joint of the worker and redesigning the motion and workplace accordingly. As a case study, several tasks taking place in a construction prefabrication shop are modeled and analyzed to modify the workplace and ensure improved ergonomic safety. The results indicate that the proposed approach enables identification and minimization of awkward worker postures in the virtual model to mitigate ergonomic risk during workplace design. (C) 2015 American Society of Civil Engineers.The authors would like to thank graduate students at the University of Alberta and the staff at Landmark Group of Builders for their help in data collection. The work presented in this paper has been supported financially by the Natural Sciences and Engineering Research Council of Canada (NSERC) Industrial Research Chairs (IRC) program, National Science Foundation (NSF), and Center for Construction Research and Training (CPWR) through the National Institute for Occupational Safety and Health (NIOSH) cooperative agreement. Any opinions, findings, and conclusions or recommendations expressed in this paper are those of the authors and do not necessarily reflect the views of NSERC, NSF, CPWR, or NIOSH