2017 Volume 82 Issue 739 Pages 2361-2371
Since many workers are involved in construction work, it is important to improve work efficiency. For this purpose, time duration and motion studies are often applied during construction work. The authors have developed methods for time and motion studies by using factor analysis. Moreover, in this paper, the authors describe analysis methods to improve work efficiency via factor analysis. The methods can be decomposed into the following seven steps:
1. Recording the tasks and activities of workers by using a camcorder
2. Observing and defining each task and activity of the workers
3. Evaluating each task and activity
4. Analyzing the task and activity times (time study)
5. Predicting the latent factors affecting task and activity times
6. Verifying the latent factor predictions via confirmatory factor analysis (CFA)
7. Analyzing the cause of work delay by using factor scores from CFA as the task and activity times
To illustrate time and motion studies by using factor analysis, we record the tasks and activities of workers. Fig. 2 shows an example of ceiling light installation work. Fig. 3 and Fig. 4 show the shape and product drawing of the luminaire, and Fig. 5 shows the office plan and layout of the luminaire employed in this study. Time to completion of the ceiling light installation work was two days with three construction workers working, as shown in Tables 1 and 2, respectively. We observed and defined the task, activities, and motions of the workers. Table 3 provides definitions for each activity and motion of ceiling light installation. The motions have been classified into 19 types, and represent a detailed analysis of work motions performed for a manual task.
Fig. 7 demonstrates the proportion of total performance duration of each installation activity. Three workers installed 49 LED luminaires. Fig. 8 and Table 4 demonstrate the proportions of activity performance duration for each worker relative to the proportional sum of duration. To analyze task and activity times, we evaluate the learning curve for installing ceiling lighting. Fig. 9 shows the task times of workers, as represented by time to complete the installation tasks. This figure shows that task time per unit decreases with increased installation task repetition.
After analyzing task and activity times, we applied factor analysis to the activity times. The purpose of this analysis was to predict the cause of work delay. First, in order to investigate the relationship between total completion time and individual activity time, we calculated the correlation between completion time and individual activity time. The results are shown in Table 4. To validate the probability distribution of task time and individual activity time, we provide histograms (Fig. 10) of the times taken to perform each task and activity.
Next, we applied exploratory factor analysis (EFA) to the six activity durations. By eigenvalue (Table 6) and parallel analysis, we determined the number of factors to be one or two. Table 7 shows the results of exploratory factor analysis. This table displays the respective loading of each variable onto each factor for the cases of one and two factors.
In the case of two factors, Factor 1 is defined as "work environment of back plate installation work, " and Factor 2 is defined as "work environment of electrical wiring." According to the results of EFA, we reevaluated the work-time relationship between each activity and the two factors F1 and F2. Next, we calculated the factor scores generated from CFA, as shown in Fig. 12, and analyzed the cause of work delay. Fig. 13 shows the relationship plot of factor scores obtained via CFA analyses, with the horizontal and vertical axes being the scores of F1 and F2, respectively.
