Proceedings of the 34th International Symposium on Automation and Robotics in Construction (ISARC)

Abstract

This research covers, in part, the growing need of computerization in the construction industry.The paper highlights the lack of computerized measurement of labor performance/productivity and proposes the optimization of an existing performance measurement method with the help of a computer software.After extensive literature review, Work sampling is selected as the measurement method while MATLAB is employed for its optimization.Data collected by observing a total of four workers from two construction sites (two from each) undertaking steel works is first analyzed using conventional Work Sampling method.The same data is them analyzed using an algorithm written in MATLAB that computes multiple iterations for all number of observations.The results obtained show that work done in each category by all the workers comes out to be approximately the same for a much lower number of observations as compared to the traditional practice of taking 384 observations.For example, the percentage value added work done by Worker A is 12.12% for 384 observations while for 155 observations it comes out as 11.92% giving a percentage error of just 1.65%.Similar pattern is observed for all the workers.It is also observed that the percentage of error below 150 observation remains quite significant even if the number of iterations are increased.This is due to the increased probability of missing certain categories of work because of reduced observations.