The current practice of labor allocation in construction schedules assumes single-skilled workforce; meaning that each worker is assumed to be skilled in only one trade. In such practice, at any instance in the project lifecycle, some of the workforce become idle waiting for other labor types to complete their work. Traditionally, companies may relocate idle workers to other projects and return them back to their original project when needed again. This complicates the resource management process and is not often performed successfully, leading to schedule and cost overruns. Alternatively, project managers may keep the idle workforce at their projects because they will be needed at a later stage and pay them in their idle days, which adds unnecessary costs to the project. Another solution would be continuously hiring and laying off labor at need, which has severe negative impacts on projects and firms. Due to the inefficiencies of these solutions, some research discussed the idea of “multi-skilled” labor, where some of the workers may have enough training to carry out different activity types. Multi-skilling decreases inefficiencies and ensures a smooth and continuous progress of works whilst maintaining the workforce and keeping their idle time to a minimum. Multi-skilling could be also used to speed up progress in construction schedules.
Previous research efforts have been made to encourage contractors in pursuing multiskilling as a solution to the non-smooth resource histograms. Yet, the literature falls short in providing a robust multi-skilling framework; specifically, one that considers the cost of training labor and solves the partial allocation problem. The objective of this research is to improve project duration and minimize unnecessary costs through the utilization of multi-skilled labor. Through a multi-step methodology, a model that optimizes the allocation of multi-skilled labor resources was developed. The novelty of the presented model is that it further minimizes the idle times of labor when compared to previous multi-skilled labor models, due to its capability in allocating resources “partially” to segments of activities rather than to full activities. In other words, unlike previous models, the developed model recognizes the fact that a crew can work for a period of time in an activity, then some workers in that crew can be allocated to another activity, leaving the rest of the crew to complete the first activity. The model allows the user to enter any number of activities and up to ten different resource types. With the use of genetic algorithms idle resources are assigned to activities that require additional manpower in order to reduce their durations, and in turn reduce the project’s indirect costs. When applied to a case study, the model generated promising results, where the reduction in duration between the single skilled allocation and multi-skilled labor allocation was 31% and this reduction jumped to 44% when partial allocation was applied. Multiskilling did not only reduce the idle labor days, but it will also shift the resource usage histogram’s end point to the left, reducing the total project duration. This did not only reduce the unnecessary costs being paid to workers on days where they have no work, but it also reduced the total indirect costs which are directly proportional to the overall project duration.
Construction Engineering Department
MS in Construction Engineering
Committee Member 1
Committee Member 2
Mohamed Mahdy Marzouk
Committee Member 3
Institutional Review Board (IRB) Approval
Not necessary for this item
(2021).Multi-skilled Labor Optimization with Partial Allocation of Resources [Master's Thesis, the American University in Cairo]. AUC Knowledge Fountain.
Saleh, Amira. Multi-skilled Labor Optimization with Partial Allocation of Resources. 2021. American University in Cairo, Master's Thesis. AUC Knowledge Fountain.