Incorporating Multiskilling and Learning in the Optimization of Crew Composition

Abstract

The presence of multiskilled workers in a crew can increase the crew’s productivity through reducing inefficiencies and supervision requirements, while also providing on-the-job learning opportunities for single-skilled workers. The effect of the presence of multiskilled workers on the learning rate of workers, which is also a function of skill level and experience, and thus on the crew’s productivity, is especially significant in repetitive construction projects. This paper presents a mathematical model for identifying the optimal combination of single-skilled and multiskilled workers with different levels of experience in the crew to minimize the duration of construction projects by accounting for the overlapping effects of multiskilling, skill level, and learning on the crew’s productivity. The model is applied to an illustrative case project to demonstrate the practicality of the model. The optimum crew compositions for different activities involved in the case project are identified using a solution technique which combines constraint programming (CP), statistical analysis (SA), and a genetic algorithm (GA).