Investigating operational workforce management decisions under uncertainty -

Abstract

Workforce management decisions (e.g., hiring, training, and staffing) have a direct impact on the cost, schedule, and quality of work. Researchers have developed several mathematical programming models to optimize such decisions. Most of these models are of a deterministic nature, i.e. they rely on well-known and pre-set input parameters such as supply and demand characteristics. However, in practice, there is significant uncertainty in these parameters, which jeopardizes the optimality of solutions obtained from these models. Moreover, they provide strategic decisions to be made over the entire project duration. This can prove to be inapplicable in projects which make use of a transient workforce, since these projects typically suffer from frequent changes in the supply and demand of workers. Lebanon has an unregulated construction labor market. It mainly depends on migrant workers from Syria who typically work on several projects in a short period of time. This highly transient workforce can cause difficulties in managing construction projects, and might lead to unpredictable rates of absenteeism, unsatisfactory productivity, and increased labor costs. This study identifies the characteristics of a transient construction workforce and measures the impact of several internal and external factors on absenteeism rates. Furthermore, it makes use of the results to present an optimization-based framework to make operational workforce management decisions for a transient workforce. The research method relies on a survey targeting 60 site engineers, construction managers and project managers who have access to labor related information. The results show that unskilled and skilled workers have different characteristics in terms of demographics, tenure of work, and wage structure. Furthermore, most of the respondents believe that external factors (e.g., holidays and political instability) have a bigger impact on absenteeism than internal factors (e.g., working conditions, interpersonal relationships). Therefore, an