Parallel machine replacement: an analysis in construction industry with considerations of horizon uncertainty, multi-purpose machines and transportation

In this dissertation, the procurement and replacement of interdependent assets is considered in which the machines satisfy demand in parallel. A number of realistic scenarios are modelled that are current limitations of the Parallel Machine Replacement Problem (PMRP). Considerations prevalent in construction management provide new formulations of the problem. A stochastic planning horizon is considered which is in line with the direction of the research field. Likewise, multi-purpose challengers are presented to offer a solution to the current heterogeneous fleet limitations. Lastly, shipping considerations for multiple demand sites are studied. New mixed-integer programming models are presented for each problem formulation. Each model considers numerous aspects that are contributions to the current literature for the parallel machine replacement problem. The work integrates the PMRP into construction management. A new solution methodology is presented that offers a usable technique for solving larger systems when shipping is of concern, without the limitations of the current models. The contributions are: considering multiple demand sites with shipping, a heterogeneous fleet, stochastic demand and planning horizon, multi-purpose machines, the ability to work and purchase used assets, applications in construction management and a solution method that is realistic and computationally efficient.