Sharing in construction projects - On determining optimal container assignments for the on-site accommodation of trades

The sharing economy is among the top trending topics in operations research. In this paper, we introduce a novel sharing application encountered in the construction industry. The problem under research, which has been developed in close cooperation with an industrial partner, is concerned with the on-site accommodation of trades in large-scale construction projects. In such projects, a construction logistics company provides containers gathered in an area of the site and assigns trades to them based on their requests issued online. Usually, a single trade will only remain on-site for a certain period of time, i. e. for as long as it is carrying out its activities in the project. Similarly, containers will only be provided for a certain time interval. This dynamic nature may require re-assigning trades as the project progresses. As such re-assignments impose costs for the logistics company, it is the goal to keep them to a minimum. We formally define the resulting optimization problem, present two mixed-integer programming formulations and develop a fix-and-optimize heuristic. In an extensive computational study, we show the latter to not only clearly outperform standard solver CPLEX operating on the holistic mixed-integer program, but to also provide substantial savings over the assignment policy currently applied in practice.