Scheduling and coordinating ongoing asset maintenance projects

点击翻译按钮获取中文摘要

The maintenance of public infrastructure such as roads, sewer systems, and transit systems is an ongoing, costly expense for any municipality. Similarly large capital intensive corporations in areas important to Canada (e.g., mining) are faced with the challenge of maintaining their equipment. Riva Modeling, the industrial partner in this proposal, markets software that allows such organizations to create maintenance plans for their different classes of assets over a 100 year time-frame. In this project, we will address the reality that large organizations often manage many simultaneous and competing maintenance projects. For example, a city may need to balance among ongoing maintenance on roads, sewers, and the urban forest. Using the output from Riva's existing software, we will investigate the problem of scheduling asset maintenance projects over a 10 year time-frame when the projects compete for the same budget, the same resources (e.g., labour force), and the same risk limitations. Using real data, we will develop and empirically compare a number of mathematical programming models using different technology (mixed integer programming, constraint programming, constraint integer programming, heuristics) and develop prototype software that will integrate with and complement Riva's existing software. This prototype will automatically transform the set of maintenance project plans plus some company data into a mathematical programming model that will then be solved, providing an optimal schedule while balancing annuals costs, resource usage, and risk.