Browsing by Author "Erquicia, Santiago"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Fleet Asset Life Cycle Costing with Intelligent Vehicles
    (Center for Transportation Studies, University of Minnesota, 2008-08) Wyrick, David A.; Erquicia, Santiago
    Life cycle costing seeks to find the optimum economic life of a particular asset considering acquisition, maintenance, operational, and disposal costs over the time it is held. The economic life can vary depending on interest rates, depreciation, maintenance, and overhead. A model was built to calculate economic life cycles for four classes of passenger cars and three classes of motor trucks and truck tractors within Minnesota’s Department of Transportation using data from the M4 information system. For class 330 snowplows in Districts 1 and 6, cost data from M4 regularly were under-reported in comparison to the Minnesota Accounting Procurement System (MAPS) from previous work. Due to high uncertainty of input data integrity in M4, various sensitivity analyses were run. Results included families of cost curves to estimate optimal life cycles for varying cost parameters. A key finding is that data may not be recorded fully, accurately, or assigned to the correct asset, indicating the need for automating as much future data collection as possible. With good data, decision makers can determine how long assets should be kept and maintained in general as a fleet, keeping in mind that results from this model are not indicative of any single unit.
  • Thumbnail Image
    Item
    Quantification of Uncertainty in Transportation Infrastructure Projects
    (2005-06-01) Rosandich, Ryan G; Erquicia, Santiago
    Monte Carlo simulation is the currently accepted method for quantifying uncertainty in projects. It was the goal of the research presented in this report to develop a purely computational technique, based on traditional probability theory, for quantifying project uncertainty with accuracy equal to or greater than that of Monte Carlo simulation. Series and parallel operators were developed for combining independent task uncertainties in project networks. The operators were used to compute overall project uncertainty given individual task uncertainty, and to calculate slack and the degree of criticality for each task. Additional techniques were developed to deal with networks where the series and parallel operator were not enough, specifically those with path dependencies. Results equal or exceed the accuracy of Monte Carlo simulation, but computational times exceed those of Monte Carlo simulation for networks with many dependencies.