Dynamically allocating expediting funds in projects with schedule uncertainty

When large-scale projects have tasks with uncertain durations, it is often necessary to expedite tasks in order to improve the probability of completing the project before a fixed deadline. Whereas traditional project scheduling tools focus managerial attention on tasks that comprise the longest path(s) through a project network, tasks with uncertain durations on other paths may ultimately delay the completion of the project. In addition, the status of a project schedule can vary over time as tasks are completed (early or late) and new information concerning future tasks becomes available. This research introduces and evaluates a matrix-based procedure for allocating expediting funds relative to the traditional deterministic approach and a method based on task Criticality Indices (CIs). The results from a large simulation experiment show this new matrix-based procedure to be the preferred method for allocating expediting funds in order to manage schedule risks over the life of a project. [Received 15 August 2008; Revised 15 February 2009; Accepted 18 February 2009]