Optimal project planning under the threat of a disruptive event

This article considers the problem of planning a complex project when there is the possibility of a Disruptive Event (DE) occurring sometime during the project. If such a disruption occurs, work on all activities will stop for some (random) time, but overhead and indirect costs will continue to accrue as well as possible penalty costs. Given information about the likelihood of such an event, how should a risk-neutral manager who wants to minimize the expected total cost of the project react (where total cost includes direct labor costs, indirect/overhead costs, and penalty costs)? Should a manager take preventive action at the start of the project (i.e., build additional slack into the project beyond that of a normal cost-minimizing schedule), act at any time during the project after gaining more information about the likelihood of a disruption, or wait until the DE occurs? The problem is formulated as a stochastic dynamic programming problem and this model is used to demonstrate several important implications for managers who face the threat of potential DEs. An efficient algorithm is described that can find the optimal compression strategies for large-scale projects; a numerical example illustrates both the algorithm and implications.