Integrating Quality as a Measure of Performance in Resource-Constrained Project Scheduling Problems

A survey conducted in the United States in 1994 (Icmeli-Tukel and Rom 1998), revealed that the issues Project Managers (PMs) face today are more complicated than those studied in the project scheduling literature. The study indicates that for over 90% of PMs the objective is to maximize the quality of projects and their outcomes. They usually measure quality by the degree to which a project’s outcome conforms to the customer’s requirements and the degree to which the project completes within budget and on schedule. In order to achieve these goals, PMs are trying to eliminate the rework associated with non-conforming activities. A job is considered to be non-conforming when it fails to meet customer/design requirements and therefore needs some rework (Ireland 1991). Non-conformities usually result in increased project implementation time and cost. In the traditional Resource Constrained Project Scheduling Problem (RCPSP) formulations, however, the general assumption is that work will be done as scheduled and that rework will not be required. This assumption clearly makes the RCPSP models more amenable to solution yet less realistic. PMs report (Icmeli-Tukel and Rom 1998) that most of the projects they implement have at least one non-conforming activity which needs to be reworked before the project moves to the next stage. Because of these reworks, they have a difficult time anticipating delays and cost overruns when using traditional techniques. If an activity is (partially) reworked to satisfy customer requirements, then the start time of all immediately succeeding activities are delayed, which, in turn might delay the project completion time. This in turn contributes to the consistently poor record of project performance (Cooper 1994).