Abstract
We consider resource-constrained scheduling of projects including activities whose resource loading is flexible. We propose a model of project-oriented production in which the application rates of the various resources required by an activity are indexed by the performance speed or intensity of the activity. Heuristic algorithms are introduced for the assignment of activity intensities through time in lieu of traditional, fixed-intensity start-time scheduling. On test problems from the literature, the assignment algorithms are shown to outperform even optimal scheduling algorithms for the fixed-intensity case, yet the assignment algorithms are cornputationally practical for scheduling actual, large-scale industrial projects. We also extend the model and the assignment algorithms to admit production-like workflow dependencies and to interface with aggregate models for multi-project planning.