Question

What are the common heuristics that can be used to schedule resources in a multi-project environment? Use example(s) to discuss and comment on their advantages and disadvantages.

Answer 1

Answer: The heuristics discussed in the text are first in line, greatest resource demand, greatest resource utilization, minimum late finish time, and mathematical programming.

The first in line heuristic allocates resources based upon which project enters the queue first. This is the easiest heuristic to follow, but it can defy common sense in favor of ease of use.

The greatest resource demand decision rule determines which projects in the company's portfolio will pose the greatest demand on available resources; these projects are allocated resources first and then other projects are scheduled. The assumption being made is that projects that are more consumptive of resources are likely sources of bottlenecks, so these activities should be accommodated first and the rest of the system subordinated to the bottlenecks' needs. This rule and the greatest resource utilization rule may result in dedicating too many resources to a project that simply isn't profitable or has poor margins when compared with projects that may represent low hanging fruit.

Greatest resource utilization is a slight variation on the greatest resource demand heuristic. Projects with high levels of resource utilization are scheduled first, resulting in good performance on the utilization performance measure. If other projects are added to the company portfolio while the first project is still viable, the utilization of resources may remain high when the new projects are considered with the existing projects.

The minimum late finish time prioritizes projects with minimal slack, in effect keeping them as close as possible to on schedule. Projects with more slack can wait a while for the resources without becoming late. The danger is that resource starvation of projects with seemingly abundant slack may result in every project in the portfolio being delivered late.

Mathematical programming can generate optimal decisions to resource constrained problems. While solutions are optimal, the problem set up and solution can be extraordinarily complex.