Multiproject Scheduling And Resource Allocation

Scheduling and allocating resources to multiple projects is much more complicate4.. than for the single-project case. The most common approach is to treat the severajg projects as if they were each elements of a single large project. (A more detailed e«* planation is given below when we consider a specific multiproject scheduling heuristic.) Another way of attacking the problem is to consider all projects as com pletely independent; see (28, 29], for example. As [28) shows, these two approach* ' lead to different scheduling and allocation outcomes. For either approach, the c ceptual basis for scheduling and allocating resources is essentially the same.

There are several projects, each with its own set of activities, due dates, and resource requirements. In addition, the penalties for not meeting time, cost, and performance goals for the several projects may differ. Usually, the multiproject problem involves determining how to allocate resources to, and set a completion time for, a new project that is added to an existing set of ongoing projects. This requires the development of an efficient, dynamic multiproject scheduling system.

To describe such a system properly, standards are needed by which to measure scheduling effectiveness. Three important parameters affected by project scheduling are: (1) schedule slippage, (2) resource utilization, and (3) in-process inventory. The organization (or the PM) must select the criterion most appropriate for its situation.

Schedule slippage, often considered the most important of the criteria, is the time past a project's due date or delivery date when the project is completed. Slippage may well result in penalty costs that reduce profits. Further, slippage of one project may have a ripple effect, causing other projects to slip. Indeed, expediting a project in order to prevent slippage may, and usually does, disturb the overall organization to the point where slippage due to resource shortages may then be caused in other projects. The loss of goodwill when a project slips and deliveries are late is important to all producers. As is the case with many firms, Grumman Aircraft, purchased by the Northrup Corporation in 1994, jealously guards its reputation for on-time delivery. During a project to install a new machine control system on a production line, Grumman insisted that the project be designed to minimize disturbance to operations in the affected plant and avoid late shipments. This increased the cost of the project, but the firm maintained delivery schedules.

A second measure of effectiveness, resource utilization, is of particular concern to industrial firms because of the high cost of making resources available. A resource allocation system that smooths out the peaks and valleys of resource usage is ideal, but it is extremely difficult to attain while maintaining scheduled performance because all the projects in a multiproject organization are competing for the same scarce resources. In particular, it is expensive to change the size of the human resource pool on which the firm draws.

While it is relatively easy to measure the costs of excess resource usage required by less than optimal scheduling in an industrial firm, the costs of uncoordinated multiproject scheduling can be high in service-producing firms, too. In the real estate syndication firm used as an example of an AON network in Chapter 8 (see Figure 8-28), the scarce resource is executive judgment time. If two deals arrived at the same time, one would have to wait. This is undesirable because other potential buyers are seeking properties, and the process must move along without delay.

The third standard of effectiveness, the amount of in-process inventory, concerns the amount of work waiting to be processed because there is a shortage of some re-source(s). Most industrial organizations have a large investment in in-process inventory, which may indicate a lack of efficiency and often represents a major source of expense for the firm. The remedy involves a trade-off between the cost of in-process inventory and the cost of the resources, usually capital equipment, needed to reduce the in-process inventory levels. It is almost axiomatic that the most time-consuming operation in any production system involving much machining of metals is an operation called "wait." If evidence is required, simply observe parts sitting on the plant floor or on pallets waiting for a machine, or for jigs, fixtures, and tools.

All these criteria cannot be optimized at the same time. As usual, trade-offs are involved. A firm must decide which criterion is most applicable in any given situation, and then use that criterion to evaluate its various scheduling and resource allocation options.

' At times, the demands of the marketplace and the design of a production/distri bution system may require long production runs and sizable levels of in-process inventory. This happens often when production is organized as a continuous system, but sales are organized as projects, each customized to a client order. Items may be produced continuously but held in a semifinished state and customized in batches.

A mattress manufacturing company organized to produce part of its output by the usual continuous process; but the rest of its production was sold in large: batches to a few customers. Each large order was thought of as a project and was organized as one. The customization process began after the metal frames and springs were assembled. This required extensive in-process inventories of semifinished mattresses.

As noted earlier, experiments by Fendley 119) revealed that the minimum-slack-first rule is the best overall priority rule, generally resulting in minimum project-slippage, minimum resource idle time, and minimum system occupancy time (l e, minimum in-process inventory) for the cases he studied. But the most commonly used priority rule is first come, first served—which has little to be said for it except that it fits the client's idea of what is "fair." In any case, individual firms may find a different rule more effective in their particular circumstances and should evaluate alternative rules by their own performance measures and system objectives.

Fendley found that when a new project is added to a multiproject system, the^ amount of slippage is related to the average resource load factor. The load factor is,1 the average resource requirement during a set time period divided by resource avaih ability for that time period. When the new project is added, the load factor for a re-, source increases and slippage rises. Analysis of resource loads is an important element in determining the amount of slippage to expect when adding projects.

Given these observations, let us examine some examples of the various types o multiproject scheduling and resource allocation techniques. We begin with a sho description of one optimization method, briefly cover several heuristics, and therjj. discuss one heuristic in greater detail.

Project Management Made Easy

Project Management Made Easy

What you need to know about… Project Management Made Easy! Project management consists of more than just a large building project and can encompass small projects as well. No matter what the size of your project, you need to have some sort of project management. How you manage your project has everything to do with its outcome.

Get My Free Ebook


Post a comment