Project Planning 2: PERT Charts

Last week, I talked about an important tool for planning and organizing the work on a project called the work breakdown structure. This week, I want to take a look at PERT (Program Evaluation and Review Technique) charts. Where the work breakdown structure simply defines what each element is, PERT charts build upon these definitions to display important information about how each element of a project relates to the others.

Breakdown

Consider a work breakdown structure for, say, a simple scenic portal, covered in duvetyne with a paint treatment applied:

1. Flat Construction/Assembly
   1. Drafting
   2. Material Order
   3. Stock Flats
      1. Pull from storage
      2. Preparation
   4. Custom Flats
      1. Framing
      2. Facing
   5. Assembly
      1. Layout units
      2. Bolt units together
   6. Soft Cover
      1. Cut Fabric
      2. Attach Fabric
2. Painting
   1. Rendering
   2. Color Mixing
   3. Paint Order
   4. Paint Application
3. Installation
   1. Rigging
   2. Jacks

This WBS isn’t exhaustive by any means, but will serve for our purposes. This outline of elements does a really great job of describing the smallest, estimable elements of the project. (It would be relatively easy, for example, to give an estimate of how much the framing for the custom flats would cost, and how long that framing would take.) However, this information doesn’t tell us much about when we can begin any of these elements, or what impact completion of a particular element has on the rest of the project. A PERT chart takes this information and describes the relationship between each element by indicating dependences: which elements require the completion of other elements before they themselves can be completed.

Nodes, nodes, nodes.

In generating a PERT chart, we’ll draw a box or node to represent each element of the work breakdown structure. (PERT charts can be drawn as either action-on-arrow diagrams, or action-on-node diagrams—we’ll discuss the latter, here, as they are more informative for our purposes.) Each node should include a list of predecessor and successor nodes; predecessor nodes are those nodes which must be completed before this node can be completed, and successor nodes are those nodes which are dependent on the completion of this node for their completion. Just to keep things (seemingly) complicated, nodes can have multiple predecessor or successor nodes.

For example, considering the work breakdown structure above, node 1.3.1, “Pull from storage” would be considered a predecessor of node 1.3.2 “Preparation,” as you couldn’t really complete the preparation of the stock flats until they’d been retrieved from storage. Similarly, node 1.5.1, “Layout units” would be a successor node to node 1.3.2, as the stock flats would be layed out after they’ve been prepared for assembly.

Additional information might be included in each node, to assist with scheduling later. We might include, for example, the estimated duration of the task. Later, after building our PERT chart, we might also want to determine such information as early start date/time, early finish date/time, late start date/time, late finish date/time, and the slack time (more on these terms below). Each node should include space to list this information.

Boy, that sure is a PERTy drawin’!

Once we’ve determined each node we’ll require for our chart, we can begin to plot each node in the drawing. A PERT chart begins with a “start” node, at the left side of the diagram. Next on the chart, place the nodes for all activities that have no predecessors, and connect these to the “start” node. Next, place successor nodes to these first nodes in the diagram, connected with arrows to their predecessor nodes. Continue in this fashion, working left to right, until you have placed those nodes which have no successor nodes; connect these nodes to a “finish” node at the right side of the chart.

The drawing below represents one version of a PERT chart for the work breakdown structure above. It is worth noting that dependencies are directly related to how you plan to accomplish a project; in this example, I assume that the scenic charge won’t make up any detailed color mixing recipes until after they’ve ordered paint, but one might also assume that they might do all of their mixing before ordering any paint—in which case, the dependencies would change, and the position of each node on the chart would need to be adjusted.

At a glance, it is easy to see from the chart what elements need to be completed before others, and which elements can be completed concurrently, should the resources be available to do so. However, one of the major drawbacks of the PERT chart—as is amply demonstrated by even our simple version, detailing the construction, painting, and installation of a scenic portal—is that it can quickly become massive and unwieldy. (Imagine if we’d actually plotted an entire scenic build!) Additionally, as projects become more complex, the dependencies can become complex as well, making the drawing more difficult to generate and to read.

These limitations aside, a PERT chart provides a valuable visualization of dependencies, as well as a quick method of seeing what is called the “critical path”—the sequence of dependent nodes whose completion—in order—determine the overall time required for the completion of a project. Determining the critical path relies on the “early start/early finish” and “late start/late finish” boxes we put in our PERT chart nodes. We’ll look these terms, and at determining the critical path, next week.