Creating a Project Network Diagram

Utilizing the Precedence Diagramming Method

You could start the PND by mapping out work units on a whiteboard, like you did for the WBS—which is a fine method. However, rather than creating a flat timeline of activity, project managers can use the Precedence Diagramming Method to create a fluid project structure. This sophisticated approach stems from the 1950s, when the original concept was called the activity-on-the-arrow (AOA) method. As you can see in Figure 7-4, each arrow represents a task in the project. The origin of the arrow is the “begin activity” sign, and the end of the arrow is the “end activity” sign.

Figure 7-4: The AOA method uses arrows to represent the start and end of activities.

As you can imagine, this representation of a project, especially a large project, could prove very cumbersome, as project managers can only have very simple relationships between tasks. What replaced the AOA method is the activity-on-the-node (AON) method. With the AON method, the focus is on the activities rather than on the start and end of activities.

This methodology of creating a network diagram is also called the Precedence Diagramming Method (PDM). PDM requires the project manager to evaluate each work unit and determine which tasks are its successors and which tasks are its predecessors. Once this information is obtained, you can begin to snap the pieces of the PDM puzzle together. You must give careful consideration to the placement of each task, as all tasks are connected even if they are scheduled to run concurrently.

Each unit of work in a network diagram using PDM is represented by a rectangle called the activity node. Predecessors are linked to successors by arrows, and are always upstream from successors, as Figure 7-5 demonstrates. One work unit can be both a successor and a predecessor; for example, the task to install and configure a web server may first require the installation of an operating system; only after the web server task is complete can the web pages the web server will host be added.

Figure 7-5: Arrows link predecessor and successor tasks using AON.

To begin creating the network diagram using PDM, a project manager will have to determine the order of tasks to be completed. Basically, a project manager asks, “What tasks must be completed before the next tasks can begin?” The activity list, which you derived from your WBS work packages, is what you are sequencing. You do not want activities that are too short in length to be useful. Based on your project, you may have work units that last days or weeks—generally not hours.

A rectangular box represents each individual task within the network diagram. As you add tasks to the network diagram, draw an arrow between tasks to connect them in the successor/predecessor relationship. As expected, the network diagram is read left to right, top to bottom. Each task within the diagram must have a successor and predecessor—except for the first and last task in the project. Once the diagram has been completed with successor and predecessor tasks, it is considered a “connected network.” A project manager, or anyone involved in the project, can trace any activity through the network.

The relationship between activities is what allows a project to move forward or to wait for other tasks to complete. An accurate description between tasks is required for the project manager to analyze and make adjustments to each task in the network diagram. Figure 7-6 shows the dependency types you may have between tasks. There are four types of dependencies:

Figure 7-6: Dependencies describe the relationship between tasks.

• Finish to start (FS) This relationship is the most common. It simply requires the predecessor task to complete before the successor task can begin. An example is installing the network cards before connecting PCs to the Internet.

• Start to start (SS) These tasks are usually closely related in nature and should be started but not necessarily completed at the same time. An example is planning for the physical implementation of a network and determining each network’s IP addressing configurations. Each are closely related and can be done in tandem.

• Finish to finish (FF) These tasks require that the predecessor task and the successor task be completed at the same time. An example is rolling out a new software package and finishing the user training sessions. While users are in the new training session, the new software should be installed and configured on their workstations by the time the training session ends.

• Start to finish (SF) These rare tasks require that the predecessor doesn’t begin until the successor finishes. You won’t find yourself using this activity all that often. It’s often used with just-in-time scheduling for inventory and manufacturing instances. For example, if you’re opening a coffee shop, you’ll want your inventory of coffee to be on hand as close as possible to when the construction of your shop is done. The coffee has to be ordered to officially open the shop, and the construction has to be done before you can open the doors. You’ll need to time the order of coffee with the completion of the construction in order to open your doors on time without your merchandise spoiling—or tying up your cash flow in inventory that’s not being sold.