© PMICOS 2002

Privacy Policy
Terms of Service
Member Agreement

 

Topic of the Month


For: January, 2004

Activity on the Arrow or Precedence Diagramming?
Submitted by: Michael J. Harris

In 1982, as a second lieutenant teaching scheduling in the Army Corps of Engineers, I co-authored an article in the COE’s Engineer magazine titled “The Critical Path Method, Is it Time to Change?” The article focused on whether the Army Engineer School should primarily teach the Activity on the Arrow (AOA) method of network diagramming or the Precedence Diagramming (“PD”) method. Our preference leaned heavily toward the PD method because the flow of logic is more easily followed by workers, fewer activities are necessary to model the project, all significant activity information is consolidated on one node, and updating the hand drawn schedule diagram was easier. We noted that the AOA method had some advantages over PD. The primary advantage we found was that calculations could more easily be manually checked using an AOA network than in a PD network containing multiple leads and lags.

AOA, also known as ADM, or the “Arrow Diagramming Method”, depicts the activity on a line or arrow, and the ends of the activity with circles, or event nodes. The activity’s beginning node is typically called the “i” node, and the ending event node the “j” node. Multiple activities could share nodes for their start or finish. Some of us actually still occasionally refer to even PD logic printouts as IJ reports. “Dummy” logic arrows were added as appropriate to properly depict relationships. Dummy arrows would be necessary, for example, when Activities A and B must precede the start of Activity C, but only Activity A, not B, precedes the start of activity D. Here, an extra node and dummy arrows would have to be added so Activity B did not erroneously constrain the start of Activity D.

The PD method, or PDM for Precedence Diagramming Method places the activity on the node, and uses logic ties, or arrows for all relationships. Most of us are familiar with PDM since it is the method currently supported by the most commonly used scheduling software programs.

Today, it would probably be quite a challenge to find a program that supports ADM. ADM networks are very rare these days. The last one I came into contact with was several years old, the result of a specification that hadn’t been updated, and only surfaced because the project was the subject of a claim that took forever to resolve. The updates actually had to be converted to PDM so they could be properly evaluated.

It also appears that the US Army Corps of Engineers has decided which method they will require. The proposed “Unified Facilities Guide Specifications”, to be used by the Army Corps of Engineers and other military construction agencies require PDM networks.

So why even discuss ADM? Why did some industry experts lament the fact that ADM is fading away in a May 26, 2003 ENR cover article? The key is in the simplicity and transparency of ADM networks. As a scheduling consultant, I review numerous networks created to support construction projects. Finding the real critical path on many PDM schedules is often like trying to find which tin cup has the ball underneath. Plugged dates, excessive use of leads and lags, retained logic issues, zero free float options, hammocks, and calendar issues divorce float values from the true longest path through the project. Many networks have the “critical path” running through the start to start leads and not the activities. If finding the critical path is difficult in a forensic analysis, it must also be difficult for the project team. CPM is intended to help identify what activities the team must concentrate on to bring the project to a timely completion. The whole point of the critical path is to highlight those key activities.

Many of these problems were not issues with true ADM. The critical path was easy to identify. Activities were broken into enough detail so that leads and lags were non-existent, or at least were infrequent. Many of the convenient, albeit potentially hazardous scheduling options were not included in the programs, and networks typically contained fewer activities than they do today.

The good news is that PDM networks can be constructed with the simplicity and transparency of the ADM networks for which so many experts are nostalgic. Limiting use of leads and lags, avoiding constrained dates, “tying” the network together so that each activity has a predecessor and successor, and forcing the activity durations and logic to actually drive the date calculations are some good scheduling techniques that were forced by ADM, but should still be used in PDM.

The scheduling sections of the proposed United Facilities Guide Specifications, UFGS 01320A, “Project Schedule”, and UFGS-01321N, “[Design – Build] Network Analysis Schedules (NAS)” specifically require many of these techniques. Although it may be a sad commentary on the state of the practice that good scheduling techniques have to be specifically spelled out in the construction contract, perhaps it will help management teams successfully complete more projects.

About the Author: Michael J. Harris, PE, is the President of Genoa Corporation. Genoa Corporation specializes in CPM scheduling, Dispute Resolution, and Project Management for the construction industry.

 

PMI College of Scheduling

About Us

[ The Vision & The Mission | An Open Invitation | How to Join | Constitution & Bylaws | Board of Directors ]

Events

News

[ Charter Status | Newsletters | ENR Article | Bibliography of Scheduling ]

Featured Topic

[ October 2003 | December 2003 | January 2004 | June 2004 | July 2004 | November 2004 ]

Volunteering

[ Newsletter ]

Interaction

[ The President | Director of Communications | Web Site Administration | The Forums ]