Hammock Activity

Let me clarify that. Say Activities B.C.D. and E with respective durations of 2,3,1,and 4 days duration, are all dependent on Activity A which has a duration of 4 days. Activities B, C, D, and E are also predecessor activities to Activity F. In fact the activities B.C,D, and E have no other importance to be separate activities than to distinguish them from each other and E being on critical path. What if I merge these 4 activities into one with a duration of 4 days and let us call this X activity. Then the critical path would be A-X-F, where X stands for B,C,D and E. I would go for this merging option as there is no other dependency creating another path put of this network diagram. Now these 4 activities B, C, D and E are internal affair of newly formed Activity X and it does not matter in what order they are done internally, but they would be considered as Activity X externally. These are the hammock activities. Hope I made it clear to you.

good explanation,

Rami, Thanks for posting the original question and for continuing to stay involved with this discussion. Although I’ve been a PMI member for close to two decades, this is my first visit to the projectmanagement.com forum.

1. I am surprised that copying and pasting a series of not-exactly-consistent “answers” from various web forums – some presumably written by others – is here lauded as “the experts judgement” and “a good and comprehensive summary.” I was expecting a higher standard of discourse.

2. What exactly is a hammock activity?
a. In a logic-driven schedule, it’s an activity whose start and finish dates are established by explicit and independent relationships to other (primary) activities, and whose duration is automatically adjusted to fill the interval between the (early) start and finish dates.
b. Although the hammock activity may have “predecessors” and “successors” (depending on the software application), these relationships exist solely as inputs to establish the hammock’s start and finish dates. None of the hammock’s components – dates, duration, or relationships – should participate in the logic flow (forward and backward pass) calculations.
c. A hammock can be shown in either PDM or AOA network notations as essentially bypass logic branch. We would like them to be specially annotated as outside the network calculations, but most software doesn’t distinguish hammocks from other activities.
d. I speculate that the name comes from their shape in hand-drawn AOA (activity-on-arrow) diagrams - i.e. a curved line slung between two logic-driven event-nodes. While originally limited to connecting two nodes in AOA networks, hammocks modeled using software based on PDM (precedence diagramming method) networks are much more powerful.

3. How does it fit into Project Management or the Schedule?
a. Hammocks are useful for ad-hoc schedule summarizing and reporting of activities independent of their coding structure. For example, a turnkey factory project schedule may include WBS branches for Procurement, Fab & Ship, Construction, and Commissioning of three production lines. Based on prior experience, top-level management is primarily interested in “Line No. 1 Construction and Commissioning.” A single hammock may be constructed to summarize the various Line No. 1 activities from the different parts of the schedule. Then the hammock is included in the Management Summary.
b. Hammocks are also useful for establishing dates for indirect/supporting activities and corresponding resources. For example, “Commissioning Team Mobilization” starts with the first of three “Line Ready for Pre-Commissioning” activities and ends with last of three “Provisional Customer Acceptance” activities. A hammock is constructed from the six primary activities. Then, depending on software capabilities, resource/costs may be loaded to the hammock.

4. Clarifications:
a. A hammock activity is NOT a group. It is a single activity that summarizes a time interval defined by other (existing) activities.
b. The last paragraph of Suhail’s definition – defining a hammock as three tasks (B, C, and D) that are squeezed between two other tasks (A and E, whose dates are independently controlled) – is completely foreign to my ~25 years of CPM/PDM scheduling experience.
c. Suhail subsequent clarifies that these now-4 tasks – all 4 with a single common predecessor (A) and a single common successor (F) – “are the hammock activities” and are to be replaced by a new Activity X. This is even further from my understanding of what a hammock is.
d. The consolidation of activities with 100% logical concurrency (i.e. exactly the same predecessors and successors) is definitely worth considering on a case by case basis. It has nothing to do with hammocks.

5. Hammocks and Level-of-Effort (LOE) activities are distinct, but they are not distinct enough for both of them to be included in the same software package. Primavera P3 had hammocks, not LOE. P6 has LOE, not hammocks. Why? Because LOE is more flexible and provides better control over the schedule. Phoenix, Spider, Deltek, and Asta all have something called a hammock, though some are implemented more like an LOE than like P3’s hammock.

6. Hammocks and LOE activities also differ substantially from Summary (note upper-case “S”) activities, whose purpose is to summarize schedule and resource/cost information on the basis of hierarchal or other activity coding.

7. Typical Microsoft Project user experience incorporates extensive reliance on (hierarchical) Summary activities, and no hammock or LOE activities exist. Several workaround methods exist: See http://wp.me/p6CCB4-ct (An Alternate Approach to Hammock Tasks in Microsoft Project).