How do we efficiently add resources to shorten a project already in process? Let us discuss.
It is common for project managers to use the global change tool when overseeing a project in progress and behind schedule to add resources so that the work can be completed on time. This is known as crashing the schedule. The term reflects intervention required to “crash” through normal schedule barriers to achieve a completion date. This approach may shorten the project’s duration, but empirical data, i.e., real-world historical project data, demonstrate that this always increases costs. In other words, you never get the time savings correlated (or equivalent) to the number of added resources for increased production.
But if compressing your schedule’s time is paramount, then these cost increases to save time may become an agreeable necessity. Increasing the number of resources on numerous tasks in the schedule can become a tedious, repetitive process.
However, in P6 Professional, schedulers can create routines in the Global Change tool to automate these processes that are otherwise data entry-intensive. Ten Six previously wrote an article to demonstrate how to create a Global Change routine to increase respective resource efforts on all their task assignments in a not-started schedule. In this investigation, we consider the same procedure, but for a schedule that has already begun and with completed tasks.
This article demonstrates how to use the global change tool to create Global Change routines to adjust the effort of resources to shorten a partially progressed schedule with some concluded tasks.
Ten Six Article – Adjust Effort on a Not-Started Schedule
Our demonstration is a continuation of the Ten Six P6 Global Change tool article at the following link:
In this article, the Global Change tool is introduced, and the process of modifying the effort of resources on their assigned tasks in a not-started schedule is shown.
Demonstration – Adjust Effort on a Progressed Schedule
Now we investigate and present how to adjust resource assignments to compress a schedule that has been progressed and has some completed tasks. The schedule at this point is time-constrained and cost-constrained. Figure 1 displays our demonstration schedule, pointing out the current finish versus the planned finish dates on the Gantt chart.
Figure 2 is a close-up view of the Install Grounding task resource assignments, and it outlines in red the lineman and groundman resource assignment attributes.
We inspected the Install Grounding task because it has all the labor resources, and its resource assignments could provide insight into other tasks’ resource assignments in the schedule. When we look at the resource assignments for the Install Grounding task, Figure 1, we find the lineman and groundman resources have a Max Units/Time of 40h/d, which means we have four of each resource available to work ten-hour days each. (The calendar assigned to each task in this schedule is 4×10 or a four-day workweek at ten hours per day.) However, the Default Units/Time of 10h/d indicates that only one lineman and one groundman were included in the initial resource assignment. The Budgeted Units/Time of 20h/d for the lineman and groundman shows that the initial assignment of these resources was later doubled.
But this is just one task. What about these resource assignments on all the other tasks in the schedule?
Activity Resource Assignment Reporting
We want to examine the lineman and groundman attributes on all their task assignments in the schedule. To do this, we generated an Activity Resource Assignment report, Figure 3, of the lineman and groundman’s Max Units/Time, Default Units/Time and Budgeted Units/Time for every one of their task assignments.
For an introduction to P6 Professional reporting, refer to the following article link:
(The difference between this article’s P6 Reporting and the report in Figure 3 is that the subject area for the Figure 3 report is ‘Activity Resource and Role Assignments.”)
In Figure 3 Activity Resource and Role Assignments report, the lineman and groundman resource assignments are consistent for each task; out of four available, the initial assignment was one each, which was later doubled to two each. The Project Manager must have concluded that doubling the lineman and groundman resources on all their task assignments was required to achieve the project’s target end date.
Scenario – A Late Schedule
The Gantt chart in Figure 1 clearly indicates that the project is behind schedule. Note on the Gantt chart that the black diamond current finish milestone is later than the yellow diamond baseline finish milestone, revealing that this project has already progressed and is late. Further, the Dig Cable Trench task was worked out of sequence to try to align the schedule’s earned value with the original plan.
For a primer on scheduling in process tasks, refer to the article at the following link:
Despite this out-of-sequence work tactic, the current schedule finish of April 25, 2029, is still nine business days behind the original April 10, 2029, planned finish date.
Late Schedule Shortening Strategy – Crashing the Schedule
In this project’s situation, completing on time was of the utmost importance, so the project manager decided to implement a “crashing” the schedule strategy. The Project manager proceeds and adds an additional lineman and groundman to each task requiring these resources. The thought is that this will shave a little time off each of these tasks. Assuming many of these tasks are on the critical path, it should result in material time savings. Hopefully, then the project can achieve its target end date.
Crashing a Progressed and Late Schedule
Let us demonstrate this crashing of the schedule for the scenario in Figure 1, a progressed schedule, and where some tasks are complete. We use the Global Change tool to modify task attributes. One rule we want for our routine is that our Global Change routine excludes completed and in-progress tasks. This is accomplished with a Global Change tool line requirement that tasks not be started before we operate on them. This way, we modify only those tasks that have not begun. (We do not include in-progress tasks because their duration, effort, and work computations are more cumbersome to explain and understand.) For the remaining not-started work, we want to see how adding resources can improve production rate and proportionately shorten the duration of the schedule.
Remember, we have four of each, lineman and groundman resources, available to work on our project. Currently, we have only assigned two each. We want to add an additional lineman and groundman resource to each task that has not begun. We hope that this way we can achieve the schedule’s baseline finish date.
Direct Approach – Increase Effort
The direct approach is to multiply the current 20h/d effort by 1.5 to make it 30h/d, so that we have three of each lineman and groundman resources working full ten-hour days. However, to operate on effort, Units/Time, the only option in the P6 Professional Global Change tool is the Activity Resource Assignments subject area. This means that individual resource assignments are modified and not the whole task, which we note. Also, our assignments work best when we use the Fixed Units duration type to add the additional lineman and groundman resources.
Refer to the following article for an introduction to P6 duration types.
Effort and Duration Correlation with Fixed Work
Why do we set the duration type to Fixed Units? When work, Units, is fixed, there are two variables: duration and effort. But in the P6 Professional hierarchy, the software wants to change the duration first, then the effort. Mathematically, in P6 Professional, with Fixed Units, the effort and duration of the resource assignments are correlated. If effort increases, the duration decreases in direct proportion. Therefore, using Fixed Units, we adjust effort and compute duration, and achieve a one-to-one correlation between effort and duration, which is what we want.
Resource Efforts – Install Bus and Jumpers Task
Let us examine the direct approach, where we adjust effort and compute duration, on a single task. Here, we investigate what happens when effort is directly increased on each respective resource assignment. We outlined in yellow in Figure 4 the Install Bus and Jumpers task’s 12-day remaining duration, and the 4, 12, 12, and 12-day remaining durations for each of the four labor resource assignments on this task.
From the computations on this Bus and Jumper’s task, we can find a pattern that repeats for the other tasks in the schedule. In Figure 4, we want to increase the lineman and groundman’s 20h/d resource assignment efforts to 30h/d each, reduce their remaining durations in proportion to this increased effort, and see if this shortens the remaining duration of the Install Bus and Jumpers task, which currently is 12 days. Note, again, that the duration type of the Install Bus and Jumpers task was set to Fixed Units in this investigation.
We proceed and make a Global Change routine to reduce the effort of the cable splicer and HEO. The Global Change routine to perform this, Figure 5, is an investigative and instructional demonstration that operates solely on the task W1120, Install Bus and Jumpers, to limit its scope.
Note that the subject area, Figure 5, is Activity Resource Assignments; the routine will operate on individual resource assignments.
The Figure 5 Global Change routine consists of an IF portion including matching criteria and a THEN section.
For a primer on matching criteria, refer to the following Ten Six article.
The IF portion of the Global Change routine (outlined in yellow) has two blocks or criteria (outlined in red and blue). These two blocks are joined by the matching criterion conjunction ‘And’, which requires both blocks’ criteria to be true to compute the THEN portion of the routine. The first block (red) limits the operation to task W1120. The second block (blue) has two lines of criteria joined by the matching criterion ‘Or’ that requires the resource assignment to be either the lineman or the groundman. When the IF section is satisfied, the THEN portion directly increases the Budgeted Units/Time by a factor of 1.5. The lineman and groundman resource assignments on this task are increased from 20h/d to 30h/d, which essentially adds one additional resource to each.
Direct Approach – Results
In Figure 6, as planned, with increased efforts, the lineman and groundman’s remaining durations were each reduced proportionately to 8 days, but this did not affect the remaining duration of the Install Bus and Jumpers task, which remained 12 days.
Why did our increased effort not affect the task’s duration? The problem is that the HEO resource assignment’s remaining duration stayed at 12 days, longer than the updated 8-day lineman and groundman remaining durations, and it is preventing the Install Bus and Jumpers task from shortening. (Note that the cable splicer’s 4-day remaining duration has no effect, as it is the shortest remaining duration for the Install Bus and Jumper’s resource assignments.)
As our investigation shows, the problem with the Activity Resource Assignments subject area is that the Global Change tool must operate on individual lines of resource assignments, and not the whole task; you could then have differing remaining durations for each resource assignment line. And the remaining duration of the task is determined by the longest Activity Resource Assignment’s remaining duration. So, your time savings on one resource assignment may not affect task duration because another resource assignment’s remaining duration is longer. The ideal is for the other task resource assignments to have the same remaining duration as the one you operated on or less.
To shorten the determinative activity resource assignment or assignments, in this case, we would need another Global Change routine to reduce the remaining duration of the unchanged HEO resource assignment; this becomes tedious. More disconcerting, our investigation reveals that the resource assignments’ remaining durations can differ in a way that prevents the task from being compressed; our increased resource assignment effort may therefore be of no advantage. We conclude that we do not want to directly specify effort, Units/Time, because we are limited to the options in the Activity Resource Assignments’ subject area, which drives each resource assignment separately.
Indirect Approach – Compress Task Duration and Indirectly Compute Effort
It is better to implement the Activities subject area to modify the task’s duration and let P6 indirectly compute the correlated effort. The Activities subject area, again, operates on attributes of the whole task. And, if we don’t want to worry about the P6 hierarchy, we could set the duration type of these tasks to Fixed Duration and Units. This way, they become time and cost-constrained, and effort, Units/Time, is the undoubted sole variable that could change.
Proceeding, we do not directly multiply effort by 1.5 to increase our production rate. Instead, and with duration type Fixed Duration and Units, we divide the remaining duration by 1.5. We chose Fixed Duration and Units because Effort then becomes the variable correlated to our remaining duration adjustment. Our reduction in the remaining duration of the whole task to achieve our compressed schedule should result in a 1.5 times increase in effort. This will achieve the 30h/d required effort (three of each resource working full ten-hour days). And there is no doubt about the schedule-shortening effect of the indirectly computed efforts.
We begin the indirect approach using the Global Change tool and by selecting the Activities subject area. We create a Global Change routine that first requires that a task be both not-started and task-dependent (excludes milestones and level of effort (LOE) tasks), then sets the duration type of these respective tasks to Fixed Duration & Units, Figure 7.
Again, the requirement, Figure 7, for the task to be not-started is how we exclude completed tasks and in-progress tasks, and operate solely on tasks that have not begun. After setting the duration type, in our routine, we divide the remaining duration of all these tasks by a factor of 1.5. As time and work are constrained, to achieve the same amount of work in this reduced time, the effort must increase proportionately.
With our newly created ‘Adjust Duration Down – Not Started’ Global Change routine, we click the Apply Change button in Figure 8 to adjust the duration down for all not-started tasks.
The Global Change report appears in Figure 9 and looks good.
Indirect Approach – Results
We commit the changes reported in Figure 9 and review Figure 10, which shows the resource assignments for the Install Grounding task; this will reveal a pattern for other tasks and their resource assignments.
We find that the lineman and groundman resource Budgeted Units/Time increase to 30h/d, which means that we have three assigned to each, Figure 10, refer to the blue outlined resource attributes.
This is what we intended, but the cable splicer and heavy equipment operator’s (HEO) effort increased from 10h/d to 15h/d, Figure 10, refer to the red outlined resource attributes. Unfortunately, in our situation, we only have one cable splicer and one HEO resource, so we cannot assign an additional cable splicer and HEO. And the current cable splicer and HEO resource each appear overallocated by five hours, which is excessive.
As our Install Grounding task results indicate, shortening the task duration to indirectly calculate respective resource assignment efforts is not without issue. We calculated attributes on the whole task; its remaining duration was reduced to 1.33 days, and each resource assignment’s remaining duration became 1.33 days, which is what we want. However, each resource assignment line could compute a unique effort.
Some of these efforts on individual resource assignment lines need adjustment to make them more realistic. These updates will require another Global Change routine, a negative we would rather avoid. But this Activities subject area approach is advantageous; there is no confusion with one resource assignment’s shortened remaining duration having no effect on the task because another resource assignment’s longer remaining duration. Here, all the task’s resource assignments would have the same remaining duration as the one operated on, if not less.
Adjust Individual Efforts
When we assessed our modified schedule, Figure 10, we found that our remaining durations for all the Install Grounding task’s resource assignments all matched, and we have three lineman and groundman resources, which is good. But our five-hour overallocated cable splicer and HEO is problematic. We consult with the line manager, who assigns actual individual resources to work on each task. Our plan is to negotiate an acceptable effort for each of these resources.
Our thought is that the cable splicer and HEO’s work requirements may have initially included some padding, which is when the estimator overestimates the required effort. Padding occurs for several reasons, including to show the project finished before the plan or so a respective resource has a full day’s work when the workload requires less. This negotiation with the line manager may require diplomacy to find an agreeable daily effort for the cable splicer and HEO.
In our situation, the line manager concurs that a reduced effort of the cable splicer and HEO would be more realistic. We proceed to decrease the cable splicer and HEO’s effort to align them with the true scheduling situation. First, we must change the duration type of these tasks. Figure 11 displays the Global Change routine, acting on Activities, to change the duration type of all not-started task-dependent activities to Fixed Duration and Units/Time.
We chose Fixed Duration and Units/Time so that when we alter the effort, work is the variable that updates, as appropriate. In Figure 12, we click to apply our Global Change routine, Adjust Duration Type – Not Started, and adjust the duration type for not-started tasks to Fixed Duration and Units/Time.
Figure 13 is the change report for the above Global Change routine; it looks correct, so we click the commit changes button.
Figure 14 shows the resulting duration types for the schedule, confirming our desired Fixed Duration and Units/Time duration type.
Now with Fixed Duration and Units/Time duration type, as mentioned, the variable becomes work, Units.
We proceed and make a Global Change routine to reduce the effort of the cable splicer and HEO, and compute the required amount of work. The routine, Figure 15, consists of an IF portion including matching criteria and a THEN section.
The IF section (outlined in yellow) consists of two blocks (outlined in red and green) separated by an ‘And’ conjunction, which means both block requirements must be met for the computation to execute. Each line in the first block is separated by the ‘Or’ conjunction, which means that if either of that block’s requirement lines is met, then that block’s criterion is met. The first block is satisfied if the Resource ID Name of the resource assignment is either a cable splicer or HEO. The second block is one line. It requires that the Assignment % Complete of the resource captured in the first block is 0% (or not started).
To summarize, the IF statement lines say, if the respective resource assignment is either the cable splicer or HEO and not started, then in the Then section, perform the computation.
The THEN portion divides the Budgeted Units/Time of the resource by 1.5. This takes a 15h/d Budgeted Units/Time effort and makes it 10h/d. And the five-hour padding on these Activity Resource Assignments is removed.
In Figure 16, we click Apply to execute the ‘Adjust Effort Down – No Padding’ global change routine.
The change report is displayed in Figure 17, which looks good.
We commit changes, Figure 17, then select the Install Grounding task in the activity table and review its resource assignments in the bottom details, Figure 18.
The outlined cable splicer and HEO Budgeted Units/Time fields now have each working a full ten-hour day, which is what we want and negotiated for. The adjustable variable in this Global Change operation was work, Budgeted Units, which decreased from 20h to 13.3h for each. The Budgeted Units and Budgeted Cost of each decreased because we eliminated the padding.
Now that our efforts are all realistic, the final step is to recalculate the schedule to see if the current finish is earlier than planned. Figure 19 shows our schedule just prior to scheduling, when the Gantt chart shows the current project finishing after the baseline finish date.
We proceed and choose Tools | Schedule and click the schedule button, Figure 19. The final task table and Gantt chart after scheduling appear in Figure 20.
The start and finish dates of all tasks, Figure 20, have been recalculated. And the current finish date of the schedule is before the baseline finish date; our schedule crash effort achieved its objective. Also, except for the small decrease in cost from the elimination of padding on the cable splicer and HEO workload, our budget remained fixed throughout the crash process.
Keep in mind that this cost is the ideal mathematical result. Empirical data from completed projects, however, demonstrates that you never get the time savings correlated (or equivalent) to the number of added resources, and cost, therefore, increases. The good news is that the current plan is to finish one week ahead of schedule, so it is probable that we will meet the finish date. But, most likely, the project manager will have to use management reserve funds to make up for the inefficiencies in executing the plan at an increased output.
Summary
Projects are unique; they have never been done before. To produce the scope of this one-off project deliverable, project managers have limited time and funding. And these constraints: scope, time, and cost each compete for the project manager’s attention. Often, a project plan begins in earnest but gets bogged down in execution. Project managers at this juncture must consider their highest priority, which often is the scope and schedule.
Their strategy usually is to risk incurring increased costs to achieve the prized scope and schedule targets. Their tactic is to increase the work completion rate through the addition of resources. This should decrease the time, which is what you want. A common crashing tactic is to add additional resources to numerous not-started tasks, hopefully, on the critical path, to shave time off each task duration and the whole project.
Ten Six recommends using the Activities Global Change tool subject area and adjusting the remaining duration, with the Fixed Duration and Units duration type. This will shorten the schedule and indirectly compute your required resource assignment efforts. The Global Change tool in P6 Professional is handy for making this scheduling operation efficient. Include the IF requirement that the activity status is not started to shorten the remaining duration of the not-started tasks, but not in-progress and completed tasks. And, with the Fixed Duration and Units duration type, where the variable is effort, modify the duration, and indirectly calculate the effort. In this way, all the task’s resource assignments’ remaining durations should be equivalent to or less than those operated on.
Then, for the efforts not aligned with resource availability, make the subject area Activity Resource Assignments and set the duration type to Fixed Duration and Units/Time, where work is the variable. Continuing, adjust these efforts as appropriate and compute the equivalent work. Achieving the schedule target is more science; stakeholders want a clearly defined schedule duration, whereas specifying the associated efforts in this process is more of a pliable art form. There is more flexibility for schedulers in allocating a resource’s effort to avoid or minimize the overtime required to achieve the target duration.
There are several ways to handle resource overallocations, as discussed in the article at the following link:
So, the tactic is to operate on the Activities subject area first and adjust durations, then operate on the Activity Resource Assignments subject area and consider effort adjustments, which have more leeway and solutions for addressing overallocation dilemmas.
In our scenario, we couldn’t add an additional cable splicer or HEO; the Max Units/Time allowed only one full-time resource for each. However, an assessment of their daily efforts and required work indicated padding. It was appropriate to therefore reduce the cable splicer and HEO’s daily effort and work, respectively. In this way, the schedule aligned with the resource availability and the true required daily output.
Conclusion
When we set aside our project’s padding situation and consider solely the crashing of the schedule, mathematically, the scheduling software shortens the project’s duration by an amount correlated directly to the added resources. The schedule shortens and the effort increases by a proportionate amount. And the cost remains fixed during this crashing operation.
However, much empirical data from historical projects raises a flag of caution, as an increased production rate comes with waste in the process. So, make sure the project has some management reserve to absorb the cost of this process loss. Crashing the schedule affects the triple constraint and management effort in that the scope and time are achieved, while the project manager must plan for increased cost due to inefficiency.