Activities that have high float or no successor make monitoring schedule progress more difficult. They also may indicate missing logic. And finding the best way to tie up these dangling activities may require some investigation.
Relationships are like gears; they are the interconnection between tasks in the schedule. These relationships are defined in the precedence diagram, where the predecessor is the cause and the successor is the effect. And the cause drives the effect.
Usually these relationships or links between activities are apparent. It just makes sense that the relationship between ‘pour foundation’ and ‘install walls’ is finish-to-start (FS), where you must completely finish ‘pour concrete’ before commencing ‘install walls’.
Sometimes though the true interconnection or relationship between activities is not so obvious. An activity, perhaps, may not have an apparent logical successor. If you do not know the successor how can you explain its linkage to a predecessor?
With few exceptions schedule activities require both a predecessor and successor. What do you do though in situations where an activity has no clear or credible successor? Do you leave it dangling in the schedule with no successor? The correct answer is no you don’t. Instead investigate the situation to find the best logical approach among alternatives.
This article investigates alternative approaches to handling dangling schedule activities and in Figure 1, is our demonstration project schedule.
Note in particular task 20 install fence. This task has no successor assignment. Its total float, therefore, measures to the end of the project, and is 31-days. This means ‘install fence’ can delay 31-days and not affect the end date of the project. Although 31-days total float is acceptable, as per most scheduling guidelines, some may consider install fence a high float activity. High float activities are discussed in detail in the following blog The DCMA 14-Point Assessment and High Float Tasks.
Install fence may not meet the precise definition of a high float activity, but it definitely is missing logic; it has no successor. You can review missing logic in the following blog The 14-Point Assessment and Schedule ‘Missing Logic’ Inspection.
So install fence is left dangling with no successor. Again, it is missing logic. At this point schedulers often connect the dangling activity to a ‘project complete’ milestone, which satisfies the missing logic guidelines. However, the respective task remains in a high float situation.
In our demonstration we really want to complete ‘install fence’ much earlier than the ‘project complete’ milestone. The natural response at this conjuncture is to constrain ‘install fence’ with a finish no later than (FNLN) date constraint. This constraint is a way to restrict the ‘install fence’ task, and specify the importance of a particular date in the life of the project.
In Figure 2 we place a FNLN constraint on ‘install fence’ and a constraint date of 24-September-2018.
Note the total float on ‘install fence’ reduces to 2-days. Perfect! So ‘install fence’ is scheduled to complete much earlier than previously. But this constraint insertion solution comes with collateral drawbacks.
One drawback is that constraints, including our FNLN, do not automatically update with schedule duration changes. A schedule that has many constraints must be manually updated for each separate constraint, which quickly becomes laborious. Constraints should also come with a note that clearly explains the purpose.
One former student admitted they inserted constraints as a way to hide total float from subcontractors. The theory goes subcontractors will delay project start execution until the last possible moment. This last possible moment, accordingly, is when all the available task total float has been burned up. Yes, you can hide total float with constraints, but it is a less than honest scheduling practice. Constraints should be used sparingly and should come with a note of explanation, so that the intent is clear to all stakeholders.
There is also a better alternative to using constraints. Try inspecting your schedule to find a logical successor to the task in question. In our schedule it makes sense to link ‘install fence’ to the ‘substantial completion’ task in a finish-to-start (FS) relationship. In Figure 3 the FNLN constraint is removed and ‘substantial completion’ made the successor to ‘install fence’ in a FS relationship.
This avoids missing logic, but our ‘install fence’ task still has 21-days total float, Figure 3. Again, many would consider 21-days total float excessive. Let’s investigate to find other potentially better solutions.
Consider the ‘install fence’ task and associated deliverable. What is the purpose of the fence deliverable? The fence provides job site security. When would job site security be critical? Before arrival of dangerous equipment. Well, then we want ‘install fence’ to complete before arrival of this hazardous equipment. In light of this the logical successor to ‘install fence’, in a FS relationship, is ‘install bus and jumpers’, which are high voltage electrical components. In Figure 4 we make ‘install bus and jumpers’ the logical successor to ‘install fence’, and the total float reduces to 13-days, which is better.
Missing logic is a definite flag in most scheduling guidelines. It is possible though to strictly adhere to the missing logic guidelines and still have a high float schedule situation.
Insertion of constraints will reduce this total float, but is less than transparent. Schedule logic should be apparent to all stakeholders and not just a chosen few. Yes, full disclosure is an inherent trait of quality schedules.
A better solution considers possible alternative successors. The right successor not only avoids missing logic, it also reduces total float while maintaining a dynamic schedule. Considering the purpose and function of a deliverable may help to find that right successor that provides a logical, minimum float, and dynamic schedule.