How We Confuse Ourselves

How We Confuse Ourselves

 A fair amount of confusion and contradiction permeates among and across the literary works that deal with Project Time Management. Some of this confusion is inherent in the subject matter itself and cannot be avoided. But much of the confusion is entirely avoidable.

However, Confusion Avoidance requires both Will and Influence on the part of those who have a vested interest and recognized voice in precisely how Project Time Management is implemented on a given project Let’s take them separately:

  • Will: Avoiding confusion requires the will to act — one’s conscious, purposeful, deliberate, and steadfast desire to do so. It requires a level of diligence and mental focus that can at times become tiring, fatiguing, and overwhelming. It is not an assignment for the faint of heart.
  • Influence: On the other hand, influence refers to one’s standing in the broader Project Time Management community. It is not enough to want, or even have a great idea about how, to avoid confusion, for there to be a meaningful and lasting effect. The catalyst for change must have broad influence.

Change of all kinds, which afflict all projects, come from only two places: externally and internally.

  • External Sources of Change: External sources of change tend to be more formal, more disruptive, and less respectful of any preconceived, agreed-upon Project Time Management approach. Included within the External class of changes that can affect a single project are government regulations, contracts and other owner mandates. It also includes “best practices” and standards that are imposed by industry groups striving to standardize approaches to Project Management. [To their credit, these institutions recognize the inherent struggle to balance multi-project standardization (even across disparate industries) with the need for project-level autonomy.]
  • Internal Sources of Change: Factors within this class of change evolve out of the Project Execution process itself. Just as rough or calm seas are merely changing states of a consistently-present ocean beneath the ship, within this class of change is the ever-modulating state of operational conditions that constitutes the platform upon which the Project is built. In Construction there are many such commonly encountered changes: to name just a few; shifting weather patterns, subsurface ground conditions, the unpredictability and uncertainty of workforce availability, quality, or performance; the timely arrival to site of critical materials; the quality and timeliness of essential information; and so forth.

It should be obvious that our greatest opportunity to influence the elimination of confusion on projects rests mainly in the External Change realm. For it is here that decisions are made that directly affect how projects are, can be, and (quite often) must be managed. Both explicitly through edict, and implicitly through accepted, recommended, or adopted practices, those with sufficient Influence are forever invoking and forcing change. And whether they realize it or not — or intend it or not — their decisions can (and often do) easily and innocently confound otherwise straight-forward understandings.

To make my point, let me give an example. In this case, a recommended or accepted practice undercuts a previously well-functioning and readily understood concept: Earliest Dates. For the sake of those possibly unfamiliar with the term, Earliest Dates (Earliest Start and Earliest Finish) indicate the earliest possible time that an Activity in a schedule can conceivably start or finish, given prior logic, activities, and durations as depicted in the CPM schedule’s interwoven network of activities.

Confusion Example #1: Earliest Dates

In this example, we will consider how the rules (formula) for determining an activity’s Earliest Start have gradually changed over time — to the extent that now the word “earliest” no longer means “earliest.” Helpful to this discussion will be a simple two-activity logic diagram, as introduced in Figure 001.

F001

Figure 1: Truly Earliest Start Interpretation

This figure presents us with a lot of important and valuable information.

  1. We see that there are two activities: Washing Dishes and Drying Dishes.
  2. We can assume that there are at least two individuals involved, since one person could not be performing both dish-washing and dish-drying at the same time.
  3. We see that it is either ten o-clock, or that the “work” is triggered to commence at ten o’clock, and no earlier, as denoted by the Start-No-Earlier-Than (SNET) designation.
  4. We see that the washing of the dishes will require fifteen minutes, and included among the dirty items in the sink are some “pots with stubborn baked-on grease.”
  5. We see that the drying of the dishes, if performed without interruption, is estimated to need seven minutes to accomplish.
  6. We see that, for whatever the reasons, the dish-drying activity is called to start after two minutes of dish-washing have been performed.  Hence, we estimate Dry Dishes will commence at 10:02.
  7. We see that, if the Dry Dishes activity commence at 10:02, as shown, it will not be able to work uninterrupted for the balance of its duration, because the Wash Dishes activity is moving at a much slower pace. The person drying the dishes can expect to be standing idle longer than he/she will be actually drying dishes!
  8. We see from the red arrows and accompanying notation that the Dry Dishes activity enjoys seven minutes of discrete Total Float (up and above any other Total Float it might have thanks to prior and subsequent logic surrounding the activity).

In terms of interpretation and usage, the Project Manager for the ”project” that this “schedule” represents has all the information that she needs to make decisions about how to execute this work. In one extreme, she can call for the Dish Dryer to be ready to commence at 10:02 (as shown). At the other extreme, she can maximize Dish Dryer efficiency by postponing the Dish Drying activity until 10:09. The take-away, from this paragraph, is that the Project Manager has several options, and the schedule, as shown in Figure 001, provides a lot of useful information upon which to make informed decisions.

Now let’s look at Figure 002, in which we see the same two activities. But this time we are exposed to a scheduling technique used in the Critical Path Method called “Continuous Durations” or “Contiguous Durations.” I first encountered this option in P3 (Primavera Project Planner) back in the late 1990s, where a scheduler was given the option of choosing a toggle setting of either “Contiguous or Interruptible Durations.”  If I chose Interruptible, the result was what we just studied in Figure 001. But if I chose Contiguous, the result was as shown in Figure 002.

F002

Figure 2: Delayed Earliest Start

Let’s look at Figure 002. The first thing we notice, jumping out at us from the bar chart on the right, is that the Dry Dishes activity is now one solid, constant block of work. We see that it starts at 10:09, and proceeds without interruption to 10:16.

You should be aware of how P3 calculates the Earliest Start of 10:09, for such awareness gets to the point of this example, and this blog article. P3 essentially violates the decades-old understanding of a Forward Pass, and instead calculates backwards in order to determine the Earliest Start.

  • First, P3 determines the Earliest Finish for Dry Dishes, by honoring the Finish-to-Finish tie of one minute. Since Wash Dishes cannot finish any earlier than 10:15, then Dry Dishes cannot finish any earlier than 10:16.
  • Next, P3 subtracts the activity duration for Dry Dishes, which is seven minutes, and arrives at an Earliest Start of 10:09.

But there is just one small problems: 10:09 is not the earliest that Dish Drying can start!! Keep in mind that for the entire life of P3, which spans more than two decades, Primavera offered the scheduler the option to choose between Contiguous or Interruptible.  And in the early years, Interruptible was the default.

But somewhere along the way, Contiguous became the default setting. P3 maintained the option to change the setting, and most seasoned schedulers would do just that. But the default setting prevailed in schedules where the scheduler was not aware of the setting. Over time, the prevalence of Contiguous Durations in schedules was rampant.

Today, in P6, there is no option for Interruptible (to my knowledge). As a result, what P6 calls the Early Start (10:02) is not, in truth, the earliest that it could start if the artificial restriction imposed by the software setting wasn’t calling the shots.

Finally, just in case you are thinking that I am merely making a well-presented argument for what is essentially a technical point of no real-world importance, consider the loss of those precious seven days of Total Float. In Figure 001, we saw that the Dish Dryer knew that he or she had a period of thirteen minutes (10:02 through 10:15) to perform six minutes of Dish Drying. Just where those seven minutes of Total Float would occur was left to the “Project Team” (in this case the Dish Dryer and Dish Washer) to decide.

Defenders of Contiguous Durations insist that there is really no downside or negative effect to its practice, but I disagree.  First, from a practical perspective I don’t happen to believe that scheduling software should have greater influence over how a project is executed than the Project Manager does. Second, any time we squander precious Total Float we endanger the project.

Let’s expand our example one last time. Let’s look into that sink of dirty dish water and see what needs to be washed. As already noted there are several pots with stubborn baked-in drips. The Dish Washer will leave these time-consuming items for last. But, other than four such pots, the remainder of the sink’s contents is comprised of dishes and silverware, each piece of which should be easy to quickly wash.

Now let us suppose that among the dishes are two ornate serving platters that the Dessert Chef is waiting on, in order to prepare a dessert for a party of six who are just about finished with their entrées in the dining room. Suppose that this urgency is communicated to the Dish Washer, who makes a point of washing those two platters in the first minute of the Wash Dishes activity.

The Dish Dryer is not scheduled to start until the end of the second minute, which is intentional: it allows the majority of water runoff to drain from the platters, so that the Dryer’s towel doesn’t get sopping wet after only a few dishes. By extension, we can see that the Dish Dryer will have dried the important platters within the first minute of his work.

F003

Figure 3: Contrasting Interruptible and Contiguous Durations

 

All of this is shown in Figure 003, left side, where we see that the Dessert Chef should have the needed platters by 10:03. The Dessert Preparation activity has an eight-minute duration, and from what we see in the logic on the left, the desserts should hit the awaiting diners at 10:11.

But now look what happens if we are using a Contiguous Duration setting, with its delayed “Earliest” Start.  Since Dish Drying is not allowed to start until 10:09, then those needed platters are not available to the Dessert Chef until 10:10, and the desserts will not arrive to the diners before 10:18. Notice that the difference between the two dessert arrival times (10:11 and 10:18) is the same seven minutes of Total Float that we enjoyed with traditional Interruptible Durations, but that had been squandered under the Contiguous Duration approach.

Folks, while the above discussion may have opened your eyes to what are clearly deficiencies in the current (Contiguous Duration) practice, the main point I am trying to make is that, from an institutional level, an External Change has been imposed on the Project Time Management world that fundamentally redefines the word “earliest.”

I would suggest that a more appropriate term might be Preferred Start. Why do I say this? Because the underlying rationale for using Contiguous Durations in the first place is to facilitate greater efficiency in the performance of the restricted activity. It traces to a well-accepted truism that uninterrupted work is far more efficient than stand-around work.

But this is a management decision, is it not? The Contiguous Duration approach favors efficiency, which in its greatest consequence, improves project economics. But the Interruptible Duration favors the project’s temporal interests. In our example, it does this in two ways. First, it supports the earliest possible delivery of desserts to the customers. Second, it holds onto Total Float for subsequent, downstream Dish Drying activities.

This second point deserves a further brief discussion. When work, any work, is performed without Total Float, it is like working without a net! There is no allowance for anything to go wrong. Suppose the Dish Dryer accidentally drops the towel into the sink of dirty dish water, and that fresh towels are downstairs in the laundry room. Or suppose the Dish Dryer accidently cuts his hand on a knife and needs to take a few minutes to contain the bleeding and bandage the wound, before returning to duty.

In the Interruptible Duration scenario there would be ample Total Float to absorb either ”incident.” But in the Contiguous Duration scenario, there would be no Total Float. And so, if we assume that either incident would delay Dish Drying by three minutes, then dish (under the Contiguous Duration setting) drying will not finish until 10:19. It would be a one-for-one delay to the end of Dish Drying, because there would be no Total Float available to absorb unexpected “surprises.”

Software Settings for Durations Just one Example

The above example of the effects of two alternative software settings for determining Earliest Dates is just one example of how External Influences can alter the meaning of concepts that were previous understood, rational, and effective. Sadly, there are many, many others similar instances of policies affecting what had been common sense understandings. Which examples can you think of? Leave your comment below?

This blog is long enough, so I won’t go into other examples. Suffice it to say that the broadly-accepted objectives and goals of Project Time Management as a discipline might be better served if those with the greatest Influence were to be as diligent as possible when checking that new approaches and innovations do not unintentionally undermine, contradict, or even reverse well-established understandings.

8 Comments to How We Confuse Ourselves

  1. Sue Backiel says:

    Informative blog regarding Contiguous and Interruptible durations. I agree, the Project Manager should have the influence over how a project is executed. Very interesting comments supporting both Contiguous and Interruptible durations.

  2. Dave says:

    As always an interesting blog explaining how scheduling confusion can happen and how to avoid it. The blog discusses change and particularly how external sources of change can be influenced to eliminate confusion on projects.

    I found the explanation of the CPM Scheduling techniques “Continuous or Contiguous Durations and Interruptible Durations” intriguing. Having never used Primavera Scheduling Software the information provided about the two previously mentioned durations was beneficial. The explanation of the forward pass used by P3 change to the earliest date calculation by actually subtracting an activity duration to calculate an early start date, was a unique solution by a software vendor and eye opening to say the least.

    By the vendor eliminating the setting option for Interruptible Durations means that recent schedulers are unaware there was ever an option to choose a different type of scheduling duration through the use of a toggle switch. Disallowing this process means that alternate what ifs scenarios will never be contemplated in these situations.

  3. Michael Neal says:

    [C] Early start seems to be determined as to whether the activity is contiguous or interruptable. More leeway is given if the activity interruptable since in theory it could start just after its pred. I like the comments that are brought forth.

  4. MurrayBWoolf says:

    Novzar, underpinning what you correctly characterize as a “generally accepted tradition that continguous durations are more efficient” is an assumption that resource efficiency is necessarily a priority of every Project Team.

    My feeling is that this decision should be made by each Project Team, and not bby sweeping, universal “best practices,” industry standards, or project management software algorithms.

    I offer the same resistance to other structured efforts that, nonetheless noble and well-intended, make the assumption that resource efficiency is a hallmark of a well-executed and efficient project. As one example, consider Location-Based Scheduling (aka, Line of Balance, Linear Scheduling) which is an extremely effective method of finding the mosty efficient use of resources for a given area.

    The problem is that localized efficiency often comes at the price of more global inefficiency. I often give the example of road construction on a rural highway. You are driving along at 55 mph when suddenly, as you come over a hill, you are faced with a sea of brake lights in front of you. You are on a two-lane road, and your lane is at a complete standstill.

    As you peer ahead you notice a flagman holding a “STOP” sign in front of the first car in your lane. To your left, a slow-moving stream of opposing traffic passes by your door window. Eventually the stream stops, the flagman rolls his fingers, and the sign now reads “PROCEED SLOWLY.”

    As you clear the flagman, you enter the Construction Zone. You see that one of the two lanes is now under construction, and both directions of traffic are sharing just one lane.

    From the perspective of the Construction Area, if you were to look down from a helicopter, you would see impressive efficiency. The major disruption to the construction, the through traffic, has been “managed” by way of two flagmen and their signs. Meanwhile, the construction work itself proceeds unabated.

    But if you were to ascend in your helicopter to a higher elevation, and pan out across the distance of miles, you would see stalled traffic extending several miles in both directions!

    From a global perspective, there is anything but efficiency. Many people will arrive late to their destinations, and the domino effect at those locations may last hours.

    The same happens when we maximize efficiency in a specific area without consideration of how overall work flow is impacted. As you may know, ICS-Global is the birthplace of Momentum Management, which uses Performance Intensity to monitor, report, and help manage overall Project Momentum.

    This is a fascinating area of Project Time Management that still remains off the “traditional” Project Management radar screen. In time, though, I am confident that Project Managers around the globe will give greater consideration to the “Big Picture.”

    Using Interruptible Durations is one step in that direction. I can give you tangible examples of how Contiguous Durations actually extend the Critical Path! So, in that light, just how important is localized resource efficiency?

  5. Connie Bremer says:

    Convincing argument in favor of maintaining an interruptible duration over a contiguous duration activity. This blog was certainly thought provoking. I also appreciate the experience and knowledge shared by other comments.

  6. Thomas Long says:

    Ultimately the CPM schedule is a tool for communication and project control. The art of scheduling is not about what you put in it, but what you leave out. Just as Aris Venetikidis discusses in his talk about creating maps that work http://www.ted.com/talks/aris_venetikidis_making_sense_of_maps.html , a CPM schedule is a tool to proof out the most efficient and maybe fastest route to your end. When looking at a transit map, your chosen path may include risk factors. For example, you might take a “longer” route because it has less transfers and thus less inherent risks in delays due to waiting on the next train.

    In tower construction I always break up the activities by trade and by floor. Then the logic is put in so that floors are “released” to a single trade, capitalizing on the inherent efficiency in allowing a trade free run of the work area. They can’t keep it for long. That would mean that they have a FS relationship between following trades and the next floor. In reality, the trade works in a performance curve: a slow start-up followed by high productivity, then slowing as backtracking occurs to fix deficiencies or tie up loose ends. Work is performed by teams, and the team composition is in constant flux. In tower construction, fission of the team occurs as some members move ahead to advance the work on the next floor as the rest of the team stays behind to wrap things up. Then the team congeals again to perform the high productivity work and the cycle continues until the last floor is reached.

    There is also the law of diminishing returns, sure you can pull more teams in, or stack the trades, but this often leads to inefficiencies as the secondary teams have lower productivity. When I walk into a construction site I look at the workflow to quickly asses how well the job is going. If one trade is dominating an area, or too many trades are working on top of each other I can easily assume that the project manager has lost control.

    FS relationships are the easiest way to sequester float in construction. If I were to build a schedule with nothing but FS relationships and filtered by trade I would see a jagged, discontinuous workflow. When the schedule is actualized there will be a lot of out of sequence to resolve as the subcontractor mitigates the mob-demob cycle with soft or late starts.

    Often when I have meetings with the owner they are looking at the early start dates. Maybe I have worked with subcontractors too much, but I first look at the finish dates. Like a fish that grows to fill its bowl, subcontractors tend to expand their work to fill the entire temporal space between early start and late finish. This isn’t just a defect, but a function of temporal osmosis driven by the subcontractor’s need to balance manpower with efficiency.

    Contiguous or interruptible settings couch the discussion in a theoretical diametric and imposes an unrealistic constraint. The art of scheduling demands that the logic and granularity of the schedule are adjusted to allow it to perform the functions of a high level communication tool and logical proofing means and methods. I have never pushed the automatic leveling tool button on many software programs and instead do it all by hand. If I am lucky enough to build the schedule I am updating, I have already built in the flexibility to do this without impacting the end date. Until scheduling software undergoes a transformation from its current 2D state to a 3D(such as line of balance) as BIM software has done to drafting, the communication gap between the theoretical schedule and the reality of work performance will continue.

    • MurrayBWoolf says:

      Thomas, your comments reflect a wealth of knowledge and experience, surely acquired through years on real projects. In all that you said, I could not agree more. Your conclusion, that “Contiguous or interruptible settings couch the discussion in a theoretical diametric and imposes an unrealistic constraint,” is particularly profound. The subtle message — written between the lines — in my blog article is that all of our best methods and techniques for modeling Project Execution Strategy are necessarily limited by the conventions and software limitations of human expression. Real Projects are dynamic, and far more than even three-dimensional. They are complex and overlapping and symbiotic.

      Whether we use Finish-to-Start ties, or choose to overlap (Start-to-Start and Finish-to-Finish) the activities in our Baseline Schedule, these are just placeholders for what will ultimately happen in Real Life. Our sketches (logic) only superficially reflect our intentions. They can never fully chart the course we will ultimately take. In the end, it is up to the Project Team, in Real Time, to make daily (and even hourly) decisions as to how the project will be executed.

      And it is to this end that I passionately argue that the Project Team, and in particular the Project Manager, never be denied the opportunity to make those in-the-moment calls. That is why I favor Continguous Durations, because they yield the truly earliest possible dates for activity start, and do not surrender or squander Total Float without offsetting gains. As Novzar notes in his brief comment, “there is a generally accepted tradition that contiguous duration is more efficient.” This may be true, that working continously is more resource-efficient. But at the end of the day, is it not for the Project Team to decide whether Labor Efficiency is the primary objective? What if expediency, timeliness, quality, safety, or some other criterion is the greater concern and objective?

  7. Novzar Dastoor says:

    There is a generally accepted tradition that contiguous duration is more efficient; but your illustration very vividly puts forward the case for merits of interruptible durations.
    Thanks for the eye opener.

Leave a Reply to Novzar Dastoor Cancel reply