In one of my favorite Gene Hackman films, Heist, there is a scene where his character is explaining how he solved a problem. His client (actually, a mobster he is planning to rip off) says, "you're a pretty smart guy." When Hackman's character declines the compliment, the mobster replies, "then how did you figure it out?" Hackman's response:

"I said to myself, 'what would a smart guy do?'"

I love that line. Unfortunately, I recently forgot that lesson. Let's set the stage. I am updating a new Primavera P6 schedule for the first time. I have a change order modification that has been added, and there is also progress on this new activity.

The client told me he is 40 percent complete with the modification. So I type in that percentage. And I move on with the update. But while checking my customized earnings report I realize there is no Earned Value for the new activity. But the activity is in fact showing 40% complete.

Okay...

Now, perhaps you are thinking I forgot to apply an actual start date. Well, no, it would not be possible to show a percent complete without first entering an actual start date. I should also mention that I am using the Physical % Complete type so that I can enter whatever percentage I want.

Hmm. Perhaps there is something wrong with the way I cost-loaded the activity. Nope, it has a budgeted total value. And I am using the same resource that all other cost-loaded activities have been assigned in this particular schedule.

What about my Earned Value settings? I almost never change them, but I verified that Performance Percent Complete = Activity % Complete (Admin > Admin Preferences > Earned Value). In the construction industry that is pretty much the setting that everyone uses. The point being that the activity percent complete determines Earned Value.

Yet here I am, still looking at zero dollars earned and 40% complete.

Aha, must be the activity type! We know that certain activity types like Level of Effort and WBS Summary behave differently than other types of activities.

Sorry - the new activity is Task Dependent. Nothing to see here folks. And if you were thinking it was a milestone, keep in mind that milestones cannot be assigned resources due to their zero durations.

What would Gene Hackman do?

He'd probably say something like, "hey buddy, if Earned Value is working on EVERY other activity in the schedule then there has to be something different about this one."

But it's not different, it's just new.

Oops...

Yes, it is a new activity! And where does Earned Value come from?

The baseline to the current schedule!

Under the Settings tab in the Project Details window there are two options for calculating Earned Value: Project baseline or User's primary baseline.

My schedule was set to Project baseline, and my Project baseline is the original plan. The new activity doesn't exist in the original plan so Earned Value is automatically zero.

I knew this, of course, but it is easy to forget when you are in a hurry. And it seems rather odd that users who never create baselines for the current project are immune to this potential problem. Because when there are no baselines the current schedule IS the baseline!

This is in my mind one of the most counter-intuitive aspects of Primavera P6. Earned Value is not always based on the current schedule. Why would I want to base my earnings on something other than the current schedule? If the budget has changed it seems logical to pay off the new budget.

Now, I prefer not to change the value of an existing activity for pretty much the same reason I don't change the original durations. I want to preserve the old values for future reference. If the duration or budget is changed by a modification I will add a new activity for that purpose.

I have had a few projects, however, where we re-balanced the activity budgets. Assigning any baseline other than the current schedule would clearly cause problems. Which brings up another concern: the Earned Value for a project can be okay one day and off the next if a different baseline is assigned. Fun!

When I mention this during training sessions I often hear someone with a little Primavera P6 experience say something like, "never happens to me". Yes, that could be true. If you don't typically create baselines then by default your baseline is the current schedule. Lucky you!

Those of us who utilize baselines on a regular basis have to be more careful. And it gets rather complicated when there is more than one baseline attached to the current schedule. Re-assigning a baseline can change Earned Value. Oh well. I'll just keep channeling Gene Hackman.

Sep

The Resource I Never Use

Categories: P6 EPPM, P6 Professional, P6 Tricks, Primavera Resources

Tags: Primavera P6, Resources

Comments Off

by Bill Pepoon

Resources add "weight" to a schedule. That is to say, resources can tell us the effort - in costs and/or hours - required to perform a task. Once resources have been added to an activity we now realize that tasks with similar durations are not so similar when the effort is considered.

Back in the early 1980s I would sometimes be asked by owner representatives to explain how many people were required each day in order to stay on schedule. I was using a proprietary scheduling program running on a mainframe computer in those days. It was a nice little program based on the Activity-on-Arrow (AOA) method of scheduling. But it was not capable of doing much with resources other than assign a cost value to each task.

In a parallel scheduling universe, however, Primavera Systems was offering the ability to budget and track both costs and units. The firm I worked threw in the towel in the face of a better product and switched over fully to Primavera in 1987. That was a good decision but at the time there were other scheduling programs to consider. Proximity might have helped. Our company's headquarters were across the Delaware River from Primavera Systems.

It might surprise schedulers who nearly always resource-load their schedules that a large percentage of companies - at least in the United States - rarely do so. The reasoning is often that it takes too much time or exposes confidential information. I know a lot of contractors do not want anyone knowing how many labor hours were figured into their projects.

Not knowing how many labor hours are required by the project once led to an awkward exchange with an opposing attorney during my testimony in a construction dispute:

Attorney: "Mr. Pepoon, how many people would have been required to perform this task that we are looking at right now?"

Me: "Enough people to perform the task within the planned duration."

Attorney: "But how many people would that be, exactly?"

Me: "The proper number to perform the task within the allotted time."

It sounded like a bad comedy skit, but of course I could not answer a perfectly relevant question because my client did not resource-load the schedule. And the reality is that when a subcontractor is responsible for the work the general contractor usually can't verify the duration either. He simply trusts that the subcontractor knows best.

The people performing the task would presumably know how long it takes. But no one else can verify this without knowing the required effort. Then why does this still happen? Well, general contractors pass on a lot of the risk (i.e. scope of work) to subcontractors and can always "sue the bastards" if they fail to meet the schedule. Of course, owners will then sue the general contractors for not controlling their subcontractors. Nobody will be happy.

Still, there is one resource I refuse to use - material. Anyone familiar with Primavera P6 understands that there are three types of resources: labor, nonlabor, and material. Nonlabor is mostly used for equipment, but I have another use for it as well. The trouble with a material resource is that Primavera P6 only tracks costs, not units. With labor and nonlabor resources the choices are to track units, costs, or both.

From my perspective, tracking material could be very useful. If I have moved 50 cubic yards of dirt and there is a total of 100 cubic yards, then certainly one-half of the work is complete. (I should also consider how long it took to complete one-half of the work, because productivity is important as well). This is production-based scheduling, but it is simply not possible with a material resource.

Thankfully, there is a workaround. Instead of using a material resource, why not use nonlabor instead? If it's not labor it can be pretty much anything, right? So nonlabor can be plumbing fixtures, dirt, gypsum board, conduit, or anything else we want to track. I have even used a nonlabor resource to track drawings being produced by the design team.

Both labor and nonlabor resources use "units" as the nominal description. Only material resources can have other labels, but we can easily create our own definitions.

For plumbing fixtures the unit can be "each", cubic yards for dirt, square feet for gypsum board, and linear feet for conduit. Each type of material will of course require its own unique nonlabor resource. I also indicate the real unit of measure in the description, such as "Dirt (Cubic Yards)" to emphasize the point.

I also do not show the unit label for these nonlabor resources because Primavera P6 would put "hours" on all of my nonlabor resources (Edit > User Preferences > Time Units). One other thing. When printing resource charts I do not lump nonlabor resources together because they are not the same materials at all, but this would be true for equipment as well.

Ultimately, we must decide what level of detail is required when adding resources. Minor amounts of material may not warrant much scrutiny. The desired output determines the level of input required. For example, do we need to track different sizes of piping or can they be lumped together? Too much detail is often as bad as not enough detail.

Dec

Why vs. How

Categories: Uncategorized

Tags: Earned Value, Primavera P6

Comments Off

by Bill Pepoon

Learning how to do something is not the same as learning why something should be done. Most of us have been told at various times by a supervisor or boss to do something without any explanation. There may be some urgency in the request and for that reason alone there is not much time to explain why it needs to be done. Still, I will never forget the time I questioned a senior partner’s instructions only to be told, “because I sign your paychecks”.

Even so, when you cannot be bothered to explain assignments to your employees, those employees are not going to learn how to think for themselves. Those employees will also become bored with their jobs. Eventually they will go somewhere else where work is more meaningful. In 2018 this was especially true in the United States, where record numbers of people quit their jobs (3.4 million in April 2018 alone).

As an instructor, I am fully aware that “why” is just as important as “how”. For example, it is one thing to explain how to set up multiple calendars in Primavera P6. Also very important is “why” this can cause issues with activity dates and float values.

Explaining “how” is fairly straightforward. “Why” takes many years of practical experience. Examples taken from real projects give depth to the explanations. Moreover, Primavera P6 is used in many industries and for that reason there is quite a lot of flexibility built into the program. Best practices for using Primavera P6 likewise vary by industry.

For example, consider the different ways to calculate Performance Percent Complete, which in turn determines Earned Value. In the construction industry we would rarely consider using any setting other than Performance Percent Complete = Activity Percent Complete.

This is in fact so common that I doubt most schedulers in the construction industry have ever bothered to consider Performance Percent Complete. However, there is actually a very good reason for showing Performance Percent Complete on a schedule. Unlike Activity Percent Complete, Performance Percent Complete rolls up, meaning we can show an overall percent complete for the project. Activity Percent Complete only displays on activities.

In the construction industry we like to take credit for every little bit of work when it comes to percent complete. If necessary, we can display up to two decimal places as well (more common when Level of Effort activities or Activity Steps are deployed). Now, I would also argue that using a schedule as a payment tool tends to interfere with the more noble purpose of scheduling; and if the schedule is not cost loaded I could not care less about the percent complete. Remaining duration is far more important!

But in other industries there are valid reasons for tracking Performance Percent Complete differently. The options are:

Use WBS Milestones
0/100 % Complete
50/50 % Complete
Custom % Complete

The 50/50 % Complete method means that the task is 50% complete as soon as an Actual Start date is recorded. I think of the classic movies where someone tears up a large denomination bill and hands half of it to the hired gun. When the job is completed the hired gun gets the other half. So we are not expecting him to report any further progress until he is done. There is no in-between.

The Custom % Complete method is flexible enough that I could see the benefit even in the construction industry. The typical task should be of a rather short duration – commonly 20 working days or less (with procurement tasks being the exception) or slightly less than one month. No task should therefore stay open very long. With this in mind, why not make progress reporting simple? The custom method allows users to set intervals such as 25%, 50%, 75% and 100%.

Quite frankly, if everything goes according to plan at most we would be recording the task as partially complete during the current month and 100% complete the following month. Activities should only be in progress for one month if the maximum duration is 20 working days! The owner should not be overly concerned with paying 50% vs. 25% or 75% vs. 50% this month knowing he will owe the balance the following month anyway.

Activity durations are another example. In the construction industry it is generally true (and often a contractual requirement) that the minimum duration is one day and all durations are integers (i.e. whole numbers). Even a project meeting must be assumed to last all day because of this convention – and keeping in mind that the maximum duration will still be roughly 20 working days. This means that lag durations should also be whole numbers. Personally, I think it makes everything much easier because I know that if I see a duration (including Total Float) that is not an integer, something is wrong with my schedule.

Partial durations tend to creep in during the update process, especially when the scheduler is using Duration % Complete as the percent complete method. Trying to force a particular percentage (such as 50% on an odd-duration task) often results in a remaining duration that is not an integer. For this reason I input the remaining duration when using Duration % Complete and let the percentage be whatever it has to be. On a schedule that is not cost-loaded the percentage is of little importance to me anyway.

In other industries the activity durations might be quite small. The minimum duration supported by Primavera P6 is one minute, if you were wondering. On a refinery shutdown that is costing the owner millions of dollars of revenue each day we clearly need to track progress on a more granular level. These schedules tend to be much larger in terms of the number of activities even though the project duration may seem rather short. Time costs a lot of money!

Likewise, even the activity relationships tend to change when the durations are very short. There is simply less need to overlap tasks when durations are in minutes or a few hours at best. Finish-to-Start durations become the norm. Not having to worry about relationship types (or lags, for that matter) makes up for the size of the schedules to some extent. Nevertheless, when tracking progress it is rarely necessary to record the exact time of day that a task is performed. Done is done.

So we can see that understanding “why” is perhaps more important than “how”. As a Primavera instructor with decades of practical scheduling experience I can think of many examples and options for how to schedule using Primavera P6 and why some options within the program might be better than others depending on the industry or the situation. There is more to Primavera P6 than just menus and buttons.

Jan

The P6 Omnibus File

Categories: Baselines, P6 EPPM, P6 Optional Client, P6 Professional, P6 Tricks, P6 Web, Primavera layouts

Tags: Baselines, Layouts, P6 Optional Client, P6 Web, Primavera P6, XML

Comments Off

by Bill Pepoon

Omnibus (adjective)

Comprising several items.

The most common Primavera export file is “XER” which harks back to the company that developed this enterprise scheduling software in the first place: Eagle Ray. Primavera Systems bought Eagle Ray, and then Oracle bought Primavera Systems. The very familiar “XER” file format in fact stands for “eXport Eagle Ray”.

XER files can only export project data associated with the current project, or all resources or all roles. For someone who needs to transfer all resources or roles from one database to another the XER file can be very useful, but I find this to be a rather esoteric function for the vast majority of users.

But there is another Primavera export file (XML) that can do so much more:

Export all project layouts associated with the current project
Export all (or some) baselines associated with the current project
Import into any other version of Primavera P6

This might also encourage you to create project layouts. By default, new layouts are user-specific and can therefore be applied to any schedule to which the user has access. Project layouts are only available to the associated project (or a copy of that project) which is desirable when the layout has specific features (such as a grouping or filter) that would not be applicable to other projects. The header or footer might likewise contain wording that is specific to one project.

Primavera P6 EPPM users are more accustomed to this method of importing files because the P6 Web interface only supports XML file imports and exports. However, P6 Professional Client (sometimes referred to as P6 Optional Client) can be used to import XER files into a P6 EPPM database. Confusing, yes, but P6 EPPM databases can be accessed via a Web or desktop interface.

The following screenshots show the process for exporting P6 XML files. Keep in mind, you are not required to export any baselines and can also choose which ones to export. Likewise, you do not have to export project layouts:

Here is the sequence for importing P6 files. Notice that we can choose which baselines should be imported:

Pretty simple!

Hopefully you will not think of me as a hypocrite if I admit to sending XER files on a regular basis. But as a consultant I do not need to keep sending baselines to clients if they have already received those files previously. For example, if I sent my client the third update last month it is somewhat redundant to send them the fourth update this month with the third update as a baseline. I also do not need to keep sending project layouts unless I have recently created new ones.

Nevertheless, for the recipient, the XML file has everything needed to view the current schedule and make comparisons to previous versions of the schedule. I find that for my construction claims work it is a great way to transmit my entire analysis of a delay to the client. The only downside might be that XML files are not nearly as compact as the text-based XER files. Roughly speaking, XML files tend to be about ten times larger, which in some cases might exceed the maximum file size for email attachments. Not surprisingly, it also takes longer to export and import XML files.

Dec

Why Make Open-Ended Activities Critical?

Categories: Constraints, Critical Path, P6 EPPM, P6 Professional, P6 Tricks, Schedule Options

Tags: Constraints, Longest Path, open ends, Primavera P6, schedule options

Comments Off

by Bill Pepoon

Primavera P6 Professional is obviously a very powerful scheduling program so naturally some of its features exceed the needs of the typical project. I have consulted on projects that span as little as 35 hours to as many as 50 years. Different industries have unique requirements for their schedules as well. Primavera P6 is designed to handle a wide variety of projects. Today I would like to address my reasons for using a feature in Primavera P6 that is rarely used by the typical scheduler: Make Open-Ended Activities Critical. You will find this feature under Schedule Options (Tools > Schedule > Options).

The concept of making open-ended activities critical was introduced many years ago in Primavera P3. And for the longest time I dismissed it as a quirky feature surely not applicable to me. After all, why would I promote an activity to critical status solely because it is missing a successor? That seems akin to me declaring myself the winner of a contest that no one else entered.

Some of my colleagues back in the 1980s figured it was an en easy way to identify activities that should not be critical. Okay, that seems backwards, but the idea was that if some task showed up as critical that did not seem “right” the scheduler would investigate further. But Primavera P3 had a report similar to what we call the Schedule Log in Primavera P6 that was a more definitive (and easier) way of identifying open ends in the schedule. With this in mind, making activities that are missing a successor critical did not seem like the best approach to finding open ends.

I started my scheduling career working solely on construction projects so my viewpoints towards Primavera P3 were based on a single industry. Years later, when I began consulting on a wide-variety of projects I realized another purpose for Make Open-Ended Activities Critical. For example, let’s say I am a manufacturer trying to track progress on several production lines. I would like each production line to have its own critical path, or what Primavera P6 refers to as Longest Path. But this would entail making each production line a separate schedule.

It would be a lot easier to track progress, however, if I incorporated all of the production lines into one schedule. That way I would not have to keep opening up another schedule to track progress. And P3 only allowed four schedules to be opened simultaneously. (Some of you P3 users undoubtedly remember the master and sub-project concept from P3 which was another alternative to what I am describing). But how could each production line have its own critical path without creating separate projects?

Yep. Make Open-Ended Activities Critical. See, by not linking the production lines to each other they would all have an open end at the end of their sequence. So every production line now has critical activities. The float values are not based on the longest of all the production lines. Each production line has activities with zero Total Float. Problem solved!

The concept I just described works the same in Primavera P6 as in Primavera P3. You might be thinking that each production line could have a constraint on the final activity to create zero Total Float and then link the final activity in each production line to some final milestone. Yes, that will work too, smarty pants! It also means that additional open ends do not need to be introduced into the schedule.

Another reason for Make Open-Ended Activities Critical is relevant to projects in any industry. One of my clients currently expects his project to finish early. Owners often don’t allow the original plan to show an early completion date because it might become the basis of a claim (“I planned to finish early and you stopped me”). In this situation the owner allowed early completion. So my client inserted two finish milestones in his schedule: “Projected Finish” and “Final Completion”. The latter milestone matched the contractual finish date.

The “Projected Finish” milestone had no constraint since the date could obviously slip without any ramifications. The “Final Completion” milestone had a Finish On constraint (Mandatory Finish also works) so the date could not move at all. But as you might have guessed, this meant that only the “Final Completion” activity appeared as critical in the schedule. The earlier milestone and all of the activities linked to it (directly or indirectly) carried Total Float values based on the later milestone.

A critical path consisting of just one (the last) activity) would obviously be acceptable to no one. But the solution was quite clear to me. While my client had linked the “Projected Finish” milestone to the “Final Completion” milestone (to avoid unnecessary open ends) we need the “Projected Finish” milestone to have no successor. Then, by choosing Make Open-Ended Activities Critical in the Schedule Options the Longest Path of activities leading up to “Final Completion” all had zero Total Float. Bingo.

Below is how the schedule looked with “Projected Finish” linked to “Contractual Finish”. Not having open ends means there is no logical critical path. Also, the “Projected Finish” milestone is a non-driving predecessor to the “Contractual Finish” milestone, as evidenced by the dotted relationship line:

By deleting the relationship between “Projected Completion” and “Final Completion” and choosing to Make Open-Ended Activities Critical under Schedule Options, the activities leading up to “Projected Completion” are now critical:

Make Open-Ended Activities Critical is not always a necessary feature, but as you can see, it certainly does have a purpose.

Aug

The Pit Stop Relationship

Categories: Activity Relationships, P6 EPPM, P6 Professional, P6 Tricks

Tags: Activity Relationships, Constraints, float, Primavera Contractor, Primavera P6

Comments Off

by Bill Pepoon

During our advanced Primavera P6 classes I like to tease my students with the “weirdo” activity relationship and challenge them to find a practical situation for applying it. This relationship type is not available in scheduling programs like Microsoft Project so most of my students have never seen it before.

I am talking about the Start to Finish relationship, whereby the predecessor must start before the successor can finish. Unless a very large lag is used, the successor will start before the predecessors starts. Think about that for a moment. The successor starts first! The following illustration will make that statement rather obvious:

The predecessor (yellow bar) starts on the same day that its successor (green bar) finishes. If we add a lag duration the yellow bar will start before its successor finishes, so they will overlap. For example, a one-day lag would cause the predecessor to start the day before the successor finishes.

Well, you can probably see why we do not unleash this relationship type on beginners!

Other than showing off, one might wonder why the Start to Finish relationship is used at all. My favorite example is the pit crew that services a race car. The pit crew is the predecessor to the car arriving in the pits (the successor).

No, that is not a typo. I am saying the pit crew is the predecessor. Put another way, the race car is not supposed to pull into the pit lane until the pit crew is ready. The pit crew is therefore logically the predecessor because they, not the car, must be ready first.

At this point I am sure you are thinking, “why not make the race car the predecessor to the pit crew and use a Finish to Start relationship?” The issue is that we do not want the car to arrive early. So it is logical to say that the successor’s start date should determine the predecessor’s finish date. Hence, Start to Finish.

While I wish that race teams used Primavera software (official scheduler for Ferrari would be a very cool job) the reality is that one of my clients was already using this relationship type before taking my class. They work at a nuclear submarine facility on the East Coast. Similar to my example, they do not want the sub to show up until they are ready to perform maintenance. So maintenance is the predecessor.

The downside to this relationship type is that you can end up with some strange looking float values, as the backwards pass algorithm seems to get confused by what almost seems like reverse logic. We expect relationship lines to always be pointing to the right, after all. Other activities tied to the ones with the Start to Finish relationship can likewise behave rather strangely.

An alternative would be to use a Finish to Start relationship and put an “As Late As Possible” (ALAP) constraint on the predecessor. This way, the race car can be the predecessor but it will not finish until the pit crew is ready. Schedulers often use this constraint to avoid having materials or equipment delivered to the jobsite too early. Sort of like Just In Time manufacturing.

Perhaps it comes down to personal preferences. Some people dislike using Start to Finish relationships and others dislike using constraints. Owners in particular view constraints as an artificial device to sequester float. The contractor is in theory suppressing float on certain activities to keep the owner from using it. Fine, but when the door hardware shows up a year early perhaps they will reconsider!

Personally, I prefer to use the ALAP constraint, albeit with discretion. A side-effect of this particular constraint is that the predecessor(s) will not be driving. This can confuse some people. While other constraints sometimes have the same effect, this always happens with the ALAP constraint.

Try the Start to Finish relationship for yourself and see what happens!

Mar

A Project That Will Save Lives

Categories: P6 Professional, Primavera Training

Tags: Golden Gate Bridge, Primavera P6, San Francisco

Comments Off

by Bill Pepoon

It is not very often that we can say that a project will save lives. But a project in San Francisco that started in February of 2017 is intended to do just that. The project has a rather long title: Golden Gate Bridge Physical Suicide Deterrent System and Wind Retrofit Project.

Our firm provided a 3-day Primavera P6 training session last week to employees of the Golden Gate Bridge Highway Transportation District and the general contractor, Shimmick/Danny’s Joint Venture. After the first day of training I drove up to an observation area on the Marin County side of the bridge to take the photo you see here.

Stainless steel netting will be strung along both sides of the 1.7 mile long bridge. The netting itself will be 25 feet below the bridge deck to avoid obstructing the view from what is of course one of the most iconic bridges in the world. Yet the sad reality is that roughly 1,600 people have chosen this location to end their lives since the bridge opened in 1937.

Suicides can be prevented. A 1978 study by University of California – Berkeley clinical psychologist Richard Seiden tracked 515 people who were restrained from jumping off the Golden Gate Bridge between 1937 and 1971. Years later, 94 percent were either still alive or had died of natural causes.

This $204 M project is expected to take nearly four years to complete. Existing paint will have to be removed, and contained. Crews will be working from the underside of the bridge some 250 feet above the water. Field measurements will have to be taken for all of the struts that support the netting. Altogether, enough netting to cover seven football fields will be hung.

Various groups and individuals lobbied decades for some sort of deterrent system. Yes, this is an expensive project but the original 4-foot high railing was hardly any deterrent at all. As one suicide note stated, “Why do you make it so easy?”

Indeed.

Feb

Are You a Pigeon?

Categories: Activity Steps, Critical Path, P6 EPPM, P6 Professional, P6 Tricks, P6 Web, Primavera Contractor

Tags: Activity Steps, Critical Path, Primavera P6, Retained Logic

Comments Off

by Bill Pepoon

Pigeons mystified Charles Darwin. He mentions them in the very first chapter of On the Origin of Species:

“The diversity of breeds of pigeons is truly astonishing. If one compares the English Messenger with the short-faced Culbutant, one is struck by the enormous difference of their beak, resulting in corresponding differences in the skull.”

There are as many species of pigeons as there are of dogs or cats. But today we are focusing on carrier pigeons and their cousins, homing pigeons. And what a noble history! Noah was the first to recognize a pigeon’s ability to carry a message. The Romans used pigeons to transmit the results of chariot races. Genghis Kahn and Charlemagne relied upon pigeons to carry messages to troops, as did the French in the Vietnam War.

Even today, pigeons are used to carry blood samples in remote parts of England and France. Not surprisingly, pigeons have also been dragged into nefarious duties. Drug traffickers in Afghanistan and Pakistan have utilized flocks of pigeons to deliver heroin, each one carrying 10 grams.

But let’s be honest. If rats could fly we might have used them instead. Pigeons deliver a message – right or wrong – from one party to another. They don’t (can’t) read the message or vouch for its accuracy.

Don’t be a pigeon. Too many schedulers are simply delivering information that is incorrect or incomplete. They fail to consider whether the current critical path makes sense, ignore activities that clearly should have had progress by now, and fail to analyze the potential impact of unforeseen events.

Not all information can be easily verified, of course. Unless the scheduler is posted to the field he or she can not independently verify actual dates, installed quantities, percent completes and the like. Still, there are times when the reported progress makes little sense, like my client who said he had started plumbing fixtures on the third floor of the building at a time when the structural steel to the second floor was being erected (he was taking credit for plumbing fixtures that had been delivered to the jobsite, but we had a procurement activity for that scenario).

That story involving the plumbing fixtures happened more than 25 years ago. Schedulers have long memories! More recently, I have been reviewing schedules on a 34-story apartment tower in the midwest for the owner. Ironically, plumbing is once again the issue.

A couple of months ago, installation of water heaters on nearly every floor showed up on the critical path. And with one week scheduled for installation per floor, we had water heaters occupying about seven months of the remaining critical path.

I understood how the water heaters wound up on the critical path. The contractor had added activity relationships between water heaters on each floor – something we were calling “crew restraints” way back in the early 1980s.

Here is the funny part. When I filed my report with the owner, the contractor accused me of modifying his schedule update! He had no idea these crew restraints existed or why they had been added. The reality is that the crew restraints had always been in the schedule but due to better progress on other paths, this water heater path had now been exposed.

If you are wondering why water heaters were such a concern, the fact is that most of the interior work (drywall, painting, etc.) had similar crew restraints. But only the water heaters assumed one floor at a time. I have no idea why. It is clearly a very conservative assumption, and as many of you who read my blog on a regular basis already know, I prefer The Schedule That Can be Beat.

Nevertheless, this project is behind schedule and letting the water heaters control the project completion date is not acceptable. Yet this trend continued for another month. Finally, the contractor revised the logic so that water heaters overlapped on some floors. The critical path is now starting to make a lot more sense.

That’s just it. A good scheduler should know what belongs on the critical path. Even when I am still building the baseline schedule I have an idea of what will be critical. During the monthly updates I believe I know what should be critical as well if my logic is still valid and the contractor makes sufficient progress.

Submittals tend to be ignored by the contractor because so many of them have large float values. Whenever a submittal pops up on the critical path after I make a preliminary run of the monthly update I get very suspicious. Most of the time, the contractor’s memory improves dramatically when I tell him a submittal is on the critical path (“oh yeah, we submitted those shop drawings weeks ago!”).

Schedulers cannot afford to be gullible. People gives us bad data all the time and expect a good result. Not going to happen! The guy who told me he was 75 percent complete last month will report that he is 60 percent complete this month. Which number is the truth? The reality is that we tend to get better information on the most critical activities because everyone more or less understands what “critical” means (fingers crossed).

So how do we avoid becoming pigeons? Here are a few things to consider:

Never publish a schedule until the draft version has been reviewed by the project stakeholders. We do not want someone coming back and saying the logic is wrong, the durations are a fantasy, etc. after the schedule has been submitted to the owner.
Use Activity Steps on complex tasks to make it easier to report progress. I think Activity Steps are one of the best features in Primavera P6 ignored by most casual users.
“Gut check” the critical path. Even if the critical path seems acceptable to other project stakeholders, they are probably not scheduling experts and they certainly do not understand the details of the schedule nearly as well.
Always use Retained Logic as the scheduling method in Primavera P6 for activities with progress. Nothing keeps you, and everyone else, honest like being confronted with out-of-sequence work.
Avoid unnecessary logic changes. If the project is not going according to plan, why? If the owner is causing work to be performed out-of-sequence we need to preserve the logic to show the delay. But if the contractor has simply changed his mind, by all means modify the logic to keep it realistic.
Educate yourself. If scheduling is not something you really understand, you need training or guidance. I studied CPM in college, but I really did not know how to apply it correctly until I started working side-by-side with professional schedulers.

There you have it. Spread your wings and fly! Okay, so perhaps that is not really the best metaphor.

Jan

The Future of Scheduling Software

Categories: P3, P6 EPPM, P6 Professional, SureTrak

Tags: Microsoft Project, P3, Primavera P6, SureTrak

Comments Off

by Bill Pepoon

Recently I read a blog by someone claiming that Primavera P6 will eventually disappear because there are much better scheduling programs now available. This is a great way to get people to read your blog, I suppose, but there is absolutely no reason to believe that Primavera P6 is a dinosaur happily living its life until some giant meteor program wipes it off the face of the Earth.

To begin with, none of us can accurately predict the future. There are the "unknown unkowns" (as popularized by former Wall Street trader Nassim Nicholas Taleb in his book, Black Swan) that are impossible to predict because we cannot even imagine that such a thing could happen. Might as well flip a coin to determine whether a program like P6 will survive because the results will be just as accurate.

Change is inevitable. I entered high school in 1974. My class was the first one not to learn how to use a mechanical slide rule, because hand-held calculators were becoming readily available. By the time I graduated in 1977 my calculus instructor had introduced us to a primitive desktop computer with a keyboard and no screen. When I graduated from college we were scheduling projects using a mainframe computer.

In 1983, Primavera Systems released its eponymous scheduling program, Primavera Project Planner (P3). Microsoft Project was released the following year. Microsoft does not like competition, to put it mildly. Here is just one example. Like many people, I used Netscape Navigator in the early days of the World Wide Web. Navigator was pretty cheap - I recall paying about $35 for the "Gold" version.

But when Microsoft released Windows 98 they included Internet Explorer for free. Internet Explorer was clunky compared to Navigator, but hey, free is hard to resist. Bye bye Navigator! (The irony is that Netscape Navigator started out as a free program to encourage its adoption, not unlike what Microsoft was doing).

My first spreadsheet program was VisiCalc. But unless you are over the age of 50 it won't ring a bell. Microsoft destroyed that product as well with Excel. Lotus 123 and Lotus Symphony were no match for Microsoft either.

So imagine you are little ol' Primavera Systems facing down the biggest software company in the world. Props for surviving even a few years! Granted, P3 was clearly a more robust scheduling program, but Microsoft Project was much cheaper. And it is not like P3 had a big head start. In the early days it would have been very tempting to just buy Microsoft Project.

Luckily, the consulting firm I was working for in the 1980s decided to buy P3 instead. It may have had something to do with the fact that Primavera's headquarters were right across the river from our offices. But the reality is that even back then it was not a two-horse race, and our firm did consider other alternatives before picking P3. We even hedged our bets by investing in an early competitor of Primavera.

Primavera has always had competition and responded accordingly. I still hear from Primavera users who pine for the days of P3. Really? I used P3 for nearly 20 years, starting with the DOS version and then Windows. It was like an old friend, my bread and butter scheduling program. But when Primavera P6 came along I kicked P3 to the curb and never looked back.

Oh sure, I griped a little during the transition to P6. It was different and required a learning curve. And when I was really busy that was sometimes annoying. I just wanted to get my work done! I am sure that most of us feel that way about technology, that sometimes change is not necessarily an improvement (Windows 8 comes to mind).

The thing is, P6 is better than P3 - demonstrably so. I became interested in selling P6 software several years ago for the simple reason that I was convinced it was much better than P3 and would become the new standard for scheduling software. It is ironic that people criticize P6 for both doing too much and too little. "I don't need all these features", some will say, while others will ask for more features. Oracle must think we are neurotic, or perhaps ungrateful.

Is P6 too complicated? Not really (pause for chuckles). Okay, so there is a lot going on with this software. I use about 70% of the features on a daily basis, which is why I have never really taught everything there is in P6. If someone like me, a professional scheduler for 33 years, cannot find reason to use some of these features I am pretty confident that most other users do not need them either.

Nevertheless, I mostly schedule construction projects. Primavera P6 was not developed for just one industry, and it certainly would not make much sense that software companies are going to develop scheduling software specific to each industry. There is one exception: linear construction such as tunneling projects does have its own standard-bearer, TILOS. This program schedules work based on time and location. P6 cannot match that.

I look at P6 the same way I looked at my first sportbike. Back in 1995 I purchased the most powerful 600cc motorcycle in the world: the Kawasaki Ninja. It was capable of a top speed of 155 MPH, which was serious velocity for an engine of that size. Redline was 14,500 RPM! Formula One race cars costing tens of millions of dollars were only reving to 16,000 RPM at the time.

Did I need that sort of performance? Obviously not. But if my bike was that good I knew the only limitation would be my skills. The bike would never let me down. The software will do more than I need it to, and that is the way it should be. Raise your hand if you use every feature in Microsoft Excel. My point exactly.

Last week I demonstrated to a manufacturer how to schedule and resource level one of their typical projects where the minimum duration for a task is two minutes and the maximum duration is only two and a half hours. That is not a level of detail I would ever see in a construction schedule. I don't schedule down to the minute of the day. But P6 can do that for people who do.

To those who believe that P6 has become stagnant, I would point out that Oracle has made 97 significant improvements since taking over Primavera Systems. How do I know this? Oracle has something called the Cumulative Feature Overview Tool. Enter the version of Primavera P6 you are currently using and the version you are thinking of upgrading to, and this tool explains what improvements have been made.

Getting back to the so-called demise of Primavera P6, I have a pretty good idea which program the author thought was the future of scheduling software. And like many of Primavera's competitors it is more expensive. In this case, you cannot just buy the scheduling module. Instead, you are buying an entire suite of programs related to planning, budgeting, cost and document control. So it is a little ironic this competitor would have the tag line, "Features Only Matter If They Get Used."

Sure, Primavera P6 can be integrated with other Oracle (and non-Oracle) programs, but it can also function as a standalone program. Many of my customers already have document control, estimating and other software. They just need a scheduling program, not a complete overhaul of their computing system.

Other competitors offer standalone scheduling programs, and I am certainly not going to debate the merits of each one here. I know my clients pretty well, and I do not believe they would feel insulted if I stated that I am better at scheduling than they are. After all, I have more than three decades of experience and I still work with this software every single day. I know what they need from a scheduling program. So here is my advice:

"Primavera P6 is your Ninja sportbike. Twist the throttle as much as you want. Fast or slow, it will take good care of you."

Software is a bit like evolution. The winners either adapt or die. Programs such as VisiCalc and Lotus 123 that were among the first to market did not last very long once competition arrived. Primavera has had competition since 1984. That is not to say that other products won't continue to steal some of Primavera's market share, but if it's my money that I'm playing with I will continue to put it on a winner. And Primavera is on a winning streak that has lasted more than 30 years.