Guild of Project Controls: Compendium | Roles | Assessment | Certifications | Membership

Tips on using this forum..

(1) Explain your problem, don't simply post "This isn't working". What were you doing when you faced the problem? What have you tried to resolve - did you look for a solution using "Search" ? Has it happened just once or several times?

(2) It's also good to get feedback when a solution is found, return to the original post to explain how it was resolved so that more people can also use the results.

Advanced CPM Training

29 replies [Last post]
Craig Shaw
User offline. Last seen 8 years 40 weeks ago. Offline
Joined: 25 Aug 2012
Posts: 4
Groups: None

Hi Folks,

I am in the process of making a mid-career transition into Planning & Scheduling.  My educational path is going well... with the exception of advanced CPM knowledge.  

Several courses that I have taken, focus almost entirely on traditional FS relationships when making network logic diagram calculations. I have found exercises that require calculations around all relationships, different calendars, and contraints.  

I decided to buy Murray Woolf's CPM Mechanics book, but it is way over my head and personally I find it an awkward read.

Does anybody have suggestions where I can learn advanced CPM knowledge in a clear and easy understand format?

Any insights would be greatly appreciated.

Cheers,

Craig

Replies

Rafael Davila
User offline. Last seen 1 hour 27 min ago. Offline
Joined: 1 Mar 2004
Posts: 5077

I understand the time-cost tradeoff procedure as implemented on few software is only valid if there are no resource limitations.

http://stat.mq.edu.au/Stats_docs/stat321/Wk6Handout2.pdf

If you look at the procedure from the above reference you will find no mention to different resources or to resource constraints other than cost.  Assuming cost can substitute individual resource loading is too simplistic.  If it does the trick then no resource loading is needed.

If you have limited resources you crash activities by increasing activity resources but if resources are limited you must borrow resources from other activities and hence the duration of these activities will increase.

The algorithm shall relate the crashing to the specific resources that are impacted and shall take into account how some activities durations are reduced at the expense of others because of resource transfer, a transfer of resources that shall be applied on a continuous time distribution and not as a one-time occurrence. Not only activities duration will change but also changes in critical path might occur. Manual shift of resources must follow if your software does not help.

http://www.deepdyve.com/lp/institute-of-electrical-and-electronics-engineers/multi-resource-constrained-discrete-time-cost-trade-off-problem-and-E0MLr7cxW1

As stated before along with every point on the crashing curve the continuously changing resource distribution shall be disclosed.   In the same way the calculation and use of Drag without considering the shift in resources and activity duration would be equally short.

Many of us solve this task through what-if analysis that can also accommodate other considerations the simplistic time-cost tradeoff cannot.

Steve.

maybe you misunderstood me.

You wrote:"Had you first used drag to optimize a CPM schedule BEFORE running it on the resource-leveled schedule? This is important because, while you are correct that the resource-leveled schedule is all-important, an optimized CPM schedule will obviate many resource bottlenecks that would otherwise crop up in the resource-leveled schedule and result in either delays or additional resource expenditures. And the difference in time and value/cost caused by each resource scarcity/bottleneck in going from a CPM schedule to a resource-leveled schedule is a vital data item!"

In resource constrained schedules many activities that were critical in CPM schedule will have floats and other activities that were not critical in CPM schedule will become critical. It means that our efforts to minimize critical path duration in CPM schedule may be useless and will not add any value. When resources are constrained it makes sense to consider crashing on Resource Critical Path.

By the way we already developed algorithms and soon will implement in Spider Project resource constrained DRAGs.

DRAG cost is easily calculated in Spider Project by those who want it. It is sufficient to define the cost of the day of delay and acceleration for major project phases.

Regards,

Vladimir 

Stephen Devaux
User offline. Last seen 30 weeks 2 days ago. Offline
Joined: 23 Mar 2005
Posts: 655

Vladimir wrote:

"CPM is still useful, it shows minimal project duration if activity durations are as estimated and all required resources are available. CPM is the first step in the schedule development but not the last."

Sigh. Vladimir, we are in violent agreement. As you say, CPM is the first vital step precisely because "it shows minimal project duration", but resource leveling is a hugely important final step. I know it's a long thread to read, but in it I have already written in various posts:

  1. "And I repeat that, in any case, a good WBS and CPM schedule is an absolute requirement for the ultimate goal of a good resource schedule."
  2. "(T)o suggest that I do not appreciate the huge value of resource leveling in scheduling, and how important Spider's functionality is in that area, is to both engage in ad hominem criticism and to create a straw man. In a nutshell, we are in complete agreement about 90% of the things you say about Spider."
  3. "They have one semester with me to teach them all I can about PM: charters, business cases, scope development, WBS, CPM, three-point estimating and Monte Carlos, resource assignments and leveling,.."
  4. "Had you first used drag to optimize a CPM schedule BEFORE running it on the resource-leveled schedule? This is important because, while you are correct that the resource-leveled schedule is all-important, an optimized CPM schedule will obviate many resource bottlenecks that would otherwise crop up in the resource-leveled schedule and result in either delays or additional resource expenditures. And the difference in time and value/cost caused by each resource scarcity/bottleneck in going from a CPM schedule to a resource-leveled schedule is a vital data item!"

That's four.

I broke into project management a quarter century ago with PROJECT/2, which ran on an IBM 370 mainframe and had what at the time was the most robust commercially-available resource leveling algorithm. I promise you, I understand and appreciate the huge importance of resource leveling (which is precisely why I emphasize the importance of, as you say, establishing the "minimal project duration if activity durations are as estimated and all required resources are available", -- precisely so we can then conduct resource leveling and show the differences resource insufficiencies are causing!).  Please tell me what I have done to in any way disagree with you, to hint to you that I don't understand resource leveling and its importance? Or, as you wrote, that "CPM is the first step in the schedule development but not the last." I will be happy to swear on a stack of PMBOK Guides or PMJs or even Dostoevsky novels that I agree with you! (Can we leave out The Adolescent?)

You also wrote:

"I like cost-time trade-offs because it is most simple case for project and portfolio management.

But unfortunately there are other criteria like social, political, safety, etc. that make decision making much more complex."

The other criteria are precisely why, while cost-time tradeoffs are helpful, they are not enough. Decisions in any project investment (and every project is an investment!) must be made on the basis of scope-cost-time tradeoffs. Social, political, safety, et cetera values all have impact on scope, cost, and/or time, and must be quantified to be managed properly. Total Project Control, the integrated methodology that I use and that is the title of my first book (which I discovered yesterday that CRC Press is now looking to publish in a new edition, to complement my new book they are bringing out next year) is significant not solely, or even primarily, for the introduction of critical path drag, but for the other techniques of value measurement, optimization and tracking, into which both drag and all the social, political and safety value drivers fit. I have long wished that you would incorporate TPC's CLUB metric (it stands for the "cost of leveling with unresolved bottlenecks", so I think most people who have read my book would say I appreciate, address and even enhance leveling theory!) and the VBS (value breakdown structure, where social, political, etc. fit in) into Spider.   

"I was involved in construction, shipbuilding, telecommunications, aerospace, government, defense, oil&gas, new product development, manufacturing, software development and everywhere some resources were constrained. Maybe it is different in health care."

The fact that you and I have worked with all these industries is precisely why we tend to agree so much (even though I seem to have trouble communicating this!). And no, health care is no different, other than in its level of ignorance and incompetence.

"So total float, free float and other floats are usually calculated in resource constrained schedules together with Resource Critical Path."

Believe me, Vladimir, I am very aware of this.

By the way, CRC Press is looking into translations of my books into Chinese, Japanese, Spanish, etc. Would you be willing to tell them that if they were translated into Russian, that Spider would recommend them to its clients and potential clients?

Now I have to go and get ready for my webinar on drag for PMI's Scheduling Community of Practice at 11am US East Coast Time. I'm told that around 2,500 people have signed up for it.

Fraternally in project management,

Steve the Bajan

Steve,

I like cost-time trade-offs because it is most simple case for project and portfolio management.

But unfortunately there are other criteria like social, political, safety, etc. that make decision making much more complex.

I was involved in construction, shipbuilding, telecommunications, aerospace, government, defense, oil&gas, new product development, manufacturing, software development and everywhere some resources were constrained. Maybe it is different in health care.

So total float, free float and other floats are usually calculated in resource constrained schedules together with Resource Critical Path. CPM is still useful, it shows minimal project duration if activity durations are as estimated and all required resources are available. CPM is the first step in the schedule development but not the last.

Best Regards,

Vladimir

Hi Mike,

It looks like two/three years ago both of us felt themselves as winners.

Remember that there are both SS and FF links with volume lags (like 500 meters distance between crews). So the progress does not prevent to keep this distance.

And you did not explain if dealing with several hundred thousands activities is reasonable solution for my pipeline construction example.

Cheers,

Vladimir

Mike Testro
User offline. Last seen 14 weeks 3 days ago. Offline
Joined: 14 Dec 2005
Posts: 4409

Hi Vladimir

A SS link with a lead lag is really a FS link with added guess work.

You are guessing where along the multi task the next downstream task can be allowed to start.

If there is only 1 SS link on the multi task then all other downstream logic stops when progress is added beyond the link.

I really enjoyed our debates on this topic two/three years ago and I think I won.

Really best regards

Mike Testro

Rafael Davila
User offline. Last seen 1 hour 27 min ago. Offline
Joined: 1 Mar 2004
Posts: 5077

Stephen,

Time-cost trade off algorithms are very narrow, they rule out the many different scenarios that are required to be considered, they do not even perform their routines considering resource limits and other constraints. 

Teaching project management shall not be exclusive of a single software as none can be considered a one size fits all, quality teaching shall expose the students to different approaches and procedures. 

I believe the main obstruction to teaching other software on the US is the bias government institutions do have to the point they brand specify scheduling software on all their jobs instead of promoting competition and free enterprise. This is forcing users to limit their options and is promoting a software monopoly on government construction jobs. This is forcing educators to limit their options, many years ago SureTrak was among their protected options and even junior colleges would use it, at least some forensic analysis not exclusive of construction contracts could be studied, something that cannot be studied using MSP. Now they have effectively brought down the options to their only two protected players Microsoft and Oracle.

Last year I submitted a complaint on a federal government job and I was ruled out as I was not directly quoting the job but as a supplier to direct bidders, seems like in federal government procurement subcontractors and independent suppliers have no standing at all. It is a huge immovable wall, just you have no standing. I suppose in Russia they use Spider Project and perhaps do not rule out American software, maybe their game rules are more fair and they shall play similar rules against American software in search of fairness. 

I take a strong stand in favor of Spider Project not only because of how good it is but also because it being technically superior to the two federal government options gives me more chances of proving how wrong our institutions are by limiting competition and by-passing the rules of fair game.  I believe we are loosing the competitive edge by avoiding competition through protective commercial actions, I believe we must compete and not run. 

The nature of critical path drag is very useful in some schedules in others is not, for the same reason time-cost trade off are at time useful and at times are not.  If your tools are poor and the job is complex it will not help much, better figure out the solution manually and use the software to communicate it. 

I do not believe Craig will be able to explore many of the advanced functionalities available in the software from the 70s if limited to MSP and P6 as they are lacking in MSP and P6.  It is up to him how far he want to go I just mention some of the several options of the many available to grab the feeling of the so called advanced features and give him some clues on what to look. 

BTW I always resource load my jobs even if on my own and provide the analysis free to my clients as resource loading in spider using multi-resources makes it fast, real fast.  But the truth is I never figured out DRAG on resource leveled jobs as you can see from the following figure. There are no parallel paths and common sense tells me Drag for activities 1 and 2 is 5.  The dotted line represent a resource dependency and not a hard link, if you make it a hard link it will show Drag. 

 photo rldrag01_zpse1e4e6af.jpg

Some of these days I will try export to Xer converting resource dependencies to links and bring back the file with the import function. 

Best Regards,

Rafael

Stephen Devaux
User offline. Last seen 30 weeks 2 days ago. Offline
Joined: 23 Mar 2005
Posts: 655

Rafael,

There are just a couple of clarifications to what you wrote, and then I at least have nothing more to say. The discussion seems to be running off in several different directions:

  1. What is the nature of construction project management?
  2. Why doesn't everyone teach project management through the prism of Spider Project and construction projects?
  3. What is the nature of critical path drag and is it useful?

Answers:

1.  I bow to your construction knowledge. But while this site has a heavy construction presence, it is not the only type of PM that members here are engaged in -- there is oil & gas, there is nuke, there is manufacturing, there is defense, there is new product development, there is health care, and there is even information systems. All are quite different, in work and resource management, from construction. Yet every project is an investment, including construction projects. If decisions are being made without, as you mentioned in an earlier post, the "time-cost trade offs" that you find "not much value" in, then not only are a significant number of your decisions likely to be mistaken from an investment perspective, but the failure to collect data about inefficiencies on a given project will be repeated for future projects. As-built drag data, whether caused by work logic or insufficient resources, will be one such inefficiency if it is ignored. I don't doubt that you know both construction and PM software -- but your dismissive attitude toward new ideas suggests an unfortunate inflexibility and closedmindedness.

2.  Your continued comments in your last post, focused as they were on construction project management and specific software, makes me wonder if you read my previous post at all: "Most of my grad students are NOT headed for the construction industry. Or even nukes or refineries. Most will be headed for IS and software, a few for design engineering/manufacturing, some for health care, and a few for DoD contracting. 1% at most will use Primavera, and everyone else will use MSP or nothing at all. They have one semester with me to teach them all I can about PM."  When you write: "The main reason why many jobs get into trouble is because of poor resource planning and follow up as usually construction sequence is a no brainer if you know how to build it," I can only repeat that my students will NOT be doing construction and that resource planning, while poor, is well down on the list of reasons that projects in other industries get into trouble. And I repeat that, in any case, a good WBS and CPM schedule is an absolute requirement for the ultimate goal of a good resource schedule.

3. Yes, as you say, anyone "can figure out DRAG by reducing the activity duration to zero on critical activities and restoring back one by one, the reduction in project duration will be the Activity DRAG." But they don't. And they won't. You are welcome to do it that way on a project that has 500 or more activities on the critical path. if so, drop me a line when you get done. But I suspect that you won't, either. But remember, if you do, and then there is a slippage or there is reported progress, you'll have to do it again because lots of drags may have changed. I have been paid to compute drag manually on DoD projects with many thousands of activities, and I assure you that I can do it a lot faster than by changing each CP activity to zero duration and back again! But it still is very time-consuming. It is, in short, just the thing that computer software was designed for and therefore should do. The only reason it doesn't do it is because they hadn't thought of it.

I am curious -- are you dismissive of all CPM computations? Total float? Free float? I mean, I've met people who are dismissive of project scheduling in general! Do you just cut to the resource-leveled chase? Because if you think total float is important (let alone free float!), how can you possibly not recognize that drag is more important because (1) it's critical;  (2) it always costs time; and (3) it always costs money (unlike float, which can cost money but often doesn't).

Again, these are my last comments in this thread -- I am clearly not going to change your mind. But I am willing to bet twenty PP credits that, ten years from now, most commercially-available PM software computes critical path drag.

Fraternally in project management,

Steve the Bajan

 

Rafael Davila
User offline. Last seen 1 hour 27 min ago. Offline
Joined: 1 Mar 2004
Posts: 5077

Stephen,

The main reason why many jobs get into trouble is because of poor resource planning and follow up as usually construction sequence is a no brainer if you know how to build it. You build it with resources, without resources you cannot build it.  If you start your planning with poor resource planning capabilities the chances are you will end up with a poor resource planning, a poor plan that will end up in trouble.

Once the job gets into trouble you will find you must get it back in track. When in not much trouble you do not have many parallel paths and use of drag helps to figure out every scheduler knows, it is on the critical path chains where you must look.

You start looking for possible solution scenarios and here is where advanced scheduling functionalities are needed. The brainstorming shall include multiple approaches and not limited to activity crashing. If you continue your scenarios analysis with limited resource planning capabilities the chances are you will end up with a poor recovery planning, a poor plan that will end up again in trouble or that will not give you a good an efficient solution.

In complex jobs to tackle the resource issue efficiently you need strategies and algorithms that assist you with an efficient planning. Usually poor resource planning capabilities forces the software to yield poor solutions.

Frequently we assign fixed crews because at the individual level they are the most optimal composition for the task at hand and forget this is frequently a major obstruction to a more efficient resource planning.  By allowing for variable resource quantities at a single click of the mouse in most cases the network can be compressed by a significant amount.  If you assign a fixed quantity to your crew of 20 and only 15 are available the software will delay the activity until there are 20 available, keeping many of the resources idle until they are available to create full crews.

If your software can provide for variable quantities (between a minimum and a maximum) and adjust dynamically the activity duration based on activity volume of work and individual/named resource productivity then the schedule model will not be unnecessarily delayed and the PM will have no need to set aside your resource plan and make-up for this deficiency with manual adjustments to the allocations and activity durations. This is the kind of advanced features that are the main topic of this thread.

If one of your what-if scenarios include the use of shifts on a single activity you will need advanced functionality that can deal with work production on each shift. Not all software provides such functionality even when they have what they call shift calendars, usually they fall very short and the best way to teach about this trap is by comparing the software.

DRAG is by no means an advanced functionality just because it is lacking in most software. Anyone can figure out DRAG by reducing the activity duration to zero on critical activities and restoring back one by one, the reduction in project duration will be the Activity DRAG. This is tedious but simple computation. If you have parallel critical paths DRAG will be zero. Obviously there is a minimum usable DRAG as it is impossible to execute work at zero duration; no book is needed for a scheduler to figure it out.

DRAG without looking at the necessary advanced functionalities needed to tackle the misleading results of poor resource loading and constraining functionality will not be of much help. Give your students the guidance to explore some advanced features.

Today many software vendors provide for free demo versions of the software that can be used to teach the advanced features lacking on cheap software. Many offer substantial academic discounts on scalable versions that are very competitive and full of advanced features. 

Very rarely you will see at Planning Planet discussions schedule fragments of over 40 activities, that is usually enough to learn and explore the advanced functions. If your budget is so tight consider Asta PP as well as Spider Project that provide such free demos that can be used to expose your students to the advanced functionalities lacking in MSP.  If you at some point want to expose them at no cost to a bigger project you can create a schedule yourself using a single license and they shall be able to take a look at it using the free viewer that have no limitation as to the amount of activities.

Keep in mind we are talking about the learning of advanced functionalities and the discussion is not limited to a single functionality or to single software.

Craig,

I suggest you download the demo versions of the above mentioned software to explore yourself their advanced features. Also consider downloading ProJack Manager demo. 

Best Regards,

Rafael

Stephen Devaux
User offline. Last seen 30 weeks 2 days ago. Offline
Joined: 23 Mar 2005
Posts: 655

Rafael, correct. It will leapfrog among paths. Optimization, with or without drag computation, is an iterative process. Drag simply helps you target the bast activities for compression. And drag cost tells you how much the compression is worth.

If we're dealing with a nuke outage, a day of compression could be worth $1M. To paraphrase the late Senator Everett Dirksen, a million dollars here and a million dollars there, and sooner of later it adds up to real money. Surely it's worth it to do compression of three days total for $3M? Even if it costs $1 million?

Fraternally in project management,

Steve the Bajan

Rafael Davila
User offline. Last seen 1 hour 27 min ago. Offline
Joined: 1 Mar 2004
Posts: 5077

It is a no-brainer under parallel critical path DRAG disappears, when crashing schedules with many parallel paths with equal or close criticality DRAG will disappear or leapfrog among the paths.

 photo disapperingdrag_zpsb6e7523f.jpg

 photo Leapfrog.gif

Stephen Devaux
User offline. Last seen 30 weeks 2 days ago. Offline
Joined: 23 Mar 2005
Posts: 655

Hi, Rafael.

To begin at the beginning, to suggest that I do not appreciate the huge value of resource leveling in scheduling, and how important Spider's functionality is in that area, is to both engage in ad hominem criticism and to create a straw man. In a nutshell, we are in complete agreement about 90% of the things you say about Spider. However, I have been working for over 25 years, both teaching PM and consulting on projects throughout North America. Unfortunately, the next company/team I meet that is using Spider will be the first. (There was one company I directed to investigate Spider, but which elected to use a product with a tiny fraction of Spider's functionality. It may have been the right decision on their part as they'll probably use far less than 50% of the functionality of the product they purchased -- they would only be confused by Spider!)

Most of my grad students are NOT headed for the construction industry. Or even nukes or refineries. Most will be headed for IS and software, a few for design engineering/manufacturing, some for health care, and a few for DoD contracting. 1% at most will use Primavera, and everyone else will use MSP or nothing at all. They have one semester with me to teach them all I can about PM: charters, business cases, scope development, WBS, CPM, three-point estimating and Monte Carlos, resource assignments and leveling, a wide range of flavors re earned value, a little risk management, and a little MSP. And I know that when they get a job, they will have the knowledge to start day one knowing a LOT more about PM than anyone they are likely to be working with! Even though I do mention other s/w packages, for me to waste their time teaching anything but MSP (where I often have to say: "If you're gonna try this in MSP, you have to trick it like this!") would be malpractice on my part.

But I promise you, as soon as I find someone who will pay me to teach whatever I want irrespective of the students' needs, the course topics will be focussed around military history, baseball, cricket, Dostoevsky and other Russian writers, Shakespeare, chess, macroeconomics, the Vietnam War, backgammon, Westerns in cinema, and the history of Barbados. If there is any time left over, I'll try to cover CPM, resource leveling, and the limitations of earned value!

Now, to focus on your specific points: It is rare that, in a discussion, one side provides such a clear example of why the position they are taking is wrong:

"Spider Project can show DRAG but with Spider and a real project it took me no time to see how limited the use of DRAG can be.  Most construction projects have many parallel paths, as soon as you compress them, critical path changes, at some point there might be parallel critical path and DRAG disappears."

Nothing could more clearly show the limitations of seeing the world solely through the keyhole that a software package, however sophisticated, provides. Prior to Spider's implementing drag computation in 2009, you had no idea about drag so you could neither have learned about it nor taught it using software. This is at a point when I had been computing and using it for years to help clients compress and recover schedules in a host of industries: pharma, nukes, refineries, communication satellites, DoD, and, yes, from IS to construction. But even though my book  and many articles on drag had been published ten years earlier, until it appeared in s/w, you weren't using it. And now, as you try to use it the way that you see it in the software, you find it not helpful, and do not think that perhaps you still don't fully understand it. You say "Spider Project can show DRAG" and then you automatically assume that what the software is programmed to show you is the be-all and end-all of the technique. What Spider is showing you is a metric, a number, a tool, and not how to use it fully -- I assume that Vladimir leaves that part up to you.

How much drag CAN be fully used in Spider I don't know -- Spider's vast functionality suggests that invaluable integration of drag not only with cost but also with resource leveling (referred to in my book as "resource availability DRAG" or RAD), ABCP analysis, and even Monte Carlo could be accomplished. But drag is still new, and if its full theory(!) and utilization is not taught to Spider users, and is not "hard-wired" into standard reports, Pareto charts, etc., then only a fraction of its value will be utilized.

But let's take your example of "how little time it took (you) to see how limited the use of DRAG can be" because of "parallel critical path and drag disappearing".

  1. By how much time had you pulled in your schedule you by using drag at that point where drag disappeared? Ten days? Five days? Fifty days? Did that time have no value? If so, it must be a VERY small project!
  2. Had you first used drag to optimize a CPM schedule BEFORE running it on the resource-leveled schedule? This is important because, while you are correct that the resource-leveled schedule is all-important, an optimized CPM schedule will obviate many resource bottlenecks that would otherwise crop up in the resource-leveled schedule and result in either delays or additional resource expenditures. And the difference in time and value/cost caused by each resource scarcity/bottleneck in going from a CPM schedule to a resource-leveled schedule is a vital data item!
  3. So you reached a point where the software was telling you there was no drag. Wow! That's fantastic! If in fact you reach a point on a project where, no matter where you search or how you combine things, nothing has drag, you have shortened the project to zero time! It's instantaneous! What an accomplishment, to be able to build a dam, or a mall, or an Olympic village in absolutely zero time! I hope you will write a paper on this, Rafael -- it will be a great boon to humanity! (Or do you think it may be just that you don't have the software computing quite the right thing when it says zero float? That perhaps you should program it to compute two things in combination instead of separately?)
  4. To explain #3 more fully, you will frequently have a path/activity with drag of zero, or 1, or 2, due to parallel paths of similar length. Far from this being a stopping point, it is often an extremely fruitful point for compression. When doing a drag report for a customer I will frequently point out: "Path A-B-C-D-E-F with duration of 40D is only one day longer than path P-Q-R-S-T-U-V-W, and therefore has only 1 day of drag. But since the next longest parallel path has float of 22D, the drag of the two paths listed above, in combination, have a drag of 22D. Anything you do to shorten BOTH those parallel paths will allow you to compress the schedule as much as 22D at $20,000/D or up to $440,000 of potential reduced drag cost." In the past, I have inferred from Vladimir that Spider would have no trouble combining two designated activities or paths and treating them as a unit for drag computation (perhaps by clicking-and-dragging). If it can't, then it's still hugely valuable info that the user needs to compute without the aid of software. Uh-oh!
  5. Finally, the items with drag -- activities, constraints, resource scarcities, etc. -- on the as-built schedule are hugely important data items that should be unearthed by the project post mortem. These can then be used for future projects to see what items actually had drag and drag cost, and to save time in the future by staffing up certain departments, or using extra shifts, or planning them differently.    

"It took me no time to find out DRAG is of not much value to me and never used it, in the same way I found of not much value competing and more advanced functionalities for crashing a network such as time-cost trade off."

There is great arrogance to dismissing techniques that others find helpful. It's important to have low cognitive barriers that accept new ideas, or to be willing to assume that we don't understand everything immediately. And whether one uses drag computation or not, there is no doubt that time-cost tradeoff techniques are valuable for justifying resources and risk, and for estimating bids for contracts.

I've been working on the drag concept for almost 20 years, and I am still learning. I gather that Bernard Ertl has already adopted new ideas that he is programming into InterPlan Systems. I am sure that, once drag's use becomes widespread, schedulers and PMs will be extending it into areas I've never even considered.  And I will be delighted when ALL TPC concepts are incorporated into software. Till then, expanding theory into functionalities that no software currently does is vital to the future evolution of project management.

Fraternally in project management,

Steve the Bajan

Hi Mike,

we already discussed your problem.

I want to remind you that we considered pipeline construction project as an example.

Let's assume that one crew does welding and another crew does coating on 1 km of the pipeline.

You have a choice to create two activities linked by SS and FF links with volume lags, or divide each of these activities to 45 smaller activities (22m each) and link them all with FS dependencies.

Real pipeline construction model for approximately 700km that uses SS and FF linke consists of approximately 8000 activities.

You may estimate the number of activities that shall be created for applying FS links only: approximately 31000 activities for each trade and there are separate trades for earthwork, welding, lowering, coating, insulation control, backfilling. This is a small part of project activity types but they alone require to create 186000 activities if to use only FS links.

There are many other activities in the same schedule, so the model will become huge and there are few packages (Spider included) that will be able to work with it. But is it convenient and practical?

Regards,

Vladimir

Mike Testro
User offline. Last seen 14 weeks 3 days ago. Offline
Joined: 14 Dec 2005
Posts: 4409

Hi Rafael

Before you can teach the basics of any planning software the student has to know how to build things.

All available software has un necessary complications in the form of functions that are either never used or are a negative influence on planning practice.

For instance no one on PP has given me a valid explanation as to why we need a FF link unless it is to close up an equally useless SS link.

For a while MSP were ahead of the game because you could not add a FF link - then they joined the herd.

Best regards

Mike Testro

Rafael Davila
User offline. Last seen 1 hour 27 min ago. Offline
Joined: 1 Mar 2004
Posts: 5077

Steve,

The disagreement is huge.

For a start we are talking about the learning of advanced features, good use and understanding of their limitations cannot come without real testing. Only by exploring the functionalities with real software the scheduler can really understand their value. 

Spider Project can show DRAG but with Spider and a real project it took me no time to see how limited the use of DRAG can be.  Most construction projects have many parallel paths, as soon as you compress them, critical path changes, at some point there might be parallel critical path and DRAG disappears.  It took me no time to find out DRAG is of not much value to me and never used it, in the same way I found of not much value competing and more advanced functionalities for crashing a network such as time-cost trade off.  Maybe for some/few single dominant critical path networks DRAG is of some value.

We usually crash a schedule by creating more shifts or changing logic with changed construction methodologies.  For this you need software capable of handling shifts and resource leveling at it's optimum, something lacking in many software, some even have dysfunctional handling of shifts and cannot effectively model different shifts on the same activity.  You have to compare handling of shifts and resource leveling not by reading a book but by using the different software. 

I had to use ProJack Manager demo as to realize how complex and error prone the use of maximal lag is, to realize this is a feature I do not want, it adds too little to my needs at the expense of too much complication that makes it very hard to verify and understand schedules prepared by others.  

Anyone with the basic knowledge of CPM knows that advanced scheduling algorithms require not only advanced mathematical models but the use of computers as it is not practical to study and test these algorithms by hand. The belief  that it is possible to teach advanced applied mathematical theories that depend on the use of computer software without exposing the students to the software is wrong, if you do not teach using the software you will teach history of scheduling but not real scheduling.

I believe no serious CPM training/learning shall leave out some real experience with activity on arrow methods, even if they are a different representation of the same thing.  Limiting your students to the exposure of a single software is short of good.  

At all my schools even the best of books were secondary to the student experience.  Not even the professors lectures were enough, the professors mainly provided guidance it was the whole experience and your showing you was capable of handling not only paper exams but also practical test was important.  At our courses it was not all about exams, it was also about our performance using the tools available to our trade, laboratory experience is important. 

Can you tell me of a single CPM software that includes among others resource smoothing different to leveling, time-cost trade off, minimal/maximal lag ? How can you really teach these, as if reading a novel or a history book? Advanced CPM training must include the experience with several software that include the advanced functionalities.  Spider Project  includes many advanced functionalities but not a single one of the three I mentioned in my prior sentence.  

In my opinion, the use of advanced scheduling should be taught using several software from different vendors as there is no single software that can expose the student to all advanced algorithms.  A school that uses single software is shortsighted.  The students should be exposed to professional practice in addition to "only theory".  This applies to medical schools, to engineering schools and culinary schools as well.

Get real, the tools do matter, but it shall be your choice as no single software has it all.  A single choice is no choice but authoritarian. The three advanced functionalities that I mentioned before that are lacking in Spider, I do not use them and would be annoying. But others such as variable quantities and workloads I cannot live without; a functionality not available on any of the other software I do know but a functionality that helps with a practical need to handle idle resources and putting them into useful work accounting for the impact of their use on the schedule.

It has been documented that schedule updating in MSP is awkward and falls short for accepted forensic claim analysis methodologies.

From: http://www.ronwinterconsulting.com/MS_Project_for_Construction.pdf

1) MSP Baseline management is cumbersome and confusing for the average scheduler.

2) MSP updating method of using % Complete is not suitable for construction schedules.

3) MSP Data Date issues which in effect complicate the schedule update process are confusing for construction schedulers.

4) Due to the difficulty of using MSP to note uncompleted work before the Data Date, the difficulties in removing status for what-ifs, and the difficulty in identifying variances between a series of MSP schedules, it is hard to do most types of delay analysis (except perhaps RP 29R-03 [6] MIP 3.1 and 3.2)

5) Due to all of the built-in update help routines that automatically insert made-up actual dates and remaining durations; it is difficult to assume that most MSP schedules contain accurate enough status data that is necessary for a detailed analysis.

Training your students to only use MSP puts them on disadvantage not only because they will lack knowledge on the specifics of other software but because they lack understanding of construction industry update requirements and forensic methods. 

I wonder how you can teach advanced forensic methods using MSP.

Best Regards,

Rafael

Mike Testro
User offline. Last seen 14 weeks 3 days ago. Offline
Joined: 14 Dec 2005
Posts: 4409

Hi Stephen

"Right now, the state of project scheduling is a disgrace to the human race. And software enhancements are much less helpful than a deeper understanding of the theoretical base -- that's the thing that might improve the use of such software functionalities that alread exist."

Precisely!

My stand point is that construction software writers are not builders (spider excepted).

That is why they allow pseudo planners to use those lazy shortcuts thet I set down in my PP thread 5 years ago "Ban these planning Abominations"

That created a whirlwind of criticism which I contend proves my point.

My crusade failed but I have not given up.

Please tell me how we can promote simple good planning practice among the dangerous fools that call themselves planners.

Best regards

Mike Testro

Steve,

I do not think that Rafael wrote about the software capabilities but on many issues that shall be considered when feasible project schedule is created.

I would certainly insist that good scheduling shall take into account project (or portfolio) resource constraints that CPM does not do. Teaching scheduling it is necessary to discuss resource constrained scheduling problems because in the real life resources are always limited.

The software creates the schedule but the project model is created by people. It is necessary to teach students how to create project models that take into account all existing restrictions. In other case scheduling results will be at least not optimal and in most cases not feasible.

Scheduling that was done in 60-s did not include resource constrained schedule optimization. Understanding its principles is certainly necessary first step to learning advanced project scheduling that can be applied today.

Stephen Devaux
User offline. Last seen 30 weeks 2 days ago. Offline
Joined: 23 Mar 2005
Posts: 655

Hi, Rafael.

I have to say that I disagree. In my humble opinion, the principles of good scheduling should be taught quite independently of any brand of software. Scheduling was being done long before the first commercially-available software (PROJECT/2) was developed in the late 1960s. Over the years, I have worked with many people who tell of the NASA network diagrams in the early 60s, hanging in corridors on butcher paper and in pencil. I suspect most of those people understood CPM far better than most users of modern software brands.

In my corporate and graduate classes, most of my students use MS Project. I teach them to use sticky notes and do their computations by hand. And I frequently run across users who say things like: "You can't have two tasks with both SS and FF relationships between them!" Then I have to explain why you sometimes need to, and how to do it if you're using a package like MSP that doesn't support that.

Yes, that's MSP. But Primavera and Asta don't compute critical path drag, and Spider doesn't (at least out-of-the-box) support the VBS, the CLUB, the DIPP, or the DPI. And almost all packages that support PDM use the continuous activity default which can result in the reverse critical path anomaly, wasting time, money, and perhaps human lives. I teach all of these with pencil-and-paper calculations, and then show either the shortcomings or workarounds in the various software packages. I'd hope that anyone teaching CPM in a software package would also do so, but the fact is that most s/w instructors (and you are definitely an exception!) don't even know that such issues exist!

As George Box said: "All models are wrong, but some are useful." And, yes, some are more useful than others. But even the best software packages are being sold to people who will do the wrong things. For example, Spider is capable of managing to ROI, but do they refuse to sell their software to customers whom they know will ignore such crucial functionality? Of course not. Nor should they.

But those of us who love this discipline and want to see it improve the lives of humanity should be pushing ALL users toward a profound understanding of project management and scheduling, regardless of the software they use. We should also be hoping that all the software improves -- but such improvement will only mean that it allows the user to come closer to doing what they should be doing. Whether they actually do it or not will always be the issue, regardless of the software. Right now, the state of project scheduling is a disgrace to the human race. And software enhancements are much less helpful than a deeper understanding of the theoretical base -- that's the thing that might improve the use of such software functionalities that alread exist.

And from that point of view, I suspect that Mr. Woolf's book and course substantially add to the improvement of our discipline by creating a better understanding of the principles, irrespective of software. At some point, I am sure that he will incorporate drag and drag cost computation (which Spider was not doing either five years ago). Maybe some of those who learn from Mr. Woolf will become not just Spider users, but Spider users who really understand PM and scheduling! And maybe some of them will force other packages to improve functionality.

And maybe some day Spider, having demonstrated great agility in adopting new functionalities including drag, will also adopt other important TPC enhancements, including the value breakdown structure, the resource CLUB (cost of leveling with unresolved bottlenecks), the DIPP and the DIPP Performance Index. And maybe someone with knowledge like you have will teach their users the right way to use all that functionality.

Old Vietnam vet saying: "All gave some. Some gave all."

Fraternally in project management,

Steve the Bajan      

Stephen Devaux
User offline. Last seen 30 weeks 2 days ago. Offline
Joined: 23 Mar 2005
Posts: 655

Hi, Mike.

My nephew from Cheshire visited me two weeks ago. He is a huge rugby fan, so I wound up watching my first ever rugby game with him explaining the rules. It was interesting, but I think I'll stick to cricket. (There is very little rugby in Barbados -- cricket was the preferred sport for civilizing the colonies. In Barbados, even soccer is far behind. As a kid, I used to cry every year when my school would plough under the cricket pitch and put up the goal posts.)

It'd be pretty much of a miracle if Australia were to win back the Ashes this year.

Fraternally in cricket,

Steve the Bajan

Rafael Davila
User offline. Last seen 1 hour 27 min ago. Offline
Joined: 1 Mar 2004
Posts: 5077

If your training is based on limited functionality software from the start it will be condemned to be non advanced learning experience.  Unfortunately in the last decades the tendency have been to lower the standards in favor of a promised ease of use that was never delivered. 

To learn about advanced scheduling you shall start with software capable of doing advanced scheduling.

Among others :

1. How can you talk about Advanced Portfolio Management if:

· Your software cannot model Risk of the whole Portfolio and limited to individual jobs.

· Your software cannot model Financial Constraints, an issue that cannot be modeled using standard resource leveling algorithms.  An issue that in order to be tackled requires that your software is not dependent on discrete financial periods.       

2.  How can you talk about Advanced Resource Assignments and Resource Constrained Scheduling if:

· Your software cannot model different resource productivities on the same activity.

· Your software cannot model volume of work, different from activity duration, and necessary to determine activity duration based on multiple resource productivities.

· Your software cannot model variable resource quantities and workloads.

  • Resources can be assigned to work on some activity for only a part of their work time. In this case it is necessary to define both assigned resource quantities and workloads Total workload (quantity multiplied by workload) is not enough.  An information that 200% of some resource are assigned does not tell us if there is a need for two resource units with 100% workload or four resource units with 50% workload.

· Your software is not capable of modeling teamed and independent resource assignments on the same activity.  Most software cannot model both on the same activity.  This is a basic requirement for good modeling of Shifts and more.

  • Team assignment means that assigned resources work as a team – if some member of the team is not available other team resources will wait, because they can work only as a team.
  • Independent resource assignment means that different assigned resources may work on activity without interaction with others using their own calendars and availability periods.

·  Your software is not capable of looking for optimized schedules.

  • Resource, supply and financing constrained schedules shall be optimized – delays cost too much and creating suboptimal schedules is a waste of time and money. Unfortunately only few packages like Spider Project and Aurora try to optimize resource constrained schedules.

3. How can you talk aboutAdvanced Scheduling if:

· Your software is not capable of disclosing float values other than basic Total and Free.

· Your software is not capable of Conditional Scheduling:

  • If the software is not able to dynamically select the way to proceed with project execution when project scope, network dependencies and activity properties depend on some conditions not known at the moment when project model is created.
  • Besides, conditional networks are required for risk simulations. In this case manual adjustments of project network are not acceptable at all

4. How can you talk about Advanced scheduling of Shifts if:

· Your software is not capable of modeling contiguous volume of work among shifts.

  • It is not known before project scheduling when each activity will be able to start. If an activity can start during certain shift work time the software shall select and assign resources of that shift and if an activity will not be finished next shift shall continue the work began in the previous shift.

http://www.spiderproject.com/images/img/pdf/Advanced%20Project%20Scheduling.pdf

I do not mean to say there is a software that has it all, you have to look among several to learn and explore most of the advanced features.

You can get as many books as you can, but if all are limited to the CPM of the 80's and make no in depht discussions of new and advanced functionalities you will have an obsolete collection of non advanced CPM knowledge. 

Mike Testro
User offline. Last seen 14 weeks 3 days ago. Offline
Joined: 14 Dec 2005
Posts: 4409

Oh Stephen!

Why don't you tell him what you really think instead of holding back.

Best regards

Mike Testro

BTW 41 points from the Lions is almost as many runs as the Aussies scored against the Windies.

Stephen Devaux
User offline. Last seen 30 weeks 2 days ago. Offline
Joined: 23 Mar 2005
Posts: 655

Hi, Murray.

I'm sure your book and course are excellent. However, I do have one question: do they cover critical path drag and drag cost computation? Any book or course (or software, for that matter!) on CPM that doesn't include these topics is, in my estimation, obsolete.

Fraternally in project management,

Steve the Bajan

Mike Testro
User offline. Last seen 14 weeks 3 days ago. Offline
Joined: 14 Dec 2005
Posts: 4409

Hi Murray

How many times did you hit that save button - I have just spent 10 minutes deleting the duplicates.

Best regards

Mike Testro

Murray Woolf
User offline. Last seen 4 years 40 weeks ago. Offline
Joined: 26 Dec 2003
Posts: 13
Groups: None
Craig, I just re-read your comment. You are referring to my current book, CPM Mechanics." I am blown away by your remark that it is over your head. This book could not be more spoon-feeding! I took six pages just to describe the Activity Box, what each side of the box symbolizes. It was written with the opening assumption that the reader had NO prior knowledge of Critical Path, Project Time Management, or Construction. There are over 225 color illustrations, and every single technical term is fully defined and discussed. Yes, the book introduces new concepts that you will not find hardly anywhere else -- like Activity Multi-Path Residency or Schedule Data Credibility Profile -- but these and other new concepts are designed to broadly increase the student's understanding of how CPM works and can be effectively used on real construction projects. I am truly shocked that you would consider this book over your head, as this is the negative comment I have read. Scores of students have praised the book as being the first one to fully explain CPM in simple language, and written for the lay person. I guess, as the saying goes, you can't please everyone! Sorry. I will return your money, as satisfaction is 100% guaranteed. Murray
Murray Woolf
User offline. Last seen 4 years 40 weeks ago. Offline
Joined: 26 Dec 2003
Posts: 13
Groups: None
Craig, Thank you for your candor. Frankly, I could not agree more with your assessment of my first book, "Faster Construction Projects with CPM Scheduling." It IS difficult to read, even though I wrote it in a friendly conversational style. The blame goes to McGraw-Hill, the publisher, who insisted on moving chapters around and changing terminology, until it became in many areas "hard to follow." In addition, in your case, it was not meant for someone looking for an introduction to CPM. As I stated in the Preface, it was written for someone already in the business for three to five years. Now ... Fast Forward five years, to 2012, when I released by second book. This time I SELF-published. And this book, "CPM Mechanics" is written for newcomers to Critical Path Method. If you still have a need, or know someone who does, I do not think there will be ANY disappointment.. You can learn all you could ever want to know about CPM Mechanics (short of actually reading the book) by going to its own website, www.cpmmechaics.com. I also offer an online, self- paced course by the same name. This course earns 40 PDUs.. We also offer a free LinkedIn 18-week seminar called CPM Mechanics Symposium. Feel free to contact me with any questions or comments. Murray
Craig Shaw
User offline. Last seen 8 years 40 weeks ago. Offline
Joined: 25 Aug 2012
Posts: 4
Groups: None

Thanks for the suggestions.

As a mature "rookie" in Project Controls, I find it amazing that advanced-level CPM logic training is so difficult to obtain.  One can be certified as a PMI-SP and barely touch advanced CPM! (Or any type of schedule logic if one may argue that CPM is not the end-all-be-all.)

It would seem many in the industry are quite comfortable accepting predefined programmed software logic (i.e. Primavera) as more important than understanding how the software is/is not calculating accurate schedules.  One comment that I have heard repeatedly from Project Controls Managers is that there are too many advanced software operators in senior level Planner-Scheduler roles.  From my quest to become educated in CPM, I see how this could happen.

In addition to the suggestions, I may pick up O'Brien's book on CPM.

Any other insights to locate CPM education are gladly welcomed!

Cheers,

Craig 

Stephen Devaux
User offline. Last seen 30 weeks 2 days ago. Offline
Joined: 23 Mar 2005
Posts: 655

Hi, Craig.

So nice to hear someone wanting to learn more about CPM (and apparently comprehending that it is not a simple topic!). I have attended project management classes where CPM was taught in twenty minutes. Unfortunately, by that standard, I am a lousy teacher, because it takes me more than eight hours to teach what I regard as the basics!

As far as I am concerned, any course or book that claims to discuss "advanced CPM" and does not discuss critical path drag is a fraud. One suggestion for you would be to search under the term "drag" on this site and read all the discussions. But to simplify, there are three threads that I would especially recommend:

The first was started by Tanvir Ahmad Niazi. I especially would refer you to my posts #10 and #7. In #10, I list many references of articles, webinars, etc., that discuss critical path drag, its computation and its implications. I think all those references are useful, but particularly this article published last year in Defense AT&L Magazine of the US Department of Defense and then re-printed as Chapter 6 in this new book (just out!) titled Project Management in the Oil and Gas Industry.

Once you have read the discussion and that article, I'd suggest looking at this Planning Planet thread where I posted a critical path quiz that requires drag computation. The answers are in this thread.

All those networks deal with computing drag in an all-FS network (just to please Mike Testro!). If you want to learn how to do the computation in networks that include SS, FF, SF and lags, it's more complex and you're probably going to have to get my book Total Project Control.

You could also discuss drag computation with Vladimir Liberzon and/or Rafael Davila in the Spider Project forum here at Planning Planet, as Spider is one of two (perhaps soon to be three!) software packages that currently compute critical path drag. The other is the Project Optimizer from Sumatra.com, an add-on to MS Project that (1) fixes the reverse critical path anomaly caused by the continuous activity assumption and (2) computes drag. Sumatra also has a walk-through exercise showing how to use drag computation to compress/recover a schedule.

Finally, there is  a presentation (EVM-1389) by a woman named Leah Zimmerman in the Earned Value Management track scheduled for Washington, D.C. in June/July at the Annual Symposium of the Association for the Advancement of Cost Engineering International. If you scroll down on that site, the abstract of her presentation says that: "The concepts of critical path drag, resource elasticity, expected monetary value and technical debt will be explored and explained. A quantitatively driven, corrective action decision-tree and process model will be proposed and the use of the model will be explored to improve project outcomes through better corrective action plans." The first three of those are all Total Project Control (TPC) concepts.

I had never spoken to Ms. Zimmerman prior to last week when I saw the abstract on-line and contacted her. Her background and emphasis are more in the cost field than the schedule field, but obviously they are closely related on projects. She very much seems to know what she is talking about, and I hope that she will take the theoretical basis of critical path drag and its implications even further.

I hope this helps, Craig.

Fraternally in project management,

Steve the Bajan

 

 

Leon Rosario
User offline. Last seen 6 years 3 weeks ago. Offline
Joined: 15 Feb 2013
Posts: 7
Groups: None

Gentlemen,

Good day.

First of all,I would like to introduce to you myself briefly.My name is Leon Rosario currently working in Hydro-Projects.My previous exposure was purely electro-mechanical(external works).Moreso,in parralel to my working experience in view of scheduling for my new job(civil works).I am facing a problem in reinforcing bars production rate for structures.

I need your help maybe you can give me some links or further information regarding rebars installation productivity rates.

Thank you in advance.

 

Best wishes,

Leon Rosario

 

 

 

 

Mike Testro
User offline. Last seen 14 weeks 3 days ago. Offline
Joined: 14 Dec 2005
Posts: 4409

Hi Craig

You are doing the right thing by focusing on FS links and use of calendars.

The best way to gain experience is to work on real projects.

I published a series of Interactive tutorials two years back on Basic Planning and Advanced Planning but never sold any.

Meanwhile on Planning Planet I suggest you search on two discussions:

1. Ban these planning abominations - which will tell you what NOT to do.

2. Sunday Lunch - a short excercise in practical planning which will introduce you to most of the methodods required.

Good luck and Best regards

Mike Testro