The last couple of months have seen the publication of two books designed to provide a practical foundation for people wanting to understand the basic ‘mechanics of scheduling’.
Murray Woolf published his CPM Mechanics, available from http://www.cpmmechanics.com and Aldo Matos published Planejamento E Controle De Obras, written in Portuguese, the book is available from http://construcao-engenharia-arquitetura.lojapini.com.br/pini/vitrines/produtos/produto3634.asp.
Both books are aimed at a similar market niche; working schedulers who have realised that understanding what their software actually does from the perspective of computational mechanics (to use Murray’s term), is essential for the creation and maintenance of effective schedules.
Simply being able ‘push the right buttons’ to drive scheduling software without understanding why the different functions exist is similar to being able to drive a car without being able to read a map and plan a route. The mechanical skills for driving the ‘hardware’ are only useful if you know exactly where you want to go! Once you have learned to ‘map read’ you can plan to go virtually anywhere and understand the potential challenges of the journey before you start.
The scope of Aldo’s book is wider, including sections on Line of Balance, Activity on Arrow, and balancing time and cost. With very limited Portuguese much of the text was beyond my reach but anyone working in Brazil, Portugal, East Timor, etc., I would certainly recommend this book as a valuable resource for understanding the technical aspects of the art of scheduling.
Murray’s book is true to its title and concentrates on the mechanics of CPM scheduling in Precedence networks. As with his other books, it is thorough and detailed, contributing another component to the library of scheduling books Murray has planned for publication.
Many of the more established tomes such as James O’Brien’s CPM In Construction Management (6th Ed. By Fred Plotnick & Jim O’Brien) are targeted at experienced practitioners and are both expensive and wide ranging.
By focusing on a specific and important skills development niche; both of the books discussed in this post are effective, affordable and useful to schedulers seeking to advance their careers and move from knowing what their software does to understand precisely why their software produces the results is does, and importantly be able to recognise any software induced errors.
Fortunately the language difference allows me to avoids the difficult job of determining which is ‘best’ and I’m happy to recommend both books – chose the one that ‘speaks to you’.
Pat Weaver
Lynda Bourne
Aavssitedev's Blog
http://1.usa.gov/LbDmvF and http://bit.ly/M4RMTF are two possible contributors to the “art” and “science” of scheduling.
Sorry, I can’t agreed with the expression “Precedence based CPM scheduling”. Or you’re using CPM saying ‘activities on arrow’ or you’re using a proper Precedence, ‘activities on node’ in case based on the possibility to use a start to start- and a finish to finish relationship with or without lags between two activities. I know it could be difficult to understand and to accept.
Please call “Precedence based CPM scheduling” in the future as “CPM on nodes”. In that case ‘professionals’ will understand what you want to say.
Jack Coolman, Belgium, Europe
Jack.coolman@skynet.be
I have no idea what you point is Jacky.
CPM was developed using the ‘Activity-on-Arrow’ notation by Kelley and Walker in the late 1950s. Dr. John Fondahl developed the ‘Activity-on-Node’ or Precedence notation and published his paper in the early 1960’s, the calculations are fundamentally the same. Both styles of drawing a CPM network use a single duration estimate which differentiates them from probabilistic systems such as PERT. For more on this see ‘A Brief History of Scheduling’: http://www.mosaicprojects.com.au/Resources_Papers_042.html
Murray Woolf’s book only deals with the Precedence calculations for the fundamental reason Arrow networks are virtually extinct. Aldo’s book covers both options.
Pat, thanks for the kind words. I hope my latest books helps clarify a number of misunderstands that engulf the Scheduling Practice. Jack’s comment provides a convenient illustration of the prior sentence.
Jack, as Pat has correctly explained, both Precedence Diagramming Method (PDM) and Arrow Diagramming Method (ADM) are fairly comparable variants of the Critical Path Method (CPM). Yours is a common misunderstanding — that ADM is synonymous with CPM, but PDM is not.
As much as 20 years ago I took issue with Jim O’Brien’s seminal book, “CPM in Construction Management,” in which he essentially made this same mistaken point. Upon further discussion it became clear how the confusion came about: Jim, like many other “Venerables” of their time, strongly resisted the introduction of PDM and, at least in their minds, refused to recognize it as an equally legitimate form of CPM.
I am happy to report that, since those discussions, the Schedulng Practice has come to accept that PDM is not only here to stay, it is the dominant form of Critical Path Method. It would be laughable to NOT regard PDM as a true CPM variant, when it constitutes well in excess of 95% of all CPM schedules in existence.
By the way, as an interesting aside, it was none other than Jim O’Brien who recognized my energy and passion for Construction Planning and Scheduling, and very much liked many of the unique ideas I was professing … so much so that he personally made the introductions that led to McGraw Hill publishing my first book, Furthermore, Jim offered to write the Foreword to that book, in which he happily accepted the label “Venerable.”
So, Jack, I hope this corrects your understanding. Both ADM and PDM are variants of CPM. My latest book, CPM Mechanics,” does what I wish a lot of the books over the last two decades of had had the courage to do: DROP discussion of ADM. We may as well be discussing the Slide Rule!
Yes, there are those who relish the return of ADM. There are even those who continue to practice ADM, and there is software that still provides automation of ADM. But, the idea of overlapped activities is here to stay, and PDM is the only scheduling technology that allows that depiction without splitting activities into non-intuitive fragments.
Happy reading!
Murray Woolf
Happy to contribute Murray – I even paid for my copy of your book at the very reasonable prices you are offering!
However I do need to correct one comment to save my friend Raf Dua from Micro Planning International having an attack of apoplexy…… using SS and FF links in a PDM is not the only option for overlapping ‘whole activities’.
In the 1970s ICL (UK) developed the concept of ‘Ladders’ in ADM networks. Ladders provide for a ‘progressive feed’ function between overlapping activities. Ladders are still part of the Micro Planner X-Pert software.
In the 1960s the European ‘Meta Potential Method’ (MPM) introduced a range of overlapping links in a scheduling system remarkably similar to PDM but developed totally separately (according to Dr. Fondahl). The key additional type of overlap being percent based rather than time based. MPM underpins a number of European tools including ACOS+1.
Neither of the above are particularly common in the USA or in the majority of tools currently dominating the world markets but they are part of our scheduling heritage. For more on ‘links, lags and ladders see: http://www.mosaicprojects.com.au/PDF/Links_Lags_Ladders.pdf
The link to the paper brings up several issues in the modeling of networks of activities with Monte Carlo Simulation here in the Space and Defense business guided by DI-MGMT-81650. If work activities are anything but FS, the results from the MCS are less than confident.
Each dependency relationship, other than FS, can be converted to FS, and the result is a “free flowing, dynamic” network needed for the credible probabilistic models.
You raise a very important issue Glen.
Every schedule model is a series of compromises resulting in an inaccurate representation of what ‘might be’ the future of the project. The art of the scheduler is firstly to know what function their model has to support and make the ‘right compromises’ to make the structure of model useful. The second key skill is keeping the probable level of error sufficiently low so that the resulting calculated outputs are useful to management.
‘One size’ does not fit all, and in my view excessive detail destroys usefulness. We need to move beyond the concept of one highly detailed ‘contract program’ developed early in a project lifecycle to a more sophisticated approach to time management. One possible approach is discussed in: /2012/05/26/new-ciob-contract-for-complex-projects/