I returned home from the AACEi (Association for Advancement of Cost Engineers International) annual conference, this year in San Antonio, a few days ago. A very good conference from a truly professional organization. As usual, I learned a lot, but the point of this posting is to discuss a few trends noted.
The terms “Scheduling” and “Scheduler” are becoming more generic. The original 1956 “ADM” version of CPM focused upon “the earliest completion” to meet the needs of the key players, the superintendents of specific projects. The more powerful computers of the 1960s and 1970s led to adoption of the “PDM” variant of CPM, and massive adoption of personal computers in the 1980s, and tools such as Primavera P3, facilitated a closer relationship between project leads and their specific project schedules.
Even in those early days of computer-assisted scheduling, some effort was expended to meet the needs of another group of key players, the facilities and resource managers of an entire company or enterprise. Early endeavors by myself and others were perhaps limited to mapping usage of key personnel, equipment and other major resources amongst the many projects of the enterprise, and then allowing individual project managers and upper management decide questions of conflicting needs for scarce resources.
This had to be repeated, at least monthly, to account for new projects added to the mix as well as to address changing opportunities for best usage as one project may be ahead and another behind schedule. Thus the accuracy of the schedules and their updates of the individual projects was key to the process.
But not all projects have the development of scope to the level of detail envisioned by the creators of CPM. And so by the mid-1980s, Microsoft marketed a previously internal tool for resource allocation (a second definition of scheduling) of personnel on its various software ventures, and so was born Microsoft Project.
It should be of no surprise, even to non-computer programmers, that the algorithms for these two types of scheduling are different and may calculate different answers to the same input, although often not evident until a few updates into the project. Put simply, one algorithm is tweaked for earliest completion of the project, the other for best productivity of tasks within the project.
The need for “the earliest completion” was driven in the 1970s and 1980s largely due to the rampant inflation and escalation then experienced. As we entered the 1990s, and threats of inflation receded, the emphasis shifted to “Enterprise” scheduling systems designed to record and compare productivity of resource utilization between projects. Thus many software products, such as Primavera P6, dropped use of the 1956 algorithm. (Work-arounds to this issue are covered in my previous blog posts).
As we enter the late 2000s and 2010s, the next waves of scheduling are beginning to take hold. Schedule risk assessment (known by the mathematicians of the 1950s, but not supportable by the computers of that era) is a growing field.
In the past I have suggested that construction specifications be written that a contractor submit a CPM not to meet a specified date, but with a 80% or 90% probability of meeting such date. Software products such as Deltek Open Plan, Oracle Primavera Risk Analysis (formerly Pertmaster), and many add-on products such as @Risk, Risk+, Barbacana Full Monte, Intaver RiskyProject, have thrived as the importance of risk assessment has grown.
The growing use of BIM (building information modeling), also enabled by faster and more powerful computers, has led to development of a number of location based scheduling products such as Synchro and Vico. These products through a combination of algorithm and manual input allow users to collaborate amongst trades or subcontractors to support sustainable productivity.
The general idea is that by slowing (right-sizing the crew of) one subcontractor, others can work in tandem without the need for untimely layoffs for any, and this boosting productivity of all.
At the AACEi meeting, a topic of concern by the CDR (Construction Dispute Resolution) subcommittee was that many of these methods render output that is not repeatable.
In some cases, such as P6, the complexity of the calculation involving enterprise wide calendars, resource pools, and multi-project links are such that the actual algorithm used for one project cannot be articulated, let alone repeated. In Risk software, the use of a random number generator as part of the algorithm bars full repeatability. Location based systems, and others which call for manual input during calculation, have similar issues.
But this should be obvious as none of these systems are designed for use in after the fact forensic analysis. As it should be. A pig designed to both fly and sing is not expected to also taste good.
The concern of the AACEi subcommittee has fostered exploration of a possible statement that analyses performed by software that cannot be repeated should be discouraged from use in a court. (The Daubert series of federal decisions also implies this outcome).
My thought is that this should not be a problem for any of the software firms. Different software, used for different purposes, should be encouraged for each project.
I as well as many other authors and practitioners have long advocated that combining scheduling and cost functions is fraught with issues. Accounting measures cost to the penny. Cost engineering (productivity analysis) measures to a far lesser tolerance. CPM scheduling to yet a lesser tolerance as the typical estimate of duration is expected to be within a range (-15% to +20% or similar) as well as having nominal crew sizes to be tweaked in the field.
There is nothing inherently wrong nor wasteful with running a project with a tool such as the old P3 or SureTrak (which provides an envelope of both early and late dates for execution of individual tasks), selecting within these envelopes for best task or collaborative productivity using location based scheduling products (best used for three week look-aheads), reviewing for risk and alternate but very possible critical paths, and uploading initial schedule and then these specially tweaked periodic updates to an enterprise wide project and portfolio controls software such as P6.
All of the current software providers have been cooperatively providing means to facilitate ease of such transfers from one to the next, minimizing the need for duplication of data entry.
Timesheet and cost control systems may require additional information, but can be configured to work with their sponsor’s enterprise system. And forensic analysts can use the pure logic and supporting data with simpler than enterprise software whose algorithms are more readily explained to the judge or jury.
A properly defined project and enterprise program requires all of these disparate analyses which can best be provided by several software solutions rather than one singing and flying pig. Hopefully such implementation will assist the many key players: field superintendents, facility and resource managers, project controls and cost engineers, portfolio managers, and our forensic specialists, such as those at our recent AACEi conference.
Speaking of conferences, your Construction CPM Conference is doing well and growing. Your first year attendance was 184. Your second year attendance hit 263 paid (and apparently some unpaid) attendees, and we look forward to over 300 for your 2013 event to be held January 27-30 in New Orleans. We have negotiated a truly great time, between the first week of Mardi Gras and Super Bowl. Your conference begins with major parades passing in front of our Sheraton hotel on Canal Street, followed by your Welcome Reception. Monday and Tuesday include 100+ keynote and technical sessions, hosted Happy Hour events, our trademarked “Dine with Five New Friends” program to get you reservations at exclusive restaurants, and late evening events culminating with your own private nightclub featuring the Charmaine Neville Band. Your last day features our Mock Trial with the Hon. John M. Marshall presiding.
Your companion program features a walking tour of the French Quarter, a bus tour of Greater New Orleans, a Jazz Cruise on the river steamboat Natchez, and a class in New Orleans cooking. Approximately 20% of our early bird registrants have signed up for this program. More information on your conference is at www.ConstructionCPM.com.