When a project moves from the honeymoon phase to finger-pointing and distrust, you have reached the dark side of Project Management.  Whether a dispute gets to the level of legal action or not, the process of disputing workflows or work product content is a costly part of project management.

“We’ve all heard “This project is different.’” Kenneth Cooper

What is needed is a way to forecast the impact of change requests and find ways to mitigate the effects those changes cause later on in a project.  An example given by Kenneth Cooper is that changes in engineering documents affect how construction teams planned to accomplish their part of the project. The impact of unplanned events and conditions has been studied for decades so there are examples today’s modelers can use as starting points.

“Undiscovered rework is the source of snowballing downstream costs.” Kenneth Cooper

From the presentation, we gained insight into the project management process, itself.  While what can go wrong might not always go wrong, every project has the potential to get over budget, to get delivered late, and then to cause morale to deteriorate. Being able to plan around and through change requests is one key advantage of system dynamics applications to projecting management.

            “Simulating the future to examine mitigating actions elevates project quality.” Kenneth Cooper

The five components of dispute cost

• Attorney and analyst fees
• Work to document the issues under dispute
• Staffing the dispute effort
• Every level of management having its focus diverted to the dispute
• Business relationships that are damaged or ruined

Watch the recording below:

Q&A

Replies to questions from Ken Cooper

1. Do you think new methodologies to manage projects incorporate some of the lessons you provided over the years?

Yes. Some of the most impactful methods & tools for project management and performance have been in Engineering, which has such a highly leveraged impact on Construction performance. This is particularly true of new methods that improve either (a) Engineering “Quality” (in the Rework Cycle, quality is simply the fraction of work being done that will not require subsequent rework) or (b) the time it takes to discover needed rework. The numerical value of Quality obviously drives the amount of revision work that will be needed, thus cost and schedule a performance in Engineering. The “rework discovery time” drives how long a time that any necessary rework remains unknown, or as-yet Undiscovered Rework. The more Undiscovered Rework persists for a longer time, the much greater the knock-on impact on later stages of Engineering, as well as on the productivity (& rework) in Construction. An excellent example in this arena is 3D CAD systems that enable avoidance of rework (thus Quality), as well as the more rapid discovery of rework–for example, physical interferences being identified much earlier, rather than being discovered later, or even during the build process. 

2. Do you think that the bulk of your work has been taking that superb model and making it work in real cases?

Yes, the bulk of work I’ve led has been applying (and improving) the model for hundreds of real-world projects. The modeling has occurred at every stage of the project life cycle, from pre-bid to post-completion retrospective, and for purposes ranging from bid support to real-time management, to policy improvement, to dispute resolution. The real cases in my own experience include large complex projects such as shipbuilding, aircraft, satellites, missiles, major tunnels, refineries, power plants, and more.

3. What aspects of the model have changed?

The core structure of the model we developed originally is mostly unchanged (i.e., the Rework Cycle, productivity and “quality” effects that work in multiple feedback loops, and multiple work stages that are affected by precedent work…), but there have been continual improvements made as we learned from more and more real-world project applications. In my own experience, those improvements have been in virtually every sector of the model. Some were made to accommodate peculiarities on a specific project, but I would say that most of the detailed changes I know about have been made to improve the clarity, accuracy, and elegance of formulations. Of course, many other modelers have worked on adaptations of the project model, and some of those changes are visible in publications.

4. What areas of model and usage process are in need of work today?

As noted above, almost every one of the hundreds of real-world project applications I’ve led contributed some degree of improvement to the project model. I’m sure that will continue.

Beyond the area of model formulation, there has been substantial improvement in the tools that assist modelers and clients, such as those developed to aid feedback loop analysis, explanatory diagnostics and graphics, and numerical validation. All of these areas will benefit from even more work.

Key point: In addition to improving analyses, there needs to be more effective “selling” of models and findings to potential and current clients—an often under-emphasized area that would benefit from more development and skill-building. An important contributor to “selling” will be publications that reach out beyond the SD community, in trade and project management journals. This should be a major focus.

5. Can you provide advice to other modelers who want to successfully take a specific type of model and generalize it for sale to others? Like Pugh Roberts does? We haven’t managed to do this in many areas.

 An excellent question. My experience developing and championing the project management application of System Dynamics suggests that this takes many years of persistence and “selling” in the face of understandable inertia and natural resistance to change. In my case, it was even a battle within my own firm at the start. That said, there are a few things that I think are necessary for successfully “genericizing” a model for selling to clients—

1) A problem recognized by many potential clients, to be important, perhaps a problem that is seen to have major financial consequences.

2) A problem that is widely acknowledged to need “a better way”.

3) Connecting with organizations that could be early adopters.

5) Connecting with a senior individual who is influential (or even decision-making) in an organization, who is willing to champion an initial use, and to continue championing repeated use. In my career, I have seen how just a handful of senior executive champions have made the “genericized” project modeling successful.

6) Ideally, early uses of the model that has large visible (& publicized) impacts, one or more uses that “make a splash” in the target market.

7) In what should go without saying…working with a highly talented team that brings together a portfolio of skills—modeling & technical, commercial managing, client relations, and presentation and selling, of course.

6. The outbreak of COVID19 and subsequent lockdowns resulted in project schedule reviews. How many of these reviews would have been used to slip in additional slippage? 

 See reply to question on “Forces Majeure” below

7. Are there any experiences using System Dynamics to improve budgeting and planning “pre-project”? 

 Yes. Because many projects build off a base of prior similar projects (new versions of ships, aircraft…), once there has been at least one model developed and proven on a project in an organization, there can be pre-project planning uses on upcoming similar projects. Examples from my own personal knowledge —

1) Shipbuilding. Two broad situations here, one for subsequent ships within a single program (e.g., CG-47, then CG-48, CG-49…), as well as for brand new program ships (e.g., the CG cruiser ships were loosely based on prior destroyers in reality and in our modeling— which we used to test, for example, cost-schedule tradeoffs during the pre-project bidding and planning process).

2) Aircraft development. For example, the modeling of the F/A18 Super Hornet fighter aircraft began before the defense contract was awarded to our contractor client. We worked closely with the program manager conducting “what if” analyses of different plans and conditions, to test in advance good impact mitigation practices. The program was later recognized by the Defense Department as one of the most successful aircraft programs ever.

3) Missile programs. Multiple missile programs at one client benefited from our modeling of the likely performance and bidding plans of competitors. Every known difference in conditions for the main competitor was injected into models, to foresee what the range of project bids by the competitor was likely to be. This led to multiple successful bids for our client. Of course, doing this requires an extremely well-proven model to start with.

4) Fluor Corporation. A notably special example comes from Fluor, with whom we worked for many years, first helping settle large disputes, but then moving on to pre-project planning to help avoid disputes (see “Managing the Dynamics of Projects and Changes at Fluor). We built a system to enable widespread use of a Fluor-specific model before projects started, or very early in their process. The work was documented as saving hundreds of millions of dollars and was the winner of the System Dynamics Society Applications Award 2009-2010, and an Edelman Award Finalist in 2011.

8. The project model is a very interesting metaphor to address industrial disputes. How to deal with class action disputes, like health impacts, environmental impacts, etc.?

Using a simulation model to help deal with other kinds of disputes would be entirely feasible in some circumstances, so long as—

1) The model of the system’s historical behavior can be shown to be accurate.

2) There are specific and documentable “direct impacts” that are included in the historical simulation. These direct impacts will be the sources of “secondary” feedback-driven broader impacts.

3) The objective of the analysis is to quantify and explain the full amount of “damages”, which can be done by removing the “direct impacts” to execute a simulation of the system’s performance in the absence of those impacts. The quantum of difference between the historical case and the “would-have” case is the number of damages to which the damaged party would be entitled.

9. Less experienced staff leads to a higher risk of safety incidents. Safety incidents can escalate to the point where work has to be stopped for the safety issues to be addressed. 

 Fair point, although, fortunately, I have seen only minor instances of this.

 10. Would the COVID epidemic be considered an exemplar of “Force Majeure” and reasonably lead to a re-thinking approach to a project(s) and then to re-estimating? Or, would you consider COVID (et al) a “normal cost of business?” 

 Acknowledging first that I am not an attorney, COVID seems almost unparalleled in the annals of Forces Majeure. There are of course the effects of the virus illness itself on, for example, staff availability.  In addition, the business disruptions from government responses, lockdowns in particular, have certainly led to delays in work conduct, availability of design information, availability and cost of supplies, and more. These areas of impact seem like they would be worthy of designation as Forces Majeure. 

Present at the Seminar Series

The Society Seminar Series consists of periodic online meetings on topics of interest to the systems thinking and System Dynamics communities. These virtual activities cover a wide range of topics that cross many domains while bringing together academics, practitioners, and students together for learning and lively discussion. Send your seminar proposal here

Sponsor a Seminar

The Society is actively looking for Seminar sponsors. This allows making a seminar open to all and free of charge. If your organization would like to sponsor one of these events, where you can promote your organization, firm or software, for instance, contact us at office@systemdynamics.org