Strategies for planning complex systems development


Tony Zink

Vice President, EPMA

Complex systems development and construction projects, collectively referred to here as “complex systems” projects, have challenges that otherwise may not require as much attention in projects of smaller scale or complexity. When missteps occur on these projects that cause cost or time overruns, the impacts can be quite large because they are greatly magnified by the scale of the project. Because of the scale of the numbers involved, missteps on complex systems projects can impact the entire company, people's safety, and the environment. A structured, integrated approach utilizing systems engineering concepts can be a very effective way to mitigate the risks of this type of project.

Complex Systems Projects and Their Challenges

Complex systems have characteristics that that set them apart from the rest of the world and provide unique management challenges. Many characterize these systems as engineering marvels, whether they are enormous in size (such as an aircraft carrier) or tiny in size (such as a computer microchip).

About Complex Systems

Fundamentally, a complex system is a sophisticated structure, device, or other entity that consists of many components that interact with one another; there are often many logical dependencies between the components, and there are many variables that affect their intricate interactions. Complex systems typically cost millions of dollars—if not billions—to develop or construct, and there are typically many parties involved in the process.

Examples of complex systems include the following:

  • Airplanes
  • Spacecraft
  • Ships
  • Offshore drilling rigs
  • Automobiles
  • Computer chips
  • Manufacturing automation systems
  • Buildings
  • Bridges
  • Computer systems

Regardless of the physical size of the finished product, each of these items involves an elaborate, time-consuming, and expensive development or construction process. Although some people may not initially think of buildings or bridges as being complex systems, such projects do consist of multiple sophisticated systems such as electrical systems, plumbing and drainage systems, heating and cooling systems, communication systems, and so on, all housed in or supported by a complex structural system. As a result, we can apply development methods to them similar to those used for other complex systems.

In the United States, the money spent annually on the design, development, construction, implementation, maintenance, and decommissioning of complex systems is immense. According to data available from the U.S. Department of Commerce Bureau of Economic Analysis (2014), of the nearly US$17 trillion gross domestic product (GDP) in the United States in 2013, 7.2% (US$1,208 billion) consisted of construction (US$619,923 million) or the manufacture of complex systems (US$588,375 million), as shown in Exhibits 1 and 2. Industries considered here as part of the manufacture of complex systems includes the following:

  • Machinery
  • Computer and electronic products
  • Electrical equipment, appliances, and components
  • Motor vehicles, bodies and trailers, and parts

Exhibit 1: 2013 Gross domestic product (US$ millions) of construction and the manufacture of complex systems versus total GDP.


Exhibit 2: 2013 gross domestic product (US$ millions) of construction and the manufacture of complex systems versus total GDP.

Complex Systems Project Management Challenges

Activities on complex systems projects such as requirements, design, procurement, build/construction/fabrication, testing, reliability, logistics, and team coordination become much more difficult than with smaller projects because of their large scale and the number of activities that require coordination. Although complex systems projects have many of the same types of issues as smaller or simpler projects, their issues can be greatly magnified and can cause much higher risks. Because of the large numbers involved, missteps on complex systems projects can have enormous impacts not only to the project, but also to the entire organization. Examples include the following:

  • Estimating miscalculations: An overlooked or miscalculated purchase can easily add tens or hundreds of thousands of dollars to the project cost. According to the CMD Group (2014), a passenger elevator intended for use in a four-story building can cost over $100,000; if one elevator were missing from the bill of materials for a new building because of poorly documented customer requirements, incomplete/incorrect engineering drawings, or a mistake in procurement system data entry, that can amount to a large overlooked expense that the customer will not tolerate.
  • Missed long-lead items: An overlooked item with a long lead time (equipment orders, materials orders, zoning and permits, and so forth) can easily add months to the project's timeline. According to an article published on The New York Times website A New Elevator: The Nuts and Bolts (Rogers, 2008), a typical lead time for an elevator installation is six months; if the project manager did not account for this long lead time in the project schedule, then major delays to the project completion could be expected.
  • Lost revenue due to launch delays: An idle automobile assembly line is not making money for the company; for every day that the assembly line is not producing new cars, the company is losing millions of dollars in potential revenue. According to an article published on the Autoblog website titled “These Are America's 15 Busiest Auto Plants” (Tutor, 2012), the top 15 auto assembly plants in the United States in 2011 ranged from 262,730 to 460,338 vehicles produced for the year. It is quite common that an automotive assembly facility can produce more than 200,000 new vehicles per year, which amounts to approximately 600 vehicles per day after accounting for 30 days of shutdowns and nonworking holidays. At this daily rate, based on information from a recent article published on the USA Today website titled “Report: Average Price of New Car Hits Record in August” (Healey, 2013), an average of US$31,252 per vehicle US$18M of revenue would be lost for every day that an auto assembly line launch is delayed.

A Structured Approach to Planning Complex Systems Projects

Although you may prepare a holiday meal or take a family vacation without a concrete plan, would you build a new home without one? Not likely. The days of men building log cabins with nothing more than a piece of wooded land, an axe, and a dream are all but gone (at least in the United States). Constructing a home in the 21st century requires permits, contracts, detailed foundation/framing, HVAC, plumbing, electrical drawings, budgets, bills of materials, timelines, and much more.

A structured approach to planning complex systems development projects can lower risks and save organizations large amounts of money…just as working from a detailed set of plans and drawings greatly improves the construction of a new home. Following are some concepts for reducing risks and improving success rates on these kinds of projects:

  • Reduce complexity through decomposition: Break down the complex system into smaller, simpler components that are easier to manage.
  • Integrate: Reconnect the simplified components to form an overall system that achieves all of the desired results.
  • Implement the right controls: Know the desired results of the overall system, constantly monitor the status, and make course corrections early and often.
  • Implement the right tools: Use sophisticated, integrated program and project management tools to manage the large volumes of complex, connected data that are generated.

The Systems Engineering Methodology

The field of systems engineering offers some important management techniques that we can use to manage large, complex projects. Focused on ways to design and manage complex engineering systems, this discipline includes a set of processes, practices, and tools designed to break down a complex system into simpler components for easier management and then integrate those components into an optimal overall system that achieves a set of desired objectives. We can learn some valuable lessons from the systems engineering methodology and apply some of these concepts to the management of large, complex projects.

Systems engineering practitioners often follow a process represented as the “systems engineering V,” as shown in Exhibit 3. The top level of the V represents the entire assembled end product—be it an automobile, a computer system, or even a building—and each subsequent lower level represents decomposition into simpler components until we reach the lowest level, the molecular level at which the most detailed work is performed.


Exhibit 3: The systems engineering V.

The process begins on the upper left with the definition of overall product requirements, then proceeds to further—and more detailed—levels of requirements decomposition until reaching the most detailed level at the bottom. At this level (the bottom of the V), the simplest components are designed, built, and tested. The components are then gradually reassembled and tested (integrated) at each upward level until finally the end product is completely reassembled, tested, and integrated.

Notice also in Exhibit 3 the process gateways that lie between each level, on the way down the left side of the V (decomposition) and on the way up the right side of the V (integration). These checkpoints provide an opportunity for the team to pause, evaluate progress at the completion of each level, perform rework as necessary, make course corrections as necessary, and proceed to the next lower/upper level when all indicators are green and the decision makers have approved. There may be situations in which the decision makers choose to stop all work entirely; in product development projects or programs, it may be decided at one of these points that market conditions have changed and major changes will be made to the direction of the product or that the product should be killed altogether.

Large Project Scheduling

Before proceeding further into applying the systems engineering V methodology to the management of complex systems projects, it is important to review techniques used for scheduling large projects. When program manager/planners/schedulers create schedules for planning and managing large projects, they have the option of creating a single large schedule for the entire project or multiple smaller schedules that can be integrated with cross-project dependencies to represent the overall project (or program), as shown in Exhibit 4. There are benefits to each approach.


Exhibit 4: Single large versus multiple small schedules for a large project or program.

The benefits of using a single, large, complex project schedule include the following:

  • Consolidation: When using a single program schedule, all of the information is consolidated into a single file or location, making it easier to locate and access.
  • Control: When using a single program schedule, access is controlled centrally for the entire schedule, so a single person or group can more easily control all the information.
  • Portability: When using a single program schedule, the information is typically stored in or exportable to a single file that is easily portable and shareable.
  • Dependency management: When using a single program schedule, it can be easier to manage dependencies throughout the entire program, because all the activities and milestones are in a single place, so no cross-project dependencies are needed.
  • Workload management: When using a single program schedule, it can be easier to manage personnel workloads and over-allocations in the schedule if all the personnel are dedicated to the program.
  • Reporting: When using a single program schedule, it can be easier to generate reports because all of the data are consolidated and accessible in a single location.

The benefits of using multiple, simple, integrated project schedules include the following:

  • Delegation: When using multiple integrated project schedules, it can be easier to delegate management and activities to multiple project managers, departments, or vendors.
  • Access control: When using multiple integrated project schedules, it can be easier to control visibility or editing privileges of individual projects within the program.
  • Ease of use: When using multiple integrated project schedules, it can be less cumbersome to manage smaller project schedules than larger program schedules.
  • Concurrent editing: When using multiple integrated project schedules, multiple project managers can work on their respective projects at the same time.
  • Critical path analysis: When using multiple integrated project schedules, you can view the critical path in a single project or across multiple projects.
  • Reporting: When using multiple integrated project schedules, you can create reports for a single project or across multiple projects in the program.

It becomes clear that the benefits of using multiple, simple, integrated project schedules far outweigh those of a single, large, complex project schedule, especially considering the frequent need for delegation, distributed and concurrent editing, and ease of use. Large projects are often managed by multiple people, and they are too busy to struggle with managing a large, complex project schedule.

Combining Systems Engineering and Large Project Scheduling

Returning to the topic of complex project management, the systems engineering V concept can be applied to the process that the team follows to systematically decompose the project work, perform and monitor the work, and integrate to create the end product. As shown in Exhibit 5, what once was a single “project” now becomes a “program” of simpler integrated projects for easier planning and management, similar to the end product being decomposed into progressively simpler components for easier design, development, and testing. At each level in the complex systems planning V, project schedules and other artifacts are created by one or more project managers, planners, or schedulers to represent the work to be performed at that level.


Exhibit 5: The complex systems planning V.

Although Exhibit 5 represents the process simplistically as two-dimensional, as you decompose the structure (the product structure or the program structure) there are more components at each subsequent level, so you need to generate a three-dimensional management map, as shown in Exhibit 6 (decomposition) and Exhibit 7 (integration). These two exhibits assume three levels below the program level, and that each component, when further decomposed at the next lower level, generates three more detailed components, although the actual number of levels required, as well as the actual number of components at each level, will vary depending upon the system being managed.


Exhibit 6: The three-dimensional complex systems planning V: decomposition.


Exhibit 7: The three-dimensional complex systems planning V: integration.

Implementation Recommendations and Considerations

EPMA recommends implementing a structured methodology for managing complex systems projects that mimics the systems engineering methodology, and that several actions are taken to implement changes to process, changes to people, and changes to tools.

Process-Related Changes

  • Program-level definition: Define the levels of program management that mimic the levels of the complex system being managed. This can be defined on an accelerated timeline with the assistance of a PPM consulting partner who can assist with complex systems mapping and bring leading practices from previous experience.
  • Systems planning V: Create a program planning process with discrete levels of plan decomposition, plan integration, and decision gateways that mimic the defined complex system levels. This can be implemented on an accelerated timeline with the assistance of a PPM consulting partner who can assist with process development and bring leading practices from previous experience.

People-Related Changes

  • Program management organization (PMO): Design and implement a multilevel program management organization with the discrete levels of organizational breakdown that mimic the defined complex system levels. This can be implemented on an accelerated timeline with the assistance of a PPM consulting partner who can assist with organizational change management and bring leading practices from previous experience.
  • Program management staffing: Staff the new program management organization with personnel at each level to act as program managers, program management analysts, and program schedulers who will decompose, manage, and integrate the work, the schedules, and the related program artifacts. This can be implemented on an accelerated timeline by working with a PPM staffing partner who can outsource any and all of the defined organizational functions.

As with outsourcing any type of organizational function to a vendor, there are benefits and risks that you should carefully consider. The benefits of outsourcing these functions to a dedicated vendor may include the following:

  • Time savings: Delegating a function such as project scheduling (which many consider to be administrative work) to a dedicated expert vendor can save a program manager several hours per week that can be redirected to more important activities.
  • Advanced skills: Delegating program management or scheduling functions to a dedicated expert vendor can bring a level of expertise that would be difficult to obtain otherwise.
  • Information quality: Delegating project scheduling functions to a dedicated expert vendor can result in a dramatic increase in data quality, because the vendor's personnel are trained in the tools of the trade and can dedicate their time to ensuring accurate data that can be trusted.
  • Focus on core competencies: Delegating program management or scheduling functions to a dedicated expert vendor allows the existing organization to focus on its core business, rather than learning and maintaining a set of skills and practices that do not directly generate revenue.
  • Efficiency and economies of scale: Delegating program management or scheduling functions to a dedicated expert vendor can result in those activities being performed very quickly and efficiently, and personnel turnover becomes a nonissue because any gaps should be backfilled immediately by the vendor.

The risks of outsourcing these functions to a dedicated vendor may include the following (depending on the vendor and how it operates):

  • Process control: You may have less control over program management, scheduling, or other processes because the work is performed in a “black box.” This may not be an issue for results-oriented organizations, and it may not be an issue if the vendor agrees to follow your established processes.
  • Personnel control: You may have little control over the personnel working on your projects, including backfills when they have personnel turnover. If this is a concern, stress your need for input into the vendor's staffing decisions.
  • Quality: You may have issues with adherence to your established processes, policies, or data quality. If this is a concern, define quality/adherence standards with the vendor.
  • Intellectual property: Many organizations are extremely sensitive about sharing intellectual property with vendors because that intellectual property constitutes their competitive advantage. If this is a concern, institute nondisclosure agreements and possibly other information security measures with the vendor.
  • Language/cultural barriers: You may have communication issues with personnel who live or originate from foreign countries who do not speak your language, literally or figuratively. If this is a concern, stress your need for input into staffing decisions, or require that the vendor's personnel be native speakers of your language.
  • Responsiveness/lead times: You may have issues with unacceptably slow responsiveness or long lead times when you submit any type of standard request. If this is a concern, establish a service level agreement (SLA) with the vendor outlining the expected response times for various types of standard requests, as well as penalties for not performing as expected.

Tool-Related Changes

  • Integrated platform: Select and implement a tool that is fully capable of handling decomposed project information (scope, timing, costs, artifacts) and integrating it with dependency logic, enterprise resource pool management, and process and milestone visibility.
  • Cloud enabled: Select and implement a tool that is cloud-enabled for distributed information access and collaboration, because many program teams are large and geographically dispersed.
  • Automation: Automate standardized processes in the selected tool to improve the efficiency of the complex processes required in your business environment. Human delays and errors can often be removed through the right level of automation.
  • Business intelligence: Develop standardized, automated reports and dashboards to show real-time program status and issues without the need for time-intensive data aggregation and formatting.
  • Data quality: Implement mechanisms and personnel to identify and correct data quality issues so that the information in the system can be a trusted basis for important decisions.
  • PPM tool partner: Select a PPM tool implementation partner who is familiar with your processes, has experience successfully implementing the selected tools, and can bring industry-leading practices to your organization.
  • Training: Provide sufficient training to personnel who will operate the tools regularly, or work with a PPM partner who is fully competent in their usage to operate the tools on your behalf.
  • Support: Implement an internal tool support mechanism to help with user or technical system issues, or work with a PPM tool support partner who will provide support according to an SLA.

Implementation Investment and ROI

The implementation of a new strategy, processes, people, or tools involves an investment of time and money, and any investment should have a sizeable benefit/return/savings in order to make it worthwhile. To justify the investment, begin by identifying your organization's average per-project budget and total annual budget for running complex systems projects, because the changes will directly affect those types of projects. There may be side effects that can also benefit other smaller/simpler projects in your organization, such as a tool implementation that can be shared by other departments that run simpler projects, but the main beneficiary is your complex systems projects.

Next, estimate the cost savings that you expect to receive per project or per year from implementing the proposed process, people, and tool changes. There are many places to find cost savings, including the following:

  • Reduced overtime
  • Reduced redundant work
  • Reduced ramp-up time
  • Reduced late penalties
  • Reduced expediting fees
  • Reduced idle time due to poor coordination or missed opportunity windows
  • Improved cash flow
  • Reduced rework
  • Improved safety

Some of these savings may be time savings (requiring conversion to cost), some may be fixed dollar amounts, and others may be percentages of per-project or total annual project budgets (thus, the first step in the exercise above).

Finally, determine the length of the payback period for this investment. How many years do you expect to pass before the measurable returns should surpass the amount of your investment? Two to three years? Five years?

Final Words/Getting Started

In order to effectively manage complex systems projects using this methodology, consider the following leading practices:

  • Systems engineering approach: Decompose complex systems into a logical multilevel breakdown of components and follow this model for all similar projects.
  • Multilevel program management: Create a multilevel program management organization that mimics the levels defined for your complex systems, and staff each level of the organization with managers/planners/schedulers who can effectively navigate the entire system structure and integrate the work throughout.
  • Multilevel scheduling with integrated tools: Implement project scheduling tools that allow program managers/planners/schedulers to create detailed schedules and artifacts and integrate them within system levels and across system boundaries to create a fully integrated program plan.
  • Gated process with controls: Design a management process that enables you to effectively break program management across system levels/boundaries and insert review checkpoints.
  • Return on Investment: Gather data for a business plan to justify the investment of implementing the complex systems methodology.
  • The right implementation partner: Ask the right questions to select the right PPM partner to help you implement the desired process, tool, and people changes in your organization.

Bureau of Economic Analysis. (2014). Gross-domestic product-(GDP)-by-industry data. Retrieved from

CMD Group. (2013). Elevators cost estimating tips. Retrieved from

Healey, J. R. (2013, September 5). Report: Average price of new car hits record in August. USA Today. Retrieved from

Rogers, T. K. (2008, November 21). A new elevator: The nuts and bolts. The New York Times. Retrieved from

Tutor, C. (2012, July 6). These are America's 15 busiest auto plants. Autoblog. Retrieved from

© 2015, EPMA
Originally published as a part of the 2015 PMI Global Congress Proceedings – Orlando, Florida, USA



Related Content