Optimising systems, not projects
To successfully manage complex projects, we must learn to understand and model the complex interactions between the systems that influence our activities, and that we affect in return. Systems dynamics and systems intelligence offer methodologies and conceptual frameworks that can help us do exactly this.
In our last post, we discussed the role of situational awareness in project management, and introduced the idea that systems thinking and systems dynamics methodologies can help us take project decisionmaking to a significantly more powerful level. What exactly are these concepts, and how can we make use of them?
Systems dynamics is a methodology for improving the decisionmaking in complex projects and processes that involve multiple levels of interacting variables. The methodology is based on identifying complex cause-and-effect relationships, forming clear mental models of these relationships, and understanding the short and long-term consequences of various actions. Systems dynamics can help identify unexpected side effects of decisions while uncovering the key variables where small changes can have huge leverage effects on the final outcome.
To understand what makes a project run well, it is often most instructive to analyse the cases where the process has failed.
Structure creates action
The larger and more complex the project, the higher the stakes. To understand what makes a project run well, it is often most instructive to analyse the cases where the process has failed. From nuclear powerplant construction projects that have been delayed by over a decade, to procurement lawsuits over multibillion-dollar warships, we find in almost every case that the issues have been caused by a failure to understand complex dynamic systems and by a lack of tools for analysing systemic impacts of decisions.
Problems in complex projects arise from the combined, simultaneous interaction of countless factors. Delayed and cumulative effects can be particularly difficult to observe, as their significance may not be intuitively obvious. A qualitative approach alone (i.e. analysing how and why a problem happens) is not enough.
To understand and predict the behaviour of complex systems, computer simulations are the only viable option. Thankfully, today we actually have simulation tools that can be used not only for individual industrial tasks, but also for extremely complex real-world projects and business processes. By combining a system dynamics approach with advanced simulation capabilities, we can model and visualise processes with a range of constraints and uncertainty factors. Simulations and modelling allow us to carry out "what if" analyses to identify the effects of various actions and decisions on the eventual outcome of the project.
Furthermore, we also need to understand the level at which we want to optimise performance. It is a universal rule of control theory that by optimising the performance of a single component, we inevitably end up hindering the performance of the system at some higher level. For an individual automation solution, the result may be a slight decrease in overall efficiency; for a major industrial project, the outcome can be a loss of billions of dollars.
Small changes, big effects
Somewhat counterintuitively, large-scale industrial projects are among the most consistently mismanaged endeavours in contemporary society. Cost overruns, delays and quality issues are the norm rather than the exception. Research has shown that on average, large commercial projects cost 140% of their original budget and take 190% more time than originally planned. For defense projects, the figures are even more drastic.
On average, large commercial projects cost 140% of their original budget and take 190% more time than originally planned.
In complex large-scale projects, one of the most common sources of delays and cost overruns is the failure to understand how small, seemingly insignificant changes to product and system specifications can cause massive indirect ripple effects throughout the entire process. Change needs can of course arise from various directions, and the source can be inadequate planning due to immature specifications from either the vendor or the customer, but also there can be changes in the operating environment that have been hard to predict in advance.
In the 1970s, US shipyard Ingalls was involved in a multibillion-dollar contract to build a new fleet of destroyers and amphibious assault ships for the United States Navy. The project was suffering from massive cost overruns that threatened to bankrupt the company. In a complicated lawsuit extending over most of the decade, Ingalls and the Navy wrestled over whether the delays were due to mismanagement by the shipyard or by excessive changes requested by the customer.
In one of the first major implementations of systems dynamics modeling, Ingalls ultimately created a complex computer model that simulated the entire shipbuilding process from start to end. The results clearly showed that the changes requested by the Navy during the process had been the major cause of the time and budget overruns. The Navy questioned the results and requested numerous changes to the model; when the simulation was run again with the revised parameters, the new results showed that even more of the overrun was caused by the Navy's design changes. The case was eventually settled out of court with Ingalls receiving $447 million in compensation.
Understanding systems of systems
How can we seek to understand projects and processes at this level of complexity? A related concept to systems dynamics is systems intelligence, which is a way of thinking that seeks to enable intelligent analysis of complex systems that involve interaction and feedback.
Problems are never independent of each other, and changes in one area are always reflected elsewhere. Consequently, the successful management of complex processes calls for a systemic approach in understanding interdependencies, instead of focusing on solving individual problems one at a time. It means implementing tools, standardised project models and management practices that allow the project to be run proactively rather than by reacting to a continuous stream of minor problems.
The good news is that systems dynamics and systems intelligence are skill sets that can be taught – and learned.
The good news is that systems dynamics and systems intelligence are skill sets that can be taught – and learned. As a result, systems intelligence is relevant not only for project management, but for the development of the entire organisation. A systems-intelligent team or organisation cannot be built with tools and procedures alone. Instead, it requires a shared mindset for internal growth and of utilising the knowledge of the entire organisation.
At its core, systems intelligence is a willingness to become aware of the systems to which we belong and that we are influenced by, while affecting them in return. It is about looking deeper and identifying structures of which we are usually not aware, and finding ways to change these structures for the better.
Running projects proactively with systems dynamics does not come for free. It requires investment and professional competence, but the returns can greatly outweigh the expenditure. As we seek to succeed with our projects in an increasingly complex, disruptively changing world, the question is not whether we should embrace this kind of approach in our daily work. It is, can we afford not to?
Jari Hämäläinen, Director, Terminal Automation, Kalmar
Peter Ylén, Principal scientist, VTT
|Jari Hämäläinen (Dr. Tech.), Director of Terminal Automation at Kalmar, is passionate about renewal. He is eager to create value with new business models and technologies, find new models to improve ways of working. During his career, Jari has led various projects, big and small.|
|Peter Ylén (D.Sc) is the principal scientist in Business ecosystems, value chains and foresight research area at VTT. His research interests are management flight simulators, decision support systems, applying data and analytics in organisations, modelling, simulating and management of complex dynamic systems.|