We set out on the path to ‘Reframe our future’, as a way to step out of the Australian response to the Coronavirus Pandemic of ‘damage control’ and bring a perspective on reframing the discussion to build national resilience. Damage control has been an important step, but it has accentuated the fragility of our economy, our sectors, supply chains, health system, border security and sovereignty. It has also highlighted another key aspect; that is, the interrelationships between each element of a resilient society. Each of these elements are a system within themselves and none of these systems are discrete. They are, in fact, a ‘system of systems’; which is where Systems Engineers can make a real contribution to enable strategy and support change. By modelling the systems and their interrelationships, we can integrate and analyse information across multi-disciplinary domains to inform decision making and enable strategy.
In this article, the last in my contribution to Shoal’s ‘Reframing our future’ series of articles, I look at what a model-based approach to National Resilience Framework could be. The question I hope to answer is what is, and how would you implement, a National Resilience Framework at the technical level to support the decision makers?
Like Shoal, numerous organisations have presented in public their thoughts and insights on how to improve decision making by employing some form of decision framework. In my last article, I explored frameworks such as a Causal Loop Diagram (CLD) or a Failure Mode Effects and Criticality Analysis (FMECA). Outside of this is the recent rise in visual-based decision frameworks. These ‘dartboard’ visualisation-based frameworks generally categorise many aspects such as leadership, society, or health in quadrants of a circle. They then provide some guiding principles and goals to address each of these categories, around the outside of the quadrants making up the dartboard. Visually, they are a useful tool and have their place, however, as a framework, they are too high-level. They lack the analytical detail of interrelationships and activities provided in tools such as CLD and FEMCA.
Regardless of the approach taken to decision frameworks, there is a need to codify these high-level conceptual guiding principles and goals, and relate, or trace, them to the analytical detail of interrelationships and activities at the implementation level. In other words, provide a digital thread of logic, from strategy to implementation. This ‘digital thread’ codifies the information and decision rationale and, as I stated in my last blog, can “…deliver the robust and contestable decisions we need.”
An entity-relationship model (ER Model) is applicable in this context. Originally designed by Peter Chen in his seminal work on a Unified View of Data an ER Model provides us with a valuable and robust approach that describes the interrelated aspects (such as ‘guiding principles’ or ‘failure modes’ referenced earlier) in a specific domain of interest. This description of the ‘Entities’ their ‘Relationships’, and attributes of each, enables both a digital thread of logic and a codified definition of a problem space that can be systematically analysed and decisions subsequently made.
As a simplistic example, and drawing on the referenced decision frameworks, consider the following thread of reasoning and logic:
- A Strategy identifies guiding principles
- Guiding Principles establishes goals
- Goals prioritise Capabilities needed
- Capabilities needed are achieved by an organisation
- An organisation performs a service
- A service exhibits a performance
- Performance includes a level of resilience needed.
This simplistic underpinning schema, or ontology, categorises the information classes of the problem of interest and provides the structure that delivers a decision framework. This can then be used for robust and contestable decisions. The structure in the schema is utilised to capture the instantiations of the real-world information of the problem, in a ‘descriptive model’. The instantiated descriptive model provides the analyst and decision makers with the visible knowledge at their fingertips to gain the insights from a digital thread of logic and reason over issues such as the impact of a shock or viable solution options.
The challenge with the complexity of ‘wicked problems’, such as enhancing our national resilience, is that some Systems Thinking tools cannot be stretched to robustly cover all aspects of the problem space. As considered by George Miller, from Harvard University, in his 1956 paper in the Psychological Review, that is still relevant today, we limit our “…judgments to about seven categories.” This is where a well-structured descriptive model can help. The structure allows for the decision maker to focus on a single digital thread of reasoning and only visualise the aspects needed to make the decision, whilst being informed on the broader issues. Other tools, that aren’t model-based, quickly become unacceptable for supporting the decision making when they extend beyond tens of components and relationships. This is the level of complexity that we expect with a National Resilience Framework.
A model-based decision framework enables better decision making by providing a structure and hence bringing clarity to the interconnectedness of cross-domain problems and solutions. This is even more important for decisions around building national resilience, where Government investment is likely required and funding will be finite. It will deliver a ‘rich picture’ of an appropriate sub-set of information to the decision makers providing the digital thread of rationale that realise contestable decisions. Better decisions. Particularly around prioritising available resources to maximise the value of the investment.
Given Shoal’s expertise in understanding the ontology of problems, how decisions are made, and capturing the information and rationale in descriptive models, we are ideally positioned to support development of a National Resilience Framework for robust and contestable decisions.