Good decision-making leads teams to operational success. That might seem obvious, until we consider the challenges that good decision-making needs to overcome. A typical predict-then-act approach to decision-making relies on an understanding of the circumstances and the likely outcomes of various courses of action. Crucially, teams need assurance of the quality of their understanding in order to make decisions with confidence. Confidence in decision-making is the lifeblood of any well-run organisation, be it in business, defence, or government; an organisation and its component parts have to be willing to rally behind a decision in order to bring its expected success to fruition.
In practice, reaching such confidence is not always a straightforward process. What happens when circumstances are volatile, ambiguous, or prohibitively complex? What about when confusion is amplified by the pressure of high stakes consequences, such as success or failure in business, victory or defeat in military operations, or navigating the dangers of a pandemic? In these cases, it is very easy for emotions to flare and for opinions to split. This can lead a team into a state known as ‘deep uncertainty’.
Deep uncertainty exists wherever decision makers and stakeholders do not know or cannot agree on the likelihood of future events or the consequences of potential courses of action. Under these circumstances, typical predict-then-act approaches to decision-making can grind to a halt behind the inability to make assertive predictions and manage expectations. Deep uncertainty can occur in any team, regardless of the amount of experience and expertise of its members, when the means of projecting a definite future outcome are unavailable. In these cases, a team needs resources for parsing the complexity of the circumstances and giving weight to one decision over another.
The role of computer-based simulations in overcoming deep uncertainty
The power of human imagination simply cannot account for the huge variety of conditions that can influence future outcomes, nor give objective calculations of their likelihood. This is why organisations facing deep uncertainty turn to computer-based means of bolstering data intelligence and boosting confidence in decision-making, such as simulation and machine learning—computers do what the human brain can’t. The results of simulation-based data intelligence analysis provide a solid foundation for acknowledging and quantifying risks, and objective projections of future outcomes that can settle arguments, bypass sentiments, and help to fill gaps in knowledge. Simulation data thereby acts as an anchor for teams caught in deep uncertainty, providing much-needed decision support in contexts where experience and intuition aren’t enough to produce confidence.
Under deep uncertainty, any projection of future outcomes must contain a range of possibilities for what will actually occur. To handle this projection confidently, an organisation must have a good framework for taking action that not only drives towards a preferred outcome, but also includes contingency plans for dealing with variable conditions. One very effective means of achieving this is through a thorough exploration of simulated what-if scenarios that are parametrised according to an organisation’s real-world circumstances. Running these what-if scenarios at sufficient scale enables a clear assessment of the effects that a range of conditions could have on possible outcomes, from the most likely to the rarest. This allows a team to confidently quantify risks and to devise agile and adaptable strategies for managing them, based on simulations that are representative of real-life operational factors and procedures.
Running what-if scenarios to support decision-making is an iterative process, requiring a broad range of permutations to be processed in concert so as to ensure that the many possibilities of future outcomes are accounted for. A simulation may thus need to run many thousands of iterations of what-if scenarios to achieve robust data, factoring in variables such as weather conditions, human behaviour, material degradation and so forth. Besides designing simulation models with the right parameters and equations in place, complex what-if scenario explorations also require a lot of computing power to pull off at a satisfying scale and speed. Decision makers who are contending with time pressure and high stakes cannot always afford to wait a long time for results, requiring means of deploying and processing their what-if simulations quickly and efficiently in line with changing circumstances.
How distributed cloud computing helps achieve data robustness
Leveraging the power of distributed cloud computing allows what-if scenario simulations to process far larger numbers of permutations in parallel than would otherwise be possible, thereby reducing the time needed to achieve robust data. No longer relying on individual machines or localised servers means that access to convenient and cost-effective compute-power is near limitless. Moreover, cloud-native distributed applications can be accessed by a large number of users, with few hardware requirements on the client side, allowing simulations to be deployed more quickly and agilely than ever before. For decision makers making tough decisions under variable conditions and time constraints, this is a huge boon.
Hadean Orchestrate, our solution for decision support, uses dynamic provisioning to scale what-if scenario explorations automatically, on demand, according to parameters set by the end user. By bypassing the inefficient DevOps processes normally required to achieve this, the keys to decision support tools are handed straight to decision makers. The agnosticism of Hadean’s technology allows organisations to implement models of their own choosing, allowing for full human-in-the-loop parametrisation, and outputting results directly to their preferred data visualisation tools. This ensures that organisations have decision support tools that fit seamlessly into their existing processes, irrespective of the intricacies of their use case. With this in place, teams can leverage robust data intelligence to tackle high-stakes, time-sensitive instances of deep uncertainty quickly and assertively, and achieve far greater confidence in decision-making.