31 Mar 2020
Taking the stress out of critical decision-making
Technology developed by tpgroup has the potential to guide complex decisions relating to capacity planning and availability of critical care assets in the health sector.
Clinicians and healthcare managers are faced with a continuous stream of difficult choices when critical resources become overloaded. There may never be a true ‘right answer’ but having clarity in how one gets to the decision leads to greater confidence in the outcomes. Building a picture of the factors that contribute to the decision, gathering good quality information, applying some discipline to the decision pathway and doing this consistently are all positive steps.
There are two ways to chase the ‘right answer’ in a decision process. One explores all the possible options within a set of variables and constraints. These approaches produce mathematically provable optimum solutions, but typically take many hours to compute.
The alternative is a heuristic approach that uses a semi-random, guided exploration of a decision space. In complex scenarios, a metaheuristic approach can explore multiple dimensions simultaneously. In an iterative process, it ‘tries out’ possible combinations to find a balanced optimisation across multiple factors. The outcome is very close to the mathematical optimum but is discovered in a fraction of the time and can adjust as the situation changes. This means that the model can be:
- rerun for changing conditions/circumstances on month-month, day-day, hour-hour basis
- run multiple times to explore different (what-if) scenarios
- adjust run time (confidence versus time) to balance criticality and immediacy of decision
The metaheuristic approach allows us to promote pragmatism over perfection.
In the oil and gas sector, tpgroup is working with upstream producers to optimise their critical assets. The team is building a digital model that reflects a collection of offshore production rigs. A range of factors that contribute to the demand for their output, their production efficiency, availability and the strategic landscape. A range of options is then considered using a tool called Optimiser, an artificial intelligence (AI) based data-driven decision support system, that uses a metaheuristic approach (as opposed to an exact method) to rapidly explore large-scale, multidimensional decision spaces.
The tpgroup team believes this could be applied in other sector such as healthcare, where complex situations evolve over time. The goal is to work with subject matter experts who understand the factors that make up a situation – and understand the relationships between them. With data combined as a model of the real-world position, recommendations can be made by the AI system and then reviewed and applied where suitable by decision makers.
This arrangement would build consistency and repeatability in the decision process and reduce the cognitive burden on those at the front line. In practice, it could be applied to:
- dynamic replanning and reallocation of scarce resources across the enterprise for day-to-day operations
- business continuity planning – assessing enterprise resilience to crisis management scenarios
- long-term planning and allocation of resources across the enterprise
- organisational design – determining the optimal investment in new and/or modified enterprise resources
This is all part of the bigger challenge to build an informatics architecture for integrated critical care. Attempts to build an information model that guides specific decisions are positive steps towards that goal – and each case can contribute to the bigger picture as they are added.
If you would like more information on this exciting challenge, or would like to talk about how these ideas might be made possible, please contact firstname.lastname@example.org.
Please click here for more information on Optimiser.
We are a provider of game-changing solutions for a safer world. As a trusted partner to the UK Government, we support highly regulated mission and safety-critical programmes as well as delivering the assured availability of naval submarine life-support systems for over 40-years.