Systems Modelling, Analysis and Trade-off Evaluation
Domain 6 tests modeling, analysis, and trade-off evaluation: selecting a model for the question, validating it for use, and comparing alternatives against explicit criteria and uncertainty.
Core rules and concepts
- A model is an abstraction for a purpose. Fidelity must match the decision, not the ambition of the modeler.
- Trade studies define objectives, measures, weights, alternatives, and uncertainty. Sensitivity analysis tests whether the ranking is robust.
- Analysis types include functional, performance, reliability, safety, and cost. They answer different questions and need different evidence.
- Model validation asks whether the model is adequate for the intended use, not whether it is universally "true."
Worked example
A remote sensor station needs power. Alternatives are battery, solar-battery, and fuel cell. Criteria are through-life cost, mass, maintenance interval, and environmental risk. Solar-battery wins on cost and maintenance but loses in polar winter. Sensitivity analysis on latitude shows the ranking reverses below a threshold. Reasoning: the trade-off is conditional, not absolute. The decision record states the winning alternative and the condition under which it wins.
Traps that cost marks
- Optimizing on a single criterion and calling it a trade study.
- Hiding weights or choosing them after seeing results.
- Ignoring uncertainty and sensitivity.
- Using a model without stating assumptions and limits.
- Comparing alternatives that do not meet the same constraints.
Key points
- Model purpose drives fidelity.
- Trade studies need explicit criteria and weights.
- Sensitivity tests robustness.
- Validate models for intended use.
- Record conditions, not just winners.