InteractiveMap of Operations Research
Back to Uncertainty & Stochastic Systems
MethodAdvanced03.08
Uncertainty & Stochastic Systems

Simulation Optimization

Optimize systems whose performance is estimated by simulation.

Overview

Simulation Optimization focuses on optimize systems whose performance is estimated by simulation. In the map of OR, it connects Ranking & selection, Response surfaces, SAA to decisions that must be modeled, solved, explained, and revised as evidence changes.

Simulation optimization uses ranking and selection, response surfaces, stochastic approximation, and Bayesian optimization to choose policies in noisy models. The practical use case is clearest in Call centers, Hospitals, Warehouses, Manufacturing lines, where the method helps turn constraints and tradeoffs into a decision artifact someone can inspect.

Core ideas

Ranking & selection

Ranking & selection is a core checkpoint for Simulation Optimization: define it concretely, attach units or rules where possible, and test whether stakeholders interpret it the same way.

Response surfaces

Response surfaces is a core checkpoint for Simulation Optimization: define it concretely, attach units or rules where possible, and test whether stakeholders interpret it the same way.

SAA

SAA is a core checkpoint for Simulation Optimization: define it concretely, attach units or rules where possible, and test whether stakeholders interpret it the same way.

Bayesian optimization

Bayesian optimization is a core checkpoint for Simulation Optimization: define it concretely, attach units or rules where possible, and test whether stakeholders interpret it the same way.

How to use it

  1. 1Start with Call centers: write the decision, time horizon, actors, and objective in operational language.
  2. 2Translate the problem into Ranking & selection, Response surfaces, and SAA; define units and data sources for each one.
  3. 3Build a small instance of Simulation Optimization that can be solved or simulated by hand inspection before using full production data.
  4. 4Compare the recommendation against a baseline policy, not just against mathematical optimality.
  5. 5Document assumptions, sensitivity results, and the conditions under which the recommendation should be revisited.

Applications

Call centersHospitalsWarehousesManufacturing lines
  • Call centers: compare feasible policies, quantify the operating tradeoffs, and make the assumptions behind the recommendation visible.
  • Hospitals: compare feasible policies, quantify the operating tradeoffs, and make the assumptions behind the recommendation visible.
  • Warehouses: compare feasible policies, quantify the operating tradeoffs, and make the assumptions behind the recommendation visible.
  • Manufacturing lines: compare feasible policies, quantify the operating tradeoffs, and make the assumptions behind the recommendation visible.

Common pitfalls

  • Applying Simulation Optimization because the label sounds appropriate while leaving the actual decision boundary vague.
  • Treating Ranking & selection as a technical detail instead of a modeling choice that affects the recommendation.
  • Reporting one answer without showing sensitivity to demand, capacity, costs, or behavioral assumptions.
  • Ignoring implementation details such as data quality, explainability, ownership, and how users will override bad recommendations.

Resources