SimPilot
DocumentationExamplesAPI Reference
Sign in

Search Documentation

Find pages across the documentation

Getting Started

IntroductionQuickstartHow It Works

Core Concepts

AI AgentSimulation WorkflowEngineering PipelineKnowledge BaseArchitecture

Features

OverviewChat InterfaceCanvas & ArtifactsSharing & CollaborationExports & ReportsVoice InputMulti-Model AIProjectsTemplates & WorkflowsFile UploadsDeep ResearchWeb SearchCode ExecutionImage GenerationMCP ConnectorsInteractive DashboardsParallel ExecutionURL RetrievalFreeCAD CADSimulationsTemplatesConvergence Monitoring

Simulation

OverviewSupported TypesMesh GenerationError RecoveryBatch & SweepsResults Comparison

Studies & Analysis

OverviewDOE & Parametric SweepsOptimizationComparison

Validation & Reviews

OverviewBaselines & VersioningEngineering ReviewsRegulatory Compliance

3D Viewer

OverviewVisualization ToolsKeyboard Shortcuts

Enterprise

OverviewAdmin PanelOrganizationsKnowledge ManagementMethod Packs

Account

Getting StartedSettingsBilling

Examples

OverviewAerodynamicsPipe FlowHeat TransferStructural
  1. Docs
  2. Studies & Analysis
  3. Optimization

Optimization

Optimization studies with Pareto fronts, response surfaces, and convergence tracking.

Optimization studies find the best parameter values to minimize or maximize your engineering objectives. SimPilot runs simulations iteratively, learning from each result to guide the next experiment toward the optimum.

Single-objective optimization

Optimize one quantity at a time:
"Minimize the pressure drop across this heat exchanger by varying fin height (5-20 mm) and fin spacing (2-8 mm)"
The AI explores the parameter space, tracks the best result found so far, and converges toward the optimal configuration. You see a running best-so-far chart that shows improvement over iterations.

Multi-objective optimization

When objectives conflict -- lower drag vs. higher downforce, better heat transfer vs. lower pressure drop -- SimPilot finds the Pareto front: the set of solutions where no objective can be improved without worsening another.
"Optimize fin height and spacing to maximize heat transfer coefficient and minimize pressure drop"

Experiment execution

The AI runs simulations across the parameter space, evaluating both objectives for each configuration.

Pareto front construction

Non-dominated solutions are identified and plotted. Each point on the Pareto front represents a different trade-off between your objectives.

Trade-off analysis

The AI highlights key trade-off regions: the knee point (best balanced solution), the extremes (best for each individual objective), and clusters of similar solutions.

Response surfaces

From the collected simulation data, SimPilot generates response surfaces -- smooth approximations of how outputs vary across the parameter space. These let you:
  • Visualize the full optimization landscape in 3D
  • Identify flat regions (insensitive to parameter changes) vs. steep regions (highly sensitive)
  • Predict approximate results for untested parameter combinations
  • Spot local vs. global optima
Response surfaces are rendered as interactive 3D surface plots that you can rotate and zoom.

Convergence tracking

Throughout the study, a best-so-far chart tracks how the optimal value improves over iterations:
  • Rapid early improvement: The AI quickly finds good regions of the parameter space
  • Plateau detection: When improvement stalls, the AI adjusts its search strategy
  • Convergence criteria: The study concludes when improvement falls below a threshold or the iteration budget is reached

Recommended visualizations

Chart combinations for optimization studies
Optimization dashboards typically combine these chart types:
  • scatter3D: Pareto front with objective values on axes, colored by a third metric
  • line: Best-so-far convergence tracking over iterations
  • surface3D: Response surface over 2-parameter domains
  • parallel: All parameters and objectives on parallel axes -- trace solutions across the design space

Study configuration

You control the optimization by specifying:
SettingDescription
ParametersWhich inputs to vary and their ranges (min/max or discrete values)
ObjectivesWhich outputs to optimize and whether to minimize or maximize each
ConstraintsHard limits on outputs (e.g., "stress must stay below 250 MPa")
BudgetMaximum number of simulations to run
The AI handles everything else: sampling strategy, experiment sequencing, and result evaluation.
📐

DOE & Parametric Sweeps

Run a DOE study first to understand the parameter space before optimizing.
📊

Interactive Dashboards

Customize the optimization dashboard with additional chart types.
PreviousDOE & Parametric SweepsNextComparison

On this page

Single-objective optimizationMulti-objective optimizationResponse surfacesConvergence trackingRecommended visualizationsStudy configuration