DoE (Design of Experiments) for Calibration - Calibration & XCP Protocol

DoE Rationale: Efficient Operating Space Coverage

A 4-factor full-factorial experiment with 5 levels each requires 5^4 = 625 test points. A Latin Hypercube Design (LHD) with 60 points can cover the same space and still fit a response surface model with quantified uncertainty — a 10× reduction in dyno time while maintaining statistical model quality.

DoE vs One-Factor-at-a-Time Comparison

  One-Factor-at-a-Time (OFAT):           Latin Hypercube Design (LHD):
  Fix RPM=2000, sweep EGR 0-40%          60 points: RPM and EGR varied
  Fix EGR=20%, sweep RPM 1000-5000       simultaneously, space-filling
  Fix both, sweep load                   No factor held constant
                                          Detects interactions between factors

  OFAT: 3 × 20 points = 60 test points  LHD: 60 points
  OFAT: misses RPM-EGR interaction       LHD: captures full joint response
  OFAT: cannot fit 2nd-order model       LHD: supports Gaussian Process model

Calibration Design Types

Design Type	Best For	Factors	Points Needed
Full Factorial	≤3 factors, full enumeration	≤3	levels^factors (e.g., 5^3 = 125)
Central Composite Design (CCD)	2nd-order response surface, curvature important	3–6	2^k + 2k + centre (e.g., k=4: 25 points)
Latin Hypercube (LHD)	No prior model, general exploration	Any	50–200 recommended
D-Optimal	Model-based, minimise coefficient variance	Any	Depends on model order; tool calculates minimum
Space-Filling Maximin	Maximum spread in factor space	Any	User-specified; typically 3-5× number of model parameters

ETAS ASCMO DoE Workflow

Pythonascmo_doe_workflow.py

# ETAS ASCMO API: DoE design, test execution, model fitting
import ascmo_api as ascmo

# 1. Define design space (factor ranges)
design_space = ascmo.DesignSpace([
    ascmo.Factor("engine_speed_rpm", 800,  5000),
    ascmo.Factor("engine_load_pct",  10,   100),
    ascmo.Factor("egr_valve_pct",    0,    40),
])

# 2. Generate Latin Hypercube test plan (50 points)
test_plan = ascmo.LatinHypercube(design_space, n_points=50, seed=42)
test_plan.export_inca("ascmo_test_plan.inca_sequence")
print(f"Generated {len(test_plan.points)} test points")

# 3. Execute test plan on dyno (INCA records mf4 automatically)
# [dyno operator runs INCA sequence — each point: set RPM/load/EGR, record 30s steady-state]

# 4. Import mf4 results and fit Gaussian Process model
results = ascmo.ImportResults("dyno_run_001.mf4",
    responses=["NOx_ppm", "HC_ppm", "fuel_cons_l100km"])
model = ascmo.FitGaussianProcess(test_plan, results)

# 5. Validate model with holdout set (10 points not used in fitting)
validation = model.Validate(test_plan.holdout_points, results.holdout_responses)
print(f"NOx model RMSE: {validation['NOx_ppm'].rmse:.1f} ppm")
print(f"HC  model RMSE: {validation['HC_ppm'].rmse:.1f} ppm")

Model-Based Calibration: Pareto Optimisation

Pythonpareto_optimisation.py

# ASCMO Pareto optimisation: minimise NOx AND fuel consumption simultaneously
import ascmo_api as ascmo
import numpy as np

model = ascmo.LoadModel("engine_gp_model.ascmo")

# Define dual objectives and constraints
objectives = [
    ascmo.Objective("NOx_ppm",          minimize=True,  weight=1.0),
    ascmo.Objective("fuel_cons_l100km", minimize=True,  weight=1.0),
]
constraints = [
    ascmo.Constraint("HC_ppm",          "<=", 150),   # Euro 6 limit
    ascmo.Constraint("engine_speed_rpm",">=", 800),   # idle stability
]

# Run multi-objective Pareto optimisation
pareto_result = model.ParetoOptimise(
    objectives=objectives,
    constraints=constraints,
    n_pareto_points=100,
    method="NSGA-II"  # Non-dominated Sorting Genetic Algorithm
)

# Display Pareto front
print("Pareto-optimal solutions:")
print(f"{'NOx (ppm)':<15} {'Fuel (L/100km)':<18} {'EGR valve (%)'}")
for pt in pareto_result.pareto_front[:5]:
    print(f"{pt['NOx_ppm']:<15.1f} {pt['fuel_cons_l100km']:<18.2f} {pt['egr_valve_pct']:.1f}")

# Validate winning point on dyno
winner = pareto_result.select_by_weighted_sum(nox_weight=0.6, fuel_weight=0.4)
print(f"\nSelected operating point: EGR={winner['egr_valve_pct']:.1f}%, "
      f"predicted NOx={winner['NOx_ppm']:.0f} ppm, fuel={winner['fuel_cons_l100km']:.2f} L/100km")
print("Validate on dyno before deploying to production calibration")

Summary

DoE replaces exhaustive testing with statistically designed experiments that cover the operating space efficiently. A 50-point Latin Hypercube Design can characterise a 3-factor engine response that would require 125 full-factorial test points — without sacrificing model accuracy. The Gaussian Process model fitted in ASCMO enables Pareto optimisation across competing objectives (NOx vs. fuel consumption), producing the optimal calibration map set that no manual dyno tuning process can match for efficiency or statistical rigor.

🔬 Deep Dive — Core Concepts Expanded

This section builds on the foundational concepts covered above with additional technical depth, edge cases, and configuration nuances that separate competent engineers from experts. When working on production ECU projects, the details covered here are the ones most commonly responsible for integration delays and late-phase defects.

Key principles to reinforce:

Configuration over coding: In AUTOSAR and automotive middleware environments, correctness is largely determined by ARXML configuration, not application code. A correctly implemented algorithm can produce wrong results due to a single misconfigured parameter.
Traceability as a first-class concern: Every configuration decision should be traceable to a requirement, safety goal, or architecture decision. Undocumented configuration choices are a common source of regression defects when ECUs are updated.
Cross-module dependencies: In tightly integrated automotive software stacks, changing one module's configuration often requires corresponding updates in dependent modules. Always perform a dependency impact analysis before submitting configuration changes.

🏭 How This Topic Appears in Production Projects

Project integration phase: The concepts covered in this lesson are most commonly encountered during ECU integration testing — when multiple software components from different teams are combined for the first time. Issues that were invisible in unit tests frequently surface at this stage.
Supplier/OEM interface: This is a topic that frequently appears in technical discussions between Tier-1 ECU suppliers and OEM system integrators. Engineers who can speak fluently about these details earn credibility and are often brought into critical design review meetings.
Automotive tool ecosystem: Vector CANoe/CANalyzer, dSPACE tools, and ETAS INCA are the standard tools used to validate and measure the correct behaviour of the systems described in this lesson. Familiarity with these tools alongside the conceptual knowledge dramatically accelerates debugging in real projects.

⚠️ Common Mistakes and How to Avoid Them

Assuming default configuration is correct: Automotive software tools ship with default configurations that are designed to compile and link, not to meet project-specific requirements. Every configuration parameter needs to be consciously set. 'It compiled' is not the same as 'it is correctly configured'.
Skipping documentation of configuration rationale: In a 3-year ECU project with team turnover, undocumented configuration choices become tribal knowledge that disappears when engineers leave. Document why a parameter is set to a specific value, not just what it is set to.
Testing only the happy path: Automotive ECUs must behave correctly under fault conditions, voltage variations, and communication errors. Always test the error handling paths as rigorously as the nominal operation. Many production escapes originate in untested error branches.
Version mismatches between teams: In a multi-team project, the BSW team, SWC team, and system integration team may use different versions of the same ARXML file. Version management of all ARXML files in a shared repository is mandatory, not optional.

📊 Industry Note

Engineers who master both the theoretical concepts and the practical toolchain skills covered in this course are among the most sought-after professionals in the automotive software industry. The combination of AUTOSAR standards knowledge, safety engineering understanding, and hands-on configuration experience commands premium salaries at OEMs and Tier-1 suppliers globally.