Project T : TestReducer - Algebraic Optimization of Engine Test Matrices Using Hecke Monoid Reduction

Read full text
Authors: Stéphane François LEFORT
Cited by: 0

Abstract

TestReducer is a software tool that applies algebraic properties from the Hecke monoid H₀(Ã₂) to optimize engine test matrices for automotive, aerospace, and marine applications. PROBLEMEngine dynamometer testing is expensive and time-consuming. A typical engine calibration involves 1,000-10,000 test points, with costs of €500-1,500 per hour. Analysis shows that 40-70% of test points in standard DoE (Design of Experiments) matrices are redundant, interpolable, or low-value. SOLUTIONTestReducer leverages two fundamental properties of the Hecke monoid:- Idempotence (a² = a): Same test conditions executed twice = single test- Braid relation (aba = bab): Some test sequences are equivalent regardless of order Combined with spatial interpolation analysis, these properties enable automatic identification of redundant tests. REDUCTION RULESThe tool implements six reduction rules:- TR001 - Idempotence: Detect duplicate test conditions (binned for tolerance)- TR002 - Interpolation: Mark interior points derivable from neighbors- TR003 - Transitivity: System tests predictable from component tests- TR004 - Equivalence: Detect equivalent test sequences- TR005 - Tolerance: Filter points within measurement uncertainty- TR006 - Prioritization: Rank by safety, certification, customer usage RESULTSIn benchmark testing with a 10×10 RPM/Load grid:- Input: 135 test points- Output: 38 test points- Reduction: 72%- Dyno time saved: 3.9 hours- Cost saved: €3,104 TECHNICAL DETAILS- Input formats: CSV (DoE matrices), JSON- Output formats: CSV, JSON, TXT reports- Dependencies: Python 3.8+ (standard library only)- License: MIT APPLICATIONS- Automotive engine development and calibration- Aerospace propulsion testing- Marine engine certification- HIL (Hardware-in-the-Loop) test optimization- Emissions certification campaigns This tool is part of Project T, exploring applications of algebraic idempotence properties to engineering verification domains including Infrastructure as Code, BIM clash detection, and engine testing.

Fields of Study & Subjects

Optimal Experimental Design Methods Engine and Fuel Emissions Technical Engine Diagnostics and Monitoring

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
25 January 2026
Language
EN