Automated Test Case Generation via Equivalence Class Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Manual generation of test cases for safety-critical software is time-consuming, and existing automated tools only achieve structural code coverage without efficiently testing representative and error-prone values at the requirements level.

Innovation Solution

The method involves automatically generating requirements-based test cases using equivalence class analysis, which differentiates valid and invalid equivalence classes, identifies boundary values, and selects tests from these classes to ensure thorough software behavior testing, employing both set-based and formal methods approaches.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If automated test generation tools are used, then test generation time is reduced, but test quality and completeness deteriorate because they only achieve structural code coverage without efficiently testing representative and error-prone values at the requirements level

Engineering Contradiction:
Improvetest generation timeVSAvoidtest quality
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent segments test generation into multiple specialized modules: equivalence class analysis unit, boundary value analysis unit, and error-prone value identification unit. Each module handles specific aspects of test case generation, allowing automated efficiency while maintaining comprehensive test quality through targeted analysis of representative values, boundary conditions, and error scenarios.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary layer between requirements and test cases that performs equivalence class analysis and boundary value identification. This intermediary automatically derives testable equivalence classes from requirements specifications, enabling automated generation of high-quality tests that cover representative values and error-prone conditions without manual intervention.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If manual test case generation is used, then test quality and completeness are improved by carefully selecting representative and error-prone values, but test generation time increases significantly

Engineering Contradiction:
Improvetest qualityVSAvoidtest generation time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent enables the system to perform self-service by automatically analyzing requirements specifications to identify equivalence classes, boundary values, and error-prone conditions without human intervention. The automated equivalence class analysis unit processes requirements and generates test cases independently, eliminating manual effort while maintaining the quality of representative value selection.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent transforms the test generation process by changing parameters from manual selection criteria to automated mathematical analysis of equivalence classes and boundary conditions. By parameterizing the analysis around equivalence class definitions and boundary value identification, the system achieves both automation and comprehensive test coverage.

Inventive Principle:
Principle #35Parameter changes

3Productivity

If conventional automated test generation tools are used, then test generation speed is improved, but test completeness deteriorates because they fail to differentiate valid and invalid equivalence classes and identify boundary values

Engineering Contradiction:
Improvetest generation speedVSAvoidtest completeness
Core Design Contradiction:
ProductivityVSLoss of information

Solution Approach 1:

The patent performs preliminary action by conducting equivalence class analysis and boundary value identification before actual test case generation. The system pre-processes requirements specifications to identify valid and invalid equivalence classes, boundary conditions, and error-prone values in advance, ensuring that subsequent test generation is both fast and complete.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent replaces conventional mechanical test generation mechanisms with automated mathematical analysis systems that perform equivalence class partitioning and boundary value identification. This substitution enables the system to automatically differentiate valid from invalid equivalence classes and identify boundary values, achieving both speed and completeness.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP3588304B1System and method for equivalence class analysis-based automated requirements-based test case generation
Publication Date: 2021.03.03 GENERAL ELECTRIC CO
  • EP3588304B1 patent drawingFigure 1
  • EP3588304B1 patent drawingFigure 2
  • EP3588304B1 patent drawingFigure 3

AI summary

A system for equivalence class analysis-based automated requirements-based test case generation includes a control processor, a data store containing textual design requirements, a textual converter unit structured to convert the textual design requirements to a machine-readable version of design requirements, a requirement partition unit configured to partition the machine-readable design requirements into one or more sets of related design requirements, an equivalence class partition unit configured to process the machine-readable design requirements and input/output variables into a set of equivalence classes, an equivalence class analyzer unit structured to analyze the set of equivalence classes to generate equivalence class tests and identify uncovered input space, and a boundary class analyzer unit structured to identify boundaries of the equivalence classes and generate boundary value tests and robustness tests. A method for equivalence class analysis-based automated requirements-based test case generation implementable on the system, and a non-transitory computer readable medium are also disclosed.