Software regulators in electric vehicles use thermal management systems to balance energy efficiency, heating, and cooling. A virtual control unit is being developed to simulate and test this complex system before full vehicle testing is done.

Richard Seidl Software Testing

This is the other podcast on Software Testing by Richard Seidl, the episodes are in spoken German but the show notes and site are written in English. Our summaries are generated from AI transcript translations.

• https://www.richard-seidl.com/en/blog/tag/podcast-software-testing
• https://www.richard-seidl.com/en/

Episode Details

• Show Notes: https://www.richard-seidl.com/de/blog/thermomanagement-testen
• Published: 2026-01-06T05:00:00Z
• Duration: 1100
• Author: Richard Seidl - Experte fur Software-Entwicklung und Testautomatisierung

Overview

The podcast outlines the challenges involved in testing thermal management software in electric vehicles, highlighting the importance of electronic control units (ECUs) in regulating heating and cooling systems. Factors such as energy efficiency, vehicle range, and varying operating conditions must be considered during testing. To address these challenges more efficiently, the team has shifted from traditional on-vehicle testing to a “Shift Left” strategy, utilizing hardware-in-the-loop (HIL) systems and virtual control units (VECU) to simulate real-world conditions early in the development process. This approach helps reduce costs and streamline development.

A virtual ECU is being developed to specifically test thermal management functions, reducing dependence on the more expensive HIL system. Automated testing tools and structured test case management are used to cover a wide range of scenarios, from simple to complex. Techniques such as simulated timers and filtering mechanisms allow for accurate simulation of long test runs and temperature changes. While virtual testing is extensive, physical testing is still required to validate critical system behaviors. Additionally, the team is investigating the use of AI-generated test specifications to further improve automation and efficiency in their testing procedures.

What If

• What if you applied a virtual control unit (VECU) to simulate only your core software component for early testing?

  • Concrete move: Develop a VECU that mimics the behavior of your software component (e.g., thermal logic) in isolation using simulation tools like MATLAB/Simulink or cloud-based HIL platforms.
  • Why now: Hardware costs for full system testing are prohibitive for solo operators, but VECU reduces reliance on physical setups, enabling rapid iteration.
  • Expected upside: Early defect detection, reduced rework costs, and faster time-to-market by decoupling component testing from unrelated system dependencies.

• What if you leveraged AI-generated test specifications to automate test creation for your software?

  • Concrete move: Use a Large Language Model (LLM) to generate test specs from your functional requirements, then automate these tests with frameworks like PyTest or Jenkins.
  • Why now: AI tools are accessible (e.g., GitHub Copilot, OpenAI Codex), and your project likely has structured requirements that LLMs can parse effectively.
  • Expected upside: 5070% faster test development, reduced manual effort, and consistent test coverage aligned with system requirements.

• What if you implemented shift-left testing using hardware-in-the-loop (HIL) simulations for your MVP?

  • Concrete move: Partner with cloud-based HIL providers (e.g., dSPACE, Speedgoat) to simulate real-world conditions (e.g., extreme temperatures) for your softwares core functionality.
  • Why now: HIL costs have decreased, and cloud access allows solo operators to simulate hardware interactions without physical prototyping.
  • Expected upside: Identify integration flaws early, validate edge cases (e.g., battery cooling priorities), and reduce costly late-stage rework by 4060%.

Takeaway

• Implement shift-left testing using hardware-in-the-loop (HIL) or virtual environments to simulate real-world conditions (e.g., extreme temperatures) early in development, reducing costs and enabling earlier error detection before full vehicle testing.
• Develop a virtual control unit (VECU) for isolated component testing (e.g., thermal management) to avoid using costly full-scale physical test systems like the Hill cabinet, allowing focused validation of specific functionalities.
• Automate test cases and reuse them across physical and virtual environments (e.g., HIL and VECU) by adapting test tools and configurations, minimizing redundant efforts while ensuring consistency.
• Structure requirements meticulously for auto-generated test specifications using Large Language Models (LLMs), ensuring clear, hierarchical input to produce accurate automated tests for critical functionalities.
• Prioritize functional testing over controller quality in early stages by focusing on core behaviors (e.g., battery cooling prioritization) and using simulated timers to handle long-running processes like vehicle preconditioning.