Abhishek Khardenavis

Building a comprehensive test suite for Autonomous Driving stacks using generative AI

Submitted Sep 26, 2025

Automotive software in general and autonomous driving systems in particular, are heavily reliant on Model-Based Software Engineering (MBSE) for their development. This approach necessitates a comprehensive and multi-layered testing process to ensure safety, reliability, and performance. Traditional test case creation can be a time-consuming and resource-intensive bottleneck in this workflow. Recognizing this, we are strategically integrating generative AI (GenAI) to automate and significantly accelerate the generation of these crucial test cases.

By leveraging GenAI, we aim to build a far more efficient development process for our autonomous driving stacks. The AI can intelligently create a diverse range of test scenarios, covering a wider breadth of potential situations than might be achievable through manual methods. This not only reduces development time but also enhances the robustness of the system by exposing it to a greater variety of challenging conditions during testing, ultimately contributing to a safer and more dependable autonomous driving experience.

Key Takeaways:

  • Approaches for streamlining extensive testing requirements using GenAI
  • Domain specific automated test case creation to increase efficiency while maintaining robustness

Target Audience:

  • Software Engineers and AI developers directly involved in automotive and autonomous vehicles software development.
  • Verification and Validation Engineers responsible for designing and executing tests for automotive systems.
  • Engineering Managers/Team Leads responsible for the overall development process efficiency and resource allocation.

About Speaker:

Ankur Deo is a Subject Matter Expert in Autonomous Driving at KPIT. His field of expertise involve Software-defined vehicles (SDVs), ADAS algorithm development and AI-ML for L2/L2+ features - accelerating software development and validation through GenAI, building AI models for collision anticipation and automotive sensor fusion. He holds a PhD in Machine Learning for Autonomous Vehicles from Coventry University.

Comments

{{ gettext('Login to leave a comment') }}

{{ gettext('Post a comment…') }}
{{ gettext('New comment') }}
{{ formTitle }}

{{ errorMsg }}

{{ gettext('No comments posted yet') }}

Hosted by

Jump starting better data engineering and AI futures