Automated Scenario Generation for Autonomous Robots using Large Language Model

The development of autonomous robots has increasingly emphasized the importance of identifying potential issues early through efficient validation and simulation-based testing. Despite this progress, significant challenges remain in automating test scenario generation and addressing the diversity and complexity of testing environments. This study introduces a method for scenario-based testing using the Hakoniwa platform, which supports IoT and robot software development in a virtual environment, in combination with Large Language Models (LLM) to achieve a more precise drone simulation tailored to user requirements.