A Generative AI-driven Application: Use of Large Language Models for Traffic Scenario Generation

Artificial Intelligence (AI) has revolutionized the way we interact with technology. One of its forefronts is Natural Language Processing (NLP), an essential sub-field that focuses on human language understanding and generation. Large Language Models (LLMs) such as PaLM, BERT, GPT have emerged as the powerhouses of NLPs with remarkable capabilities. Today, we also utilize AI in other industries like automotive. While using AI-backed powerful instruments in autonomous vehicle technologies, we are still suffering to create virtual traffic scenarios for special purposes such as safety testing and validation, autonomous driver training and regulatory compliance. In this study, we introduce a solution which harnesses the power of LLMs to overcome scenario creation difficulties. Thus, our approach provides creation of the traffic scenarios in linguistic way rather than using user interfaces of the editors or writing special formatted structured scenario files.