Creating Test Scenarios by Natural Language Input

dSPACE Works with AWS to Drive Innovation in Autonomous Driving: Unleashing the Potential of Generative AI for Advanced Scenario Generation

Paderborn, June 24, 2024. dSPACE, one of the world's leading solution providers for simulation and validation of automated and connected vehicles, and an Amazon Web Service (AWS) Automotive Competency Partner, is developing an advanced scenario generation solution by leveraging generative AI technologies from Amazon Web Services (AWS). 

The new scenario generation feature, which dSPACE is developing with support from the AWS Generative AI Innovation Center, will be integrated into the dSPACE portfolio for SIL/HIL testing and data-driven development. It is based on a large language model (LLM) available on Amazon Bedrock and retrieval-augmented generation (RAG) architecture, which leverages scenario data provided by dSPACE. As a result, users will be able to easily create a vast number of complex and realistic scenarios, including edge cases according to the ASAM OpenSCENARIO standard, simply by entering a textual scenario description. This allows automotive manufacturers and suppliers to simulate and validate their autonomous driving systems more comprehensively and efficiently, reducing the time and cost associated with physical prototyping and testing.

"dSPACE is an AWS Automotive Competency Partner. In this new initiative, we are creating innovative capabilities in scenario generation for autonomous vehicle validation. Using generative AI technologies, we ensure that our customers can easily create realistic and diverse simulation scenarios to develop safer and more reliable autonomous vehicles," says André Rolfsmeier, Director of Strategic Product Management at dSPACE. 

dSPACE's close collaboration with AWS and the integration of generative AI technologies into dSPACE products pave the way for more comprehensive and efficient validation pipelines, ultimately contributing to safer autonomous vehicles. 

dSPACE is demonstrating these capabilities for the first time to the market at the AWS Automotive Symposium in Munich on June, 26.