AImotive, MathWorks team to improve autonomous vehicle simulation

The partners are integrating MathWork’s RoadRunner 3D scene-editing capabilities into AImotive’s aiSim to enable more accurate 3D scenes for simulations.
A simulator must include a high-fidelity environment and realistic sensor simulation to enable end-to-end testing of automated driving software. To enable this, the simulator needs precise, well-defined maps and quality assets as inputs.

To address this challenge, leading developer of mathematical computing software MathWorks and Europe’s well-funded automated driving technology supplier AImotive have begun collaborating to integrate RoadRunner into aiSim. This collaboration will allow users to easily create road and 3D networks for simulations into aiSim, the world’s first ISO 26262-certified automotive-grade simulation platform — said to be the most deterministic and robust solution on the market.

The result of the collaboration provides a user-friendly way to create various environments using RoadRunner’s intuitive map editor, ranging from dense urban scenes to compliance-specific test cases. The integration allows users to create and edit OpenDrive maps, import GIS (geographic information system) data, and place the built-in high-fidelity assets of aiSim mixed with RoadRunner’s extensive asset library. The created content can be readily imported to the ASIL-D certified simulator, exercising its advanced sensor simulation capabilities.

“Automated driving engineers use virtual testing to reduce the cost and risk of in-vehicle testing,” said Avinash Nehemiah, Principal Product Manager — Autonomous Systems, at MathWorks. “To achieve the most benefit from virtual testing, eng
Click on the link below to read the full post online.