Motional said it will be launching nuPlan, an expanded public dataset created to help self-driving system efficiency and safety.
In order to build safe performance in complex environments for self-driving systems, nuPlan contains a “virtual driving test” that measures the performance of a vehicle’s planning techniques. Motional, the autonomous-driving joint venture of Hyundai Motor Group and supplier Aptiv, calls this the world’s first benchmark for autonomous vehicle planning.
“The goal in planning is to teach the vehicle how to find its way through a dynamic environment full of obstacles and ever-changing circumstances — like a human driver does every day,” Holger Caesar, Motional’s Data Algorithm Teams lead, wrote in a blog post.
nuPlan will be available to the public late this year as an expansion of Motional’s nuScenes, which was released in March 2019.
Its predecessor has been downloaded by 12,000 students, researchers and developers and referenced in more than 600 academic publications, according to Caesar.
Caesar noted nuScenes ushered in an era in which almost all AV companies can share their datasets and advance the community.
Motional’s goal is to further this effort by providing a machine learning-based planning dataset.
“With the introduction of nuPlan, we hope that by providing a large-scale dataset and common benchmark, we will now pave a path towards progress in planning, which is perhaps one of the final frontiers in autonomous driving,” Caesar wrote.
