Robotaxi and Autonomous Vehicle Safety

Provides tools and guidelines to build a credible safety case from L2 ADAS to L4 robotaxis, drawing on 330+ research papers and 1,000+ patents within NVIDIA Halos.

Comprehensive evaluations with open-loop metrics, closed-loop simulation, and real-world vehicle tests demonstrate that Alpamayo 1 is state-of-the-art in multiple aspects (including reasoning, trajectory generation, alignment, safety, latency, and more).

Collecting and annotating real-world data for safety-critical physical AI systems, such as Autonomous Vehicles (AVs), is time-consuming and costly. To address this challenge, we introduce the Cosmos-Drive-Dreams, a synthetic data generation (SDG) pipeline for generating challenging scenarios to facilitate downstream tasks such as perception and driving policy training.

Safety remains a fundamental challenge in autonomous driving, with a key step being the development of a safety evaluator that can reliably identify unsafe (i.e., collision-prone) scenarios.

HydraSafe is a framework that addresses the challenge of ensuring autonomous vehicle safety in hazardous scenarios by improving data availability and planner robustness.

The advent of end-to-end autonomy stacks—often lacking interpretable intermediate modules—has placed an increased burden on ensuring that the final output, i.e., the motion plan, is safe in order to validate the safety of the entire stack.