# Autonomous Vehicle Safety Ensure vehicle safety across the full AV stack with NVIDIA Halos. [Read the Safety Report](https://images.nvidia.com/aem-dam/en-zz/Solutions/auto-self-driving-safety-report.pdf) Overview ## Autonomous Vehicle Safety, From the Cloud to the Car NVIDIA Halos is a full-stack, comprehensive safety system that unifies vehicle architecture, AI models, chips, software, tools, and services to ensure the safe development and deployment of autonomous vehicles (AVs) from cloud to car. The system covers the full development lifecycle with design-time, deployment-time, and validation-time guardrails that collectively build safety and explainability into AI-based AV stacks. These guardrails are implemented using three powerful computers—[NVIDIA DGX](https://www.nvidia.com/en-us/data-center/dgx-platform.md)™ for AI training, [NVIDIA Omniverse](https://www.nvidia.com/en-us/omniverse.md)™ and [Cosmos](https://www.nvidia.com/en-us/ai/cosmos.md)™ for [simulation](https://www.nvidia.com/en-us/use-cases/autonomous-vehicle-simulation.md), and [NVIDIA AGX](https://www.nvidia.com/en-us/self-driving-cars/in-vehicle-computing.md)™ for deployment. At the heart of the vehicle, NVIDIA Halos OS provides the unified software foundation necessary to bridge these AI capabilities with production-ready safety. NVIDIA Halos complements existing industry-standard safety practices, while introducing unique elements for autonomous vehicles. This ensures regulatory compliance and advances safe and reliable AV stacks, together with [NVIDIA’s Halos AI Systems Inspection Lab](https://www.nvidia.com/en-us/ai-trust-center/physical-ai/safety-certification.md). Halos is also extending its comprehensive safety framework beyond AVs to [robotics](https://blogs.nvidia.com/blog/european-robot-makers-isaac-omniverse-halos-safe-physical-ai), further enhancing the reliability and safety of intelligent systems. ### NVIDIA Releases New AI Models and Developer Tools to Advance Autonomous Vehicle Ecosystem To accelerate the development of next-generation AV architectures, NVIDIA released NVIDIA Cosmos Predict-2 — a new world foundation model with improved future world state prediction capabilities for high-quality synthetic data generation — as well as new developers tools. [Read the Blog](https://blogs.nvidia.com/blog/autonomous-vehicle-ecosystem-ai-models-developer-tools) ### NVIDIA Launches Halos, a Full-Stack, Comprehensive Safety System for Autonomous Vehicles NVIDIA unifies vehicle architecture to AI models; chips, software, and tools to services for safely developing AVs from cloud to car. [Read the Blog](https://blogs.nvidia.com/blog/halos-safety-system-autonomous-vehicles) Highlights ## Autonomous Vehicle Safety Leadership NVIDIA Halos is the result of continuous investment in autonomous vehicle safety—from research to engineering to active engagement with international safety standards—validated by independent third-party assessments. ### 18,600+ Engineering years invested in vehicle safety to date ### 21 Billion+ Safety transistors safety assessed ### 7,000,000 Lines of safety-assessed code ### 2,000,000 Daily end-to-end integration tests for validation ### 22,000+ Platform safety monitors ### 20,000+ Hours of safety test data ### 1,000+ Patents filed ### 330+ Research papers published on AV safety ### 30+ Certificates and assessment reports issued Technology ## Engineered for Safety, Designed for Trust As AV companies transition to AI-based, end-to-end architectures, NVIDIA Halos provides the critical safety foundation to ensure system-level reliability and iterative improvement for automated driving systems. This includes integration of third party-assessed hardware, software, and processes with a diverse algorithmic architecture and validation pipelines. ### NVIDIA DGX **Design-time safety guardrails** for built-in hardware/software safety and trustworthy development processes. [Learn More](https://www.nvidia.com/en-us/self-driving-cars/ai-training.md) ### NVIDIA Omniverse With Cosmos **Validation-time guardrails** for data generation, simulation, evaluation, and lifelong safety assurances. [Learn More](https://www.nvidia.com/en-us/use-cases/autonomous-vehicle-simulation.md) ### NVIDIA DRIVE AGX™ **Deployment-time guardrails** for run-time monitoring and real-time introspection. [Learn More](https://www.nvidia.com/en-us/self-driving-cars/in-vehicle-computing.md) Benefits ## A Comprehensive System for Autonomous Vehicle Safety NVIDIA Halos helps to ensure AI-driven AV systems are safe and secure. Partners can tap into NVIDIA’s investments in AI safety to accelerate development and enhance AV reliability. NVIDIA Halos is also open to developers, enabling adoption or customization of safety elements to drive the shared mission to create safe and reliable autonomous vehicle technology. Design-time, deployment-time, and validation-time guardrails collectively build safety and explainability into several layers of technologies spanning platform safety, AI algorithmic safety, and ecosystem safety. At the top of the NVIDIA Halos elements sits the [NVIDIA Halos AI Systems Inspection Lab](https://www.nvidia.com/en-us/ai-trust-center/physical-ai/safety-certification.md), which allows customers and ecosystem partners to verify the safe integration of their products with NVIDIA Halos elements. The lab is the first worldwide program to be accredited by ANAB for AI functional safety. Use Cases ## A Full-Stack Comprehensive Safety System NVIDIA Halos integrates foundational models and a diverse algorithmic stack, combining classical and AI-based, end-to-end models to drive system-level safety in the shift toward AI-driven AV architectures. [Explore NVIDIA Halos Product Brief](https://www.nvidia.com/content/dam/en-zz/Solutions/self-driving-cars/nvidia-halos-product-brief.pdf) 1. ### Platform Safety 2. ### Algorithmic Safety 3. ### Ecosystem Safety 4. ### Halos AI Systems Inspection Lab ## Platform Safety The robust foundation for autonomous driving systems includes: * A [System-on-a-Chip](https://www.nvidia.com/en-us/solutions/autonomous-vehicles/in-vehicle-computing.md) (SoC) that’s designed for safety, with hundreds of built-in safety mechanisms. * NVIDIA Halos OS, a unified, three-layer safety foundation built on ASIL D certified [NVIDIA DriveOS™](https://developer.nvidia.com/drive/os). The architecture integrates Halos Core (safety operating system), Halos SDK (safety middleware), and Halos applications (safety applications) to deliver a production-ready environment for AV. * A safety-assessed base platform that delivers the foundational safe computer needed to enable safe systems for all types of applications * [NVIDIA DRIVE](https://www.nvidia.com/en-us/self-driving-cars/in-vehicle-computing.md) Hyperion™, a diverse hardware platform that connects the SoC, OS, and sensors in a vehicle architecture. This enables a vehicle to safely execute contingency plans if needed. ## Algorithmic Safety Algorithmic AI safety spans: * A diverse AV stack that combines a modular stack and [NVIDIA Alpamayo](https://www.nvidia.com/en-us/solutions/autonomous-vehicles/alpamayo.md) reasoning VLA models for algorithmic AI safety. * Training, simulation, and validation environments that use NVIDIA Omniverse and Cosmos platforms to build safe AVs. * A separate safety dataset that ensures AV performance is tested against diverse, unbiased data. ## Ecosystem Safety Building a safer AV ecosystem includes: * Continual improvements through a safety data flywheel, which continually learns from the road how to expand the set of operational design domains for safe deployment. * Seamless integration of physically based and diverse sensor simulation into existing workflows to safely train, test, and validate AVs with the NVIDIA Omniverse Blueprint for AV Simulation. * Open-source data, such as the [NVIDIA Physical AI Dataset](https://huggingface.co/collections/nvidia/physicalai-67c643edbb024053dcbcd6d8), to enable critical safety research throughout the industry. * A growing list of partners using the NVIDIA Halos system, listed [here](https://www.nvidia.com/en-us/solutions/autonomous-vehicles/partners.md). ## Halos AI Systems Inspection Lab NVIDIA is the first company to establish an ANSI National Accreditation Board (ANAB)-accredited [Halos AI Systems Inspection Lab](https://www.nvidia.com/en-us/ai-trust-center/physical-ai/safety-certification.md), integrating functional safety, cybersecurity, AI, and regulations into a unified safety framework. The lab helps to ensure that member system integrations meet rigorous safety and cybersecurity standards through impartial assessments. By providing inspection reports and streamlining technical validations, the lab accelerates compliance with global safety standards for AV safety and cybersecurity. This empowers the automotive ecosystem to deploy safer, more reliable AI-driven technologies while advancing compliance with international standards. The NVIDIA Halos AI Systems Inspection Lab has now expanded from AV to robotics. [Learn More About the AI Systems Inspection Lab](https://www.nvidia.com/en-us/ai-trust-center/physical-ai/safety-certification.md) “Joining NVIDIA's Halos AI Systems Inspection Lab marks our commitment to advancing driving safety. By combining Bosch's comprehensive in-house ADAS sensor expertise with NVIDIA's AI validation framework, we're setting new standards for safe and reliable ADAS solutions.” **— Dennis Raabe,** Senior Vice President ADAS Components, **Bosch** “NVIDIA’s latest evaluation with ANAB verifies the demonstration of competence and compliance with internationally recognized standards, helping ensure that developers of autonomous machines—from automotive to robotics—can meet the highest benchmarks for functional safety.” **— R. Douglas Leonard Jr.,** Executive Director, **ANAB** “Aligned with our shared mission to enhance safety, efficiency and productivity, we congratulate NVIDIA on the launch of its Halos AI Systems Inspection Lab. The Lab's reports will provide valuable insights to support our certification efforts.” **— Thomas Steffens,** Head of Certification Body Functional Safety and Cybersecurity, **TUV Rheinland** “We are pleased to hear of NVIDIA's commitment to advancing autonomous vehicle safety and welcome NVIDIA's Halos Inspection Lab efforts for structured and thorough development of AI for safety-relevant applications.” **— Dominik Strixner,** Global Lead Functional Safety Auto and Mobility, **TÜV Rheinland (automotive)** “UL Solutions is a global leader in applied safety science. We are pleased to announce our intent to collaborate with the NVIDIA Halos AI Systems Inspection Lab to harmonize testing activities and reports for companies who are pursuing safety certification with us.” **— Alex Dadakis,** EVP, Chief Business Ops and Innovation, **UL Solutions** “We are committed to fostering digital trust and delivering rigorous AI assurance at scale. By recognizing the inspection reports of the NVIDIA Halos AI Systems Inspection Lab, we’re supporting the industry’s move toward more transparent, reliable, and secure AI — while enabling developers to bring innovative systems to market faster and more safely.” **— Vincent Sabot,** CEO, **CertX** Certification ## Assessed by Experts Independent third-party safety and cybersecurity assessments of NVIDIA Halos elements demonstrate NVIDIA’s ongoing commitment to AV safety. ### ANSI National Accreditation Board ANAB accredited the [NVIDIA Halos AI Systems Inspection Lab](https://blogs.nvidia.com/blog/drive-ai-lab-ces/) as an ISO/IEC 17020 Inspection Body. NVIDIA is the first company accredited by ANAB for an inspection plan that combines cybersecurity, AI, and functional safety. ### TÜV SÜD TÜV SÜD certified the core NVIDIA hardware and software process to Automotive Safety Integrity Level (ASIL) D.  Under the ISO 26262 standard, NVIDIA [DriveOS](https://developer.nvidia.com/drive/os) 6.0 is certified  ASIL D conformant and Thor-X SoC is assessed as ASIL D conformant. NVIDIA also received ISO/SAE 21434 Cybersecurity Process certification for its automotive system-on-a-chip, platform, and software engineering processes. ### TÜV Rheinland TÜV Rheinland performed an independent United Nations Economic Commission for Europe safety assessment of NVIDIA DRIVE AV related to safety requirements for complex electronic systems. Research ## NVIDIA Research for Autonomous Vehicles Our research and development have published 330+ research papers on autonomous vehicle safety. ### Alpamayo 1: Bridging Reasoning and Action Prediction for Generalizable Autonomous Driving in the Long Tail 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). [Learn More About Alpamayo](https://research.nvidia.com/publication/2025-10_alpamayo-r1) ### Cosmos-Drive-Dreams: Scalable Synthetic Driving Data Generation With World Foundation Models 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. [Learn More About Cosmos](https://arxiv.org/abs/2504.15362) ### Sim2Val: Using Correlation Across Test Platforms for Variance-Reduced Metric Estimation Learning-based robotic systems demands rigorous validation to assure reliable performance. However, extensive real-world testing is often prohibitively expensive and, if conducted, may still yield insufficient data for high-confidence guarantees. That’s where Sim2Val comes in. [Visit GitHub](https://github.com/NVlabs/sim2val) ### SafeVL: Driving Safety Evaluation Using Meticulous Reasoning in Vision Language Models 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. [Visit GitHub](https://safevl.github.io/) ### Enhancing Autonomous Driving Safety With Collision Scenario Integration HydraSafe is a framework that addresses the challenge of ensuring autonomous vehicle safety in hazardous scenarios by improving data availability and planner robustness. [Visit the Project Page](https://nvlabs.github.io/SafeHydra/) ### Safety Evaluation of Motion Plans Using Trajectory Predictors as Forward Reachable Set Estimators 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. [Read the Paper](https://arxiv.org/abs/2507.22389) For a list of additional AV Research papers, [click here](https://research.nvidia.com/publications?f%5B0%5D=research_area%3A3445). ## NVIDIA Alpamayo Designed to handle real-world scenarios, Alpamayo accelerates safe Level 4 AV development by enabling reasoning-based autonomy through a family of open VLA models, physical AI datasets, and simulation frameworks. [Learn More About Alpamayo](https://nvidianews.nvidia.com/news/alpamayo-autonomous-vehicle-development) [Watch the Demo (3:44)](https://www.youtube.com/watch?v=5EEYgxoT8kc) Partners ## Partners Using NVIDIA Halos System Leading robotaxi companies, OEMs, industry safety pioneers, mapping and simulation companies, and software and sensor providers worldwide are using the system to deliver autonomous vehicle safety at all levels of automation. [View All NVIDIA DRIVE Partner Ecosystem Members](https://www.nvidia.com/en-us/self-driving-cars/partners.md) ### Cars ### Trucks ### Robotaxis ### Suppliers ### Simulation ### Sensors ### Software ### Mapping Resources ## The Latest in NVIDIA Halos Resources 1. ### News 2. ### Sessions 3. ### Live Streams [View All AV Safety Sessions](https://www.nvidia.com/en-us/on-demand/playlist/playList-187e39a0-9b5f-4655-bc6a-af749f397332/) ### Scaling Functionally Safe AVs With NVIDIA DriveOS and Hyperion How do you scale autonomous vehicles to large fleets while meeting the world’s strictest safety standards? This livestream presented by NVIDIA safety experts shows how the [DriveOS](https://www.nvidia.com/en-us/solutions/autonomous-vehicles/drive-av.md) stack and [Hyperion](https://www.nvidia.com/en-us/solutions/autonomous-vehicles/drive-hyperion.md) reference architecture solve the safety-scalability problem. [Watch the Replay](https://www.youtube.com/watch?v=5SgAUKKL2fo) ### Open Dataset for Autonomous Vehicle Safety Research In this livestream, we showcase how the newly launched [NVIDIA Alpamayo](https://www.nvidia.com/en-us/solutions/autonomous-vehicles/alpamayo.md) autonomous vehicles open dataset creates new opportunities to advance AV safety. [Watch the Replay](https://www.youtube.com/live/-5jfeuv2Lkw) ### Accelerating Physical AI Certification With NVIDIA Halos Learn about NVIDIA Halos AI Systems Inspection Lab—the first ANSI National Accreditation Board (ANAB)-accredited lab dedicated to physical AI systems. [Watch the Replay](https://www.youtube.com/watch?v=gaWaUa27ufQ) ### Autonomous Vehicle Safety Validation Through Simulation In this livestream, we present a general, theoretically grounded framework for AV safety validation whereby real-world tests are paired with simulated tests on corresponding reconstructed scenarios. [Watch the Replay](https://www.addevent.com/event/Xg26329868) ## Next Steps ### Redefining Safe Autonomy Learn how cutting-edge AI, rigorous validation frameworks, and global standards are shaping autonomous vehicle safety. [Read the Product Brief](https://www.nvidia.com/content/dam/en-zz/Solutions/self-driving-cars/nvidia-halos-product-brief.pdf) #### Automotive News Sign up for the latest news and updates from NVIDIA. [Stay Informed](#automotive-news) #### NVIDIA Automotive NVIDIA solutions give you the performance and scalability to design, visualize, develop, and simulate the future of driving. [Explore NVIDIA Automotive](https://www.nvidia.com/en-us/industries/automotive.md) ## Read the NVIDIA Autonomous Vehicles Safety Report Welcome back. Not you? Log Out Welcome back. Not you? Clear form