opens up new avenues for advancement at the intersection of autonomy and robotics, accelerating, in particular, the development of autonomous vehicles ( AVs ). The proper tools and systems are essential to ensuring the health of both drivers and pedestrians.
NVIDIA Halos, a thorough security system that combines NVIDIA’s line of mechanical hardware and software safety solutions with its cutting-edge AI study in AV safety, is announced today by NVIDIA.
With a focus on AI-based, end-to-end AV stacks, Halos spans cards and technology to tools and services to help maintain healthy growth of Vcs from the fog to the car.
With the release of Halos, we’re empowering partners and developers to select the most advanced engineering components needed to create their own individual offerings, furthering our shared goal of creating secure and trustworthy autonomous vehicles, according to Riccardo Mariani, NVIDIA’s vice president of business safety. Halos enhances already-established security techniques and may lead to more stringent standards and regulation compliance.
At the center of Halos
Halos is a comprehensive security system with three distinct but comparable levels.
It covers system, analytic, and ecosystem safety at the technological level. It includes design-time, deployment-time, and validation-time scaffolding at the growth stage. And at the computing level, it runs from for AI education to for NVIDIA Omniverse and NVIDIA Cosmos running on for modeling, and for implementation, using three powerful computers.
” Halos ‘ holistic approach to safety is especially important in a situation where businesses want to use the power of generative AI to develop increasingly capable AV systems that don’t require traditional compositional design and verification,” said Marco Pavone, lead AV researcher at NVIDIA.
Inspection Lab for AI Systems
Serving as an entry point to Halos is the , which allows automakers and developers to verify the safe integration of their products with NVIDIA technology.
The Inspection Lab for AI Systems, announced at the CES trade show earlier this year, is the first worldwide program to be accredited by the for an inspection plan integrating functional safety, cybersecurity, AI safety and regulations into a unified safety framework.
Inaugural members of the Inspection Lab for AI Systems include Ficosa, OMNIVISION, onsemi and Continental.
“Being a member of the Inspection Lab for AI Systems means working at the forefront of automotive systems innovation and integrity,” said Cristian Casorran Hontiyuelo, advanced driver-assistance system engineering and product manager at Ficosa.
According to Paul Wu, brain of product marketing for mechanical at OMNIVISION,” Vehicles are so much more than just transportation.” They’ve even become our entertainment and information centers. Our safety may be constantly improved by cars. As a testament to our commitment to providing the highest standards of safety in our item services, we are pleased to meet NVIDIA’s new AI Systems Safety Lab.
“We are delighted to be working with NVIDIA and included in the launch of the NVIDIA Inspection Lab for AI Systems,” said Geoff Ballew, general manager of the automotive sensing division at onsemi. “This unique initiative will improve road safety in an innovative way. We look forward to the advancements it will bring.”
“We are pleased to participate in the newly launched NVIDIA Drive Inspection Lab for AI Systems and to further intensify the fruitful, ongoing collaboration between our two companies,” said Nobert Hammerschmidt, head of components business at Continental.
The essential components of a orbs
Halos is built on three main areas of target: ecology safety, algorithm safety, and system safety.
Software Safety
Halos has hundreds of built-in safety mechanisms and a system-on-chip ( SoC ) that has been safety-assessed.
Additionally included are Advance AGX Hyperion, a technology platform that connects SoC, DriveOS, and sensors in an electronic control unit architecture, as well as software, a safety-certified operating system that extends from CPU to GPU, a safety-assessed base platform that delivers the basic computer needed for healthy systems for all types of applications.
Algorithmic Safety
Halos includes libraries for safety data loading and accelerators, as well as application programming interfaces for safety data creation, curation, and reconstruction to filter out, for instance, undesirable behaviors and biases before training.
Additionally, it has rich training, simulation, and validation environments that make use of the NVIDIA Omniverse Blueprint for AV simulation and world foundation models to train, test, and validate AVs. Additionally, it has a diverse AV stack that combines modular components with end-to-end AI models to ensure safety with cutting-edge AI models in the loop.
Ecosystem safety
Halos includes safety datasets with varied, objective data, as well as safe deployment workflows, which include automated safety evaluations and triaging workflows, as well as a data flywheel for ongoing safety improvements, demonstrating leadership in AV safety standardization and regulation.
Safety Track Record
Halos brings together a sizable number of NVIDIA-focused technology research, development, deployment, partnerships, and collaborations, including:
- Engineering years and more than 15 000 spent on vehicle safety
- Contributions to international standards committees totaling 10,000+ hours
- 1, 000+ AV-safety patents filed
- 240+ AV-safety research papers have been published.
- more than 30 certificates in cybersecurity and safety
Additionally, it aligns with recent significant safety certifications and assessments of NVIDIA automotive products, such as:
- The NVIDIA DriveOS 6.0 operating system complies with ASIL D ( automotive safety integrity ) standards.
- NVIDIA’s automotive platform, software engineering processes, and ISO/SAE 21434 certification were granted by TÜV SÜD.
- For the United Nations Economic Commission for Europe, TÜV Rheinland conducted an independent safety assessment of NVIDIA DRIVE AV in relation to safety requirements for sophisticated electronic systems.
Attend  , a global AI conference taking place through Friday, March 21, to learn more about NVIDIA’s approach to automotive safety.
See the for information on software products.