Mercedes-Benz is marking 140 years of automotive innovation with a brand new S-Class constructed for the AI period, bringing collectively automotive security and NVIDIA’s superior autonomous driving platform to allow a degree 4-ready structure designed for belief.
The brand new S-Class with MB.OS, which shall be geared up with the NVIDIA DRIVE Hyperion structure and full-stack NVIDIA DRIVE AV L4 software program, is designed to help future robotaxi operations — delivering safety-first autonomy with the NVIDIA Halos system and end-to-end AI and classical driving stacks operating in parallel to make sure dependable operation.
The S-Class allows a premium, chauffeur-style autonomous expertise. As a part of NVIDIA’s beforehand introduced partnership with Uber, the businesses will work collectively to make these autonomous automobiles obtainable to riders by way of Uber’s mobility community.
It showcases how legacy automakers and AI pioneers can work collectively to construct automobiles which might be safer, smarter and more and more autonomous — with out compromising the excessive requirements of high quality and security prospects count on.
“Mercedes-Benz has set the usual within the automotive market, constructing automobiles outlined by beautiful craftsmanship and security engineering,” stated Jensen Huang, founder and CEO of NVIDIA, within the above video celebrating the S-Class launch. “5 years in the past, NVIDIA started working with Mercedes-Benz to assist carry that legacy into the AI period.”
L4-Prepared Structure Powered by NVIDIA DRIVE AV
Conventional autonomous driving approaches typically depend on predefined guidelines or discovered responses to acquainted conditions. However real-world driving is crammed with uncommon and complicated edge instances — from unpredictable pedestrian conduct and particles to uncommon street situations and aggressive cut-ins.
NVIDIA DRIVE AV gives Mercedes-Benz’s new S-Class with a full-stack automated driving system designed to deal with this lengthy tail of driving situations, whereas remaining anchored to a safety-first structure.
NVIDIA DRIVE AV is skilled at scale on NVIDIA DGX methods and designed to be validated  utilizing high-fidelity simulation with NVIDIA Omniverse NuRec libraries and NVIDIA Cosmos world fashions.
Constructed on NVIDIA’s broader AI basis — together with superior notion, planning and reasoning applied sciences — NVIDIA DRIVE AV is optimized, validated and distilled to run reliably in manufacturing automobiles, tailor-made to Mercedes-Benz’s car platforms and sensor configurations.
NVIDIA DRIVE AV allows the system to investigate complicated environments — somewhat than merely reacting to recognized patterns — consider a number of choices and choose the most secure potential final result in actual time.
Variety by Design With NVIDIA DRIVE Hyperion for Actual-World Mobility
For degree 4 autonomy, security is dependent upon greater than easy redundancy. Autos should stay operational within the face of {hardware} faults, sensor degradation and surprising software program conduct.
The brand new S-Class shall be constructed on NVIDIA DRIVE Hyperion, a reference structure that integrates sensor range and {hardware} redundancy right into a unified platform, to function a robotaxi.
DRIVE Hyperion is designed based mostly on defense-in-depth rules:
- Redundant compute to assist preserve operation if one processing ingredient fails.
- Multimodal sensor range — spanning cameras, radar and lidar — to help strong notion.
- Software program stack range, pairing AI-driven decision-making with a parallel classical security stack to maintain the car working inside secure boundaries.
Developed in accordance with NVIDIA Halos security system, NVIDIA DRIVE Hyperion helps get rid of single factors of failure and gives the inspiration wanted for L4-ready methods.
This safety-first, resilient platform is principally designed for premium robotaxi and chauffeured mobility providers — enabling dependable, large-scale deployment in real-world environments.
From AI Foundations to Manufacturing-Prepared AutonomyÂ
NVIDIA’s broader AI ecosystem — together with the NVIDIA Alpamayo household of open fashions, simulation instruments and datasets for autonomous automobiles — allows builders and companions to advance autonomous driving analysis and construct their very own driving software program.
Inside NVIDIA DRIVE AV, these AI capabilities are additional refined, optimized and engineered for manufacturing. It ensures dependable operation on automotive-grade {hardware}, with NVIDIA Halos making use of strict security requirements to the AI pipeline, in addition to seamless integration with Mercedes-Benz’s particular sensor and car architectures.
This production-grade strategy — combining large-scale coaching, high-fidelity simulation, rigorous security validation and deep system integration — is what permits NVIDIA DRIVE AV to help each degree 2 point-to-point and degree 4-ready automated driving methods in buyer automobiles.
Constructing on this basis, Mercedes-Benz and NVIDIA are partnering to ship an L4-ready model of the brand new S-Class, bringing superior AI and safety-focused autonomy to the street.
On the core of this work is NVIDIA Alpamayo, which allows automobiles to drive easily and naturally like a human driver whereas reasoning step-by-step by way of complicated conditions to decide on the most secure potential motion — since security is paramount.
Bringing Security Engineering Into the Autonomous Driving PeriodÂ
As AI turns into central to car intelligence, the definition of “the most secure automobile” is evolving. Past defending occupants in a crash, fashionable automobiles are more and more designed to assist stop accidents within the first place.
Constructed on NVIDIA DRIVE Hyperion and full-stack NVIDIA DRIVE AV software program, the next-generation S-Class extends Mercedes-Benz’s long-standing security management into the AI period. Its L4-ready structure combines end-to-end AI with parallel classical driving stacks, delivering predictable, dependable operation by way of a various, multi-layered system design.
This strategy displays a broader shift towards energetic, clever security — a pattern already acknowledged by unbiased testing, together with the Mercedes-Benz CLA’s designation as Euro NCAP’s Finest Performer of 2025.
Collectively, Mercedes-Benz and NVIDIA are demonstrating how legacy automakers and AI pioneers can collaborate to ship automobiles which might be safer, smarter and more and more autonomous — with out compromising the craftsmanship, consolation and high quality prospects count on.
