NVIDIA DRIVE AV Software Debuts in All-New Mercedes-Benz CLA
Source: NVIDIA AI Blog
What’s happening?
NVIDIA is bringing its DRIVE AV software—now with enhanced Level 2 point‑to‑point driver‑assistance—to U.S. streets by the end of 2024. The rollout starts with Mercedes‑Benz, a long‑standing partner in safe, intelligent mobility.
The All‑New Mercedes‑Benz CLA
- First MB vehicle on the MB.OS platform
- Powered by NVIDIA’s full‑stack DRIVE AV software, AI infrastructure, and accelerated compute
- Supports over‑the‑air (OTA) updates for future upgrades and new features
- Planned enhancements to MB.DRIVE ASSIST PRO will be available ex‑factory and via the Mercedes‑Benz Store
Safety Achievement
- The CLA earned a five‑star rating from the European New Car Assessment Programme (EuroNCAP)【EuroNCAP】.
- The top score reflects the performance of MB.DRIVE active‑safety features in accident mitigation and avoidance.
Executive Quote
“As the automotive industry embraces physical AI, NVIDIA is the intelligence backbone that makes every vehicle programmable, updatable and perpetually improving through data and software. Starting with Mercedes‑Benz and its incredible new CLA, we’re celebrating a stunning achievement in safety, design, engineering and AI‑powered driving that will turn every car into a living, learning machine.”
— Ali Kani, Vice President of Automotive, NVIDIA
Why It Matters
- Programmable & Updatable: Vehicles become platforms that evolve long after purchase.
- AI‑Powered Safety: Continuous improvements in perception and decision‑making raise safety standards.
- Industry Shift: Demonstrates how AI is moving from a feature add‑on to the core architecture of modern cars.
NVIDIA DRIVE AV — A Dual‑Stack Architecture for Intelligent Safety & Performance
NVIDIA DRIVE AV combines two complementary stacks:
| Stack | Purpose | Key Features |
|---|---|---|
| AI End‑to‑End Stack | Core driving intelligence | • Learns from massive real‑world & synthetic datasets • Human‑like decision‑making in complex scenarios |
| Classical Safety Stack (NVIDIA Halos) | Redundancy & safety guardrails | • Built on the NVIDIA Halos safety system • Guarantees operation within defined safety parameters • Provides fail‑safe checks and redundancy |
What This Means for Consumers
- Confidence & Comfort: Built‑in redundancy and fail‑safe mechanisms protect passengers.
- Secure Journeys: Halos continuously monitors vehicle behavior to keep it inside safe operating limits.
Capabilities Enabled by the Unified Architecture
- Advanced Level‑2 Automation – Point‑to‑point urban navigation through dense city streets.
- Proactive Active Safety – Collision avoidance that anticipates hazards before they occur.
- Automated Parking – Precise maneuvering in tight spaces.
- Co‑operative Steering – Seamless hand‑off between driver and system for a smoother driving experience.
Humanlike Urban Driving with End-to-End AI Models
NVIDIA deep‑learning models power a new generation of AI‑assisted urban‑driving systems. These models interpret traffic holistically, enabling vehicles to:
- Navigate intelligently through lane selection, turns, and route‑following in congested or unfamiliar areas.
- Recognize vulnerable road users (pedestrians, cyclists, scooter riders) and respond proactively—by yielding, nudging, or stopping—to prevent collisions.
- Assist drivers throughout trips, safely navigating from any address to any other (e.g., from home to work).
Speeding Manufacturing with Physical AI and Digital Factories
NVIDIA and Mercedes‑Benz are reshaping vehicle production through a digital‑first approach built on the NVIDIA Omniverse ecosystem.
- Digital twins of factories and assembly lines let engineers design, plan, and optimize operations in a virtual environment.
- This reduces downtime, shortens iteration cycles, and accelerates time‑to‑market.
- The combination of Omniverse and the NVIDIA Cosmos platform enables developers to test and validate intelligent‑driving software in realistic simulations before any real‑world deployment.
Cloud‑to‑Car Development: NVIDIA’s Three‑Computer Architecture
All NVIDIA‑powered intelligent‑driving systems share a cloud‑to‑car development pipeline that turns real‑world data into billions of simulated miles, accelerating improvement.
1. Training Infrastructure
- NVIDIA DGX systems provide massive‑scale GPU computing to train DRIVE AV foundation models on diverse, global datasets.
- These models capture human driving behavior across millions of real‑world scenarios.
Learn more about NVIDIA DGX for autonomous‑vehicle AI training →
2. Simulation & Validation
- NVIDIA Omniverse and Cosmos create physically accurate simulation environments for testing and scenario generation.
- Developers can validate new features in thousands of edge cases before deployment, converting real‑world miles into billions of virtual ones.
Explore NVIDIA Omniverse & Cosmos for autonomous‑vehicle simulation →
3. In‑Vehicle Compute & Hyperion Architecture
- NVIDIA DRIVE AGX delivers accelerated compute for perception, sensor fusion, and decision‑making in real time, handling complex urban and highway scenarios simultaneously.
- DRIVE Hyperion provides the compute and sensor architecture that adds redundancy and diversity for a safe, highly advanced automated‑driving experience.
Discover NVIDIA DRIVE AGX in‑vehicle computing →
Why This Closed‑Loop Approach Matters
- Rapid iteration on driving algorithms.
- Exceptional accuracy through massive‑scale training.
- Safety validation in rare or hazardous edge cases.
- Scalable deployment across multiple vehicle platforms.
The Era of AI‑Defined Transportation Has Arrived
In addition to Mercedes‑Benz, NVIDIA collaborates with a wide range of global automakers to deliver its full‑stack software, AI infrastructure, and advanced computing technologies across diverse vehicle lineups. This approach enables car and truck makers to integrate intelligent driving capabilities and safety innovations at scale, while simplifying development, deployment, and future upgrades.