Nuro’s autonomous driving platform will power a new robotaxi built on the Lucid Gravity SUV, with the three-way partnership between Nuro, Lucid Group, and Uber reaching a critical milestone as supervised on-road testing began in December 2025 in the San Francisco Bay Area.
Highlights
- Nuro’s end-to-end AI foundation model combines machine learning with verifiable safety logic for Level 4 autonomous operation
- The sensor array delivers 360-degree perception using solid-state lidar, high-resolution cameras, and radar integrated throughout the vehicle body and roof-mounted halo module
- NVIDIA DRIVE AGX Thor provides compute power for real-time AI processing and system integration
- Validation combines on-road, closed-course, and simulation testing to evaluate the full autonomy stack before commercial deployment
Autonomy Architecture
Nuro’s Level 4 autonomous system, branded the Nuro Driver, serves as the robotaxi’s core intelligence. The company describes the underlying technology as an end-to-end AI foundation model that blends machine learning capabilities with what Nuro calls “clear, verifiable safety logic.”
This hybrid approach aims to deliver both the adaptive capabilities of neural network-based systems and the predictability required for commercial autonomous operations.
“By bringing together Nuro’s proven level 4 autonomy, Lucid’s advanced vehicle architecture, and Uber’s global reach, we’re building a robotaxi service designed for real-world operations and long-term growth,” said Dave Ferguson, Co-Founder and Co-CEO of Nuro.
Sensor Suite and Perception System
The robotaxi achieves 360-degree environmental awareness through a multi-modal sensor array distributed across two locations: integrated throughout the Lucid Gravity body and within a purpose-built roof-mounted module called the halo.
The perception hardware includes:
- High-resolution cameras for visual scene understanding
- Solid-state lidar sensors for precise ranging and 3D mapping
- Radar systems for velocity detection and all-weather redundancy
The halo module maintains a low profile to maximize sensor visibility while preserving the vehicle’s exterior design. Integrated LEDs in the halo serve operational functions including vehicle identification, rider initial display, and status communication during pickup and dropoff sequences.
Compute Platform
Real-time AI processing runs on NVIDIA DRIVE AGX Thor, part of the NVIDIA DRIVE Hyperion platform. This compute architecture handles the processing demands of sensor fusion, perception, planning, and vehicle control required for Level 4 autonomous driving.
The Thor platform supports the system integration necessary to coordinate inputs from the distributed sensor array with Nuro’s autonomy software stack.
Validation and Testing Framework
Nuro leads the testing program using engineering prototypes supervised by autonomous vehicle operators. The validation framework draws on Nuro’s experience from previous commercial autonomous deployments, including its driverless delivery operations.
The program evaluates capabilities across the full autonomy stack using three complementary methodologies:
- On-road testing for real-world performance validation
- Closed-course testing for controlled scenario evaluation
- Simulation for edge case coverage and scalable testing
Testing addresses dozens of critical capabilities including pedestrian yielding, traffic light response, lane change execution, and passenger pickup and dropoff procedures.
In-Vehicle Visualization
The robotaxi provides passengers with real-time visualization of the autonomous system’s perception and planning. Displays show what the vehicle sees and its intended trajectory, including upcoming maneuvers.
This transparency feature allows riders to observe how the system interprets its environment and makes driving decisions throughout the journey.
Path to Commercial Deployment
Pending completion of the validation program, production is scheduled to begin at Lucid’s Arizona manufacturing facility later in 2026. The robotaxi service will launch initially in the San Francisco Bay Area through Uber’s ridehail platform.
The production intent vehicle is on display at NVIDIA’s CES showcase at the Fontainebleau Hotel through January 8, 2026.
Sign up for our weekly emails to stay up to date on the latest news!
