Ouster Unveils Stereolabs ZED X Nano Wrist-Mounted Camera

Ouster provides digital lidar, cameras, AI computing, sensor fusion and perception software, and AI models. | Source: Ouster

Ouster, Inc. announced the release of the Stereolabs ZED X Nano yesterday, a compact wrist-mounted stereo camera designed for robotic manipulation, imitation learning, and high-volume data collection.

“Building on Stereolabs' leadership in AI vision and perception solutions, the ZED X Nano allows us to penetrate deeper into the industrial and robotics markets, securing new applications that require more compact form factors,” said Ouster CEO Angus Pacala. “The future of Physical AI relies on vast amounts of high-quality, low-latency image data gathered at the edge. With the ZED X Nano, we are providing roboticists with a significant upgrade to their vision systems, enabling machines to perceive, reason, act, and learn with exceptional precision.”

As robotics teams scale up imitation learning and reinforcement learning for manipulation tasks, RGB image quality and end-to-end capture latency have become major bottlenecks. Conventional cameras often depend on USB connectivity, capture lower-resolution 720p RGB and depth data, and require CPU-mediated pipelines that limit throughput and increase latency.

Ouster engineered the Stereolabs ZED X Nano to address these challenges. It is 40% shorter in height than comparable solutions and mounts directly onto robotic wrists and end-of-arm tooling. It utilizes the same reliable 1920×1200 global shutter sensor found across the flagship ZED X camera line. The camera can capture high-resolution RGB and depth images at up to 120 frames per second for training data and manipulation tasks.

The ZED X Nano is available for pre-order starting today, with shipments commencing in May 2026.

ZED X Nano is engineered for physical AI, delivering sub-millimeter accuracy

The core of the ZED X Nano features an ultra-low-latency capture pipeline with a fully zero-copy data path from sensor to GPU. This allows frames to stream directly into NVIDIA hardware encoders and AI inference pipelines simultaneously.

For data collection teams, this enables higher-throughput dataset capture at full resolution. For deployment teams, it allows perception, segmentation, and policy networks to run in parallel on the same video feeds, freeing up more GPU resources.

Depth sensing is powered by Stereolabs' Neural Depth Engine, an AI-driven stereo depth system that achieves sub-millimeter accuracy along the Z-axis. Ouster states this depth engine provides significantly better lateral (XY) positioning than traditional structured-light or time-of-flight cameras. It offers critical advantages for grasp pose estimation, precise placement, and assembly tasks where lateral error can lead to manipulation failure. With a minimum depth-sensing range of 3 centimeters, the camera can also detect objects closer than many competing solutions on the market.

Ouster designed the ZED X Nano for durability. It features an onboard vibration-resistant IMU and is powered by a ruggedized GMSL2 connection with next-generation cabling built to withstand the repeated motion and cable stress of robotic arms. This replaces a fragile USB-C port with an industrial-grade link that can transmit video up to 15 meters while offering EMI resistance and secure locking connectors.

The ZED X Nano is the latest Stereolabs product to offer seamless native integration with NVIDIA Isaac Sim and Isaac Lab for sim-to-real transfer, alongside full support for ROS and ROS 2. Teams developing imitation learning or reinforcement learning pipelines can capture high-fidelity demonstrations, train in simulation using accurately matched camera models, and deploy to physical hardware—all utilizing the same sensor and software stack.