A ready-to-integrate vision pack for Leo Rover, ROS 2, and 3D perception
- Complete kit for Leo Rover: adds a vision and AI computing layer without designing a custom mount or sourcing components separately.
- Jetson Orin Nano + ZED 2i: a combination designed for 3D vision, edge processing, and mobile robotics applications.
- Ready for ROS 2: preconfigured Docker image to speed up deployment and avoid manual driver compilation.
- Camera data available directly: RGB, depth, point cloud, IMU, and TF frames accessible in ROS 2.
- Mechanical integration included: mounts, 3D-printed enclosure, and screws included in the pack.
- Powered from the rover battery: the Jetson Orin Nano is powered directly by Leo Rover via the supplied cable.
- Simplified network integration: transparent Ethernet bridge to make ROS 2 topics visible on the rover network.
- URDF-ready: the ZED 2i integrates directly into Leo Rover’s TF tree.
The Vision Pack is designed for integrators, laboratories, engineering schools, and R&D teams looking for a faster-to-deploy solution than a fully in-house vision integration. The goal is to move more quickly to experimentation by focusing on algorithms, perception, and use cases rather than hardware assembly.
Why choose the Vision Pack for Leo Rover?
This pack adds stereo vision and edge computing capabilities to Leo Rover in a fast-deployment setup. It provides a solid base for projects involving visual navigation, 3D mapping, scene understanding, or edge AI inference on a mobile platform.
- Visual SLAM and 3D mapping: for environment reconstruction and localization.
- Obstacle detection and navigation: embedded perception to improve rover autonomy.
- Scene understanding with AI models: use of RGB and depth data for advanced applications.
- Real-time pipelines on a mobile platform: onboard processing directly on the Jetson.
Package contents
- 1 × NVIDIA Jetson Orin Nano
- 1 × Stereolabs ZED 2i stereo camera (non-polarized lenses)
- 1 × USB to Ethernet adapter
- 1 × modified MEB power cable with barrel jack connector
- 1 × Ethernet patch cable
- 1 × USB-C cable
- 1 × 3D-printed Vision Pack enclosure
Please note: three Allen keys (2.5 mm, 3 mm, and 4 mm) are required for mechanical installation and are not included.
Technical specifications of the Vision Pack for Leo Rover
| AI computing |
NVIDIA Jetson Orin Nano |
| GPU |
Ampere — 1024 CUDA cores, 32 Tensor Cores |
| AI performance |
Up to 40 TOPS |
| Jetson power consumption |
7 to 15 W, configurable |
| Required software version |
JetPack 6.2.1 (L4T ≥ 36.4.0) |
| Stereo camera |
Stereolabs ZED 2i |
| RGB resolution |
Up to 2K |
| Depth range |
Up to 20 m |
| Horizontal field of view |
110° |
| Integrated sensors |
IMU, barometer, magnetometer |
| Camera interface |
USB 3.0 (camera to Jetson) |
| Data interface to the rover |
USB-C → USB to Ethernet → Ethernet bridge on the rover |
| Power interface |
Barrel jack from the Leo Rover battery |
Resources for the Vision Pack for Leo Rover