reComputer Mini Carrier Board for Jetson Orin Nano/Orin NX

• Compact and Lightweight for Deployment: Weighs only 80 grams, making it ideal for UAVs and UGVs, with a compact form factor and rich I/O ready for field integration.

• Powerful AI Performance for Real-Time Drone Autonomy: Delivers up to 100 TOPS with low power and latency, compatible with the NVIDIA Orin Nano/Orin NX SoC , enables real-time navigation, obstacle avoidance, object detection, and tracking for autonomous systems.

• Versatile Sensor Support: Compatible with RGB, infrared, hyperspectral, LiDAR, and 3D cameras, enabling rich environmental perception for advanced autonomy.

• Seamless Communication Interfaces: Integrates easily with flight controllers and peripherals via UART, and expandable Ethernet and CAN ports for smooth system coordination.

• Flexible Power Input: Supports a wide voltage range (4S to 14S battery packs) through the XT30 DC-IN connector, ideal for mobile and aerial platforms.

• Real-Time Wireless Connectivity: Connecting your 5G/4G module on it to support for real-time image, video, and data transmission, enhancing situational awareness and remote operation.

• Robotic Control

• Multi-Sensor Fusion for Autonomous Mapping & Navigation

reComputer Mini supports sensor-rich autonomy by integrating data from LiDAR (via Ethernet/CAN), 3D cameras (via USB/CSI), GPS, and IMU (via UART). Using AI models and algorithms like ORB-SLAM,A-LOAM, or Nav2, it performs sensor fusion, real-time localization, path planning, and obstacle avoidance. The low-latency processing on the Orin NX SoC allows autonomous systems to adapt to changing terrain or obstacles on the fly. Its wide voltage input and compact form factor make it suitable for mobile deployment, enabling robust, edge-level autonomy without external compute.

Learn more about sensor options compatible with reComputer Mini to get started.

• Build Multi-streams Object Detection & Tracking

For building real-time object detection, localization, and multi-object tracking, Jetson Orin NX incorporates the NVIDIA Multi-Standard Video Decoder, running YOLOV8/YOLO11 AI models to process high-resolution video input from RGB or infrared cameras connected via MIPI CSI or USB. This enables deployment in UAVs/UGVs for dynamic environments where identifying and following object, and assets is critical—without relying on cloud computation. See our tested YOLOv8 performance using NVIDIA Deepstream of single model and multi model on multi streams.