Hiwonder ROSOrin ROS2 Smart Car, Multimodal Large AI Model ChatGPT / Gemini / Grok / Llama, SLAM Navigation Programming, Mecanum Wheels Robot
【ROS2 & High-Performance Hardware】 ROSOrin runs on ROS2, compatible with Jetson and Raspberry Pi. It features LiDAR, a 3D camera, and a six-microphone array for tasks like mapping and navigation, 3D recognition, and voice interaction.
【Multimodal Large AI Model & OpenClaw】 Multimodal large AI model deployment and deep OpenClaw agent integration unlock advanced embodied intelligence. This enables cross-platform collaboration, proactive responsiveness, and spatial vector memory for advanced robotic applications.
【Versatile Chassis Expansion】 Switches between Mecanum, Ackermann, and differential drive chassis, paired with a swing arm suspension system for stable movement and precise control on complex terrains.
【AI Vision Algorithms】 Equipped with YOLO26, OpenCV, MediaPipe, and KCF algorithms, ROSOrin delivers powerful visual perception for advanced applications like autonomous driving, face recognition, and target tracking.
【Learning Resources】 Includes comprehensive tutorials, source code, and AI project examples from beginner to advanced, everything needed to move from first build to advanced ROS robot applications.
Check your ROSOrin Tutorials here: https://wiki.hiwonder.com/projects/ROSOrin/en/jetson-orin-nano-version/.
Product Description
Hiwonder ROSOrin is a multimodal AI robot developed on ROS2. It features a modular chassis design that allows fast switching between Mecanum, Ackermann, and differential drive chassis, enabling flexible adaptation to diverse scenarios.
ROSOrin is equipped with high-performance hardware, including NVIDIA Jetson, Raspberry Pi 5, LiDAR, a 3D depth camera, and an integrated 6-microphone array. It supports motion control, mapping and navigation, path planning, 3D perception, tracking and obstacle avoidance, object recognition, target tracking, gesture interaction, and voice interaction.
Supporting hybrid online and local deployment, ROSOrin integrates multimodal large AI models with the OpenClaw agent framework for natural voice interaction, scene understanding, and cross-platform collaboration.
OpenClaw Framework Deployment
Deep OpenClaw integration utilizes versatile controls and multimodal inputs to precisely perceive complex environments, make real-time decisions, and autonomously plan and execute intricate workflows.
Multimodal AI Integration
ROSOrin combines advanced AI for text, voice, and vision, delivering powerful cognition for natural conversations and smart visual understanding.
AI Vision & Tracking
With built-in OpenCV vision algorithms, ROSOrin can quickly identify colors and objects, calculate their positions in real time, and track them accurately.
Built-in AI Vision Algorithms
With YOLO26, MediaPipe, and OpenCV, ROSOrin handles object detection, face recognition, and gesture control with no complex setup required.
3D Depth Camera
The 3D depth camera can not only realize AI visual game-play, but also enable advanced gameplay such as depth image data processing and 3D visual mapping navigation.
STL-19P D500 TOF Lidar
ROSOrin is equipped with Lidar, which can realize SLAM mapping and navigation, and supports path planning, fixed-point navigation and dynamic obstacle avoidance.
ROS Control System
ROS Control allows you to run multiple neural networks, object detection, segmentation and speech processing applications in parallel.
6-Ch Far-field Microphone Array
The 6-Ch far-field microphone array and speakers support sound source positioning, voice recognition control, voice navigation and other functions.
Large AI Model–Driven Embodied AI Applications
ROSOrin Advanced Kit are equipped with a circular 6-microphone array. Going beyond the one-way command-response pattern of traditional AI models, ROSOrin—powered by ChatGPT—enables a cognitive leap from semantic understanding to physical execution, significantly enhancing the naturalness and fluidity of human-machine interaction. Combined with advanced machine vision, ROSOrin delivers outstanding capabilities in perception, reasoning, and action, making it ideal for developing sophisticated embodied AI applications.
Vision Tracking
With the advanced perception capabilities of a vision language model, ROSOrin can intelligently identify and lock onto target objects even in complex environments, allowing it to perform real-time tracking with adaptability and precision.
Voice Control
With ChatGPT integration, ROSOrin can comprehend spoken commands and carry out corresponding actions, enabling intuitive and seamless voice-controlled interaction.
Autonomous Patrolling
Utilizing semantic understanding from a large language model, ROSOrin can accurately detect and track lines of various colors in real time while autonomously navigating obstacles, ensuring smooth and efficient patrolling.
Distance Awareness
Hiwonder ROSOrin car combines a visual AI model with a depth camera to understand its environment and perceive distances. By combining visual recognition with distance data, it enables intelligent question answering.
Integration of Large AI Model with SLAM-Based Mapping and Navigation
ROSOrin combines multimodal large model to understand user voice commands via a large language model, enabling multi-point navigation. Once it arrives at the designated location, it uses a vision language model to gain a deep understanding of the surrounding objects and events. This approach greatly enhances the robot’s intelligence, adaptability, and overall user experience, making it better suited to meet real-world needs.
Intelligent Navigation
ROSOrin continuously sends environmental data to the vision language model for real-time in-depth analysis. It dynamically adjusts its navigation path based on user voice commands, allowing it to autonomously navigate to designated areas and deliver intelligent, adaptive routing./p>
Emotion Perception
Leveraging an extended RAG knowledge base, ROSOrin can recognize intent and analyze environmental context, anticipating potential needs without detailed instructions, autonomously planning tasks and responding dynamically.
Large AI Models × OpenClaw
Integrated SLAM Mapping & Navigation
Large AI Models × OpenClaw × Automated Task Decomposition
Driven by multimodal large model reasoning, ROSOrin deconstructs remote commands into action primitive sequences, directly translating ambiguous semantics into precise robotic arm execution via OpenClaw.
Large AI Models × Multi-Point Navigation × Scene Understanding
ROSOrin builds deep spatial cognition through environmental understanding. During multi-waypoint navigation, it autonomously plans optimal paths, avoids obstacles, and analyzes real-time visual feeds to precisely complete navigation and interaction tasks.
Large AI Models × OpenClaw × Intelligent Information Summary
By capturing field-of-view data and analyzing scene data, ROSOrin automatically synthesizes environmental information into intuitive decision reports, closing the intelligent loop from onsite perception to data delivery.
Large AI Models × OpenClaw × Proactive Response
Powered by global situational awareness, ROSOrin proactively investigates anomalies or ambiguous commands, proposing optimal alternatives and requesting authorization to ensure uninterrupted, efficient workflows.
Large AI Models × OpenClaw × Dynamic Low-Level Invocation
Leveraging multimodal AI models to evaluate environments and object states in real time, ROSOrin autonomously invokes low-level algorithms and routines, including chassis kinematics, to maximize efficiency in complex scenarios.
Large AI Models × OpenClaw × Spatial Vector Memory
Driven by environmental memory from daily interactions, ROSOrin retains historical target-scene associations. In similar scenarios, it autonomously recalls and reuses past execution strategies, closing the cognitive-operational loop efficiently.
Key ROS Features
Large AI Model Robot Comparison
Gazebo Simulation
Hiwonder ROSOrin employs ROS framework and supports Gazebo simulation. Gazebo brings a fresh approach for you to control ROSOrin and verify the algorithm in simulated environment, which reduces experimental requirements and improves efficiency.
Body Simulation Control
Through robot simulation control, algorithm verification of mapping navigation can be carried out to improve the iteration speed of the algorithm and reduce the cost of trial and error.
URDF Model Display in RViz
Provide an accurate URDF model that can be visualized using the RViz tool to observe mapping and navigation performance, facilitating algorithm debugging and improvement.
| Aurora930 Pro Specifications | |||||
| Module parameter | Size | 76.5 × 20.7 × 21.8 mm | Imaging performance | Depth data format | 16-bit Raw |
| Baseline | 40mm | Depth resolution/Frame rate | 640×400 @12fps (FOV: 74°×51°) | ||
| Interface | USB 2.0 Wafer connector | RGB data format | NV12 | ||
| Depth accuracy | ±8mm @1m | RGB resolution /Frame rate | 640×400 @12fps (FOV: 74°×51°) | ||
| Working distance | 15–300 cm | IR data format | 8-bit Raw | ||
| Operating temperature | -10℃ to 55℃ | IR resolution /Frame rate | 640×400 @12fps (FOV: 74°×51°) | ||
| Operating humidity | 0% to 95% RH (non-condensing) | Firmware capabilities | Firmware upgrade | Supports USB OTA Update | |
| Operating illuminance | 3–80,000 Lux | Hot start delay | <300ms | ||
| Power supply | 5V±10%, 1.5A | System compatibility | OS compatibility | Linux / ARMv8 / ROS / Windows | |
| Power consumption | Average <1.6W | ||||
| Safety rating | Class 1 Laser safety | ||||
| Aurora930 Pro Specifications | ||
|---|---|---|
| Module parameter | Size | 76.5 × 20.7 × 21.8 mm |
| Baseline | 40mm | |
| Interface | USB 2.0 Wafer connector | |
| Depth accuracy | ±8mm @1m | |
| Working distance | 15–300 cm | |
| Operating temperature | -10℃ to 55℃ | |
| Operating humidity | 0% to 95% RH (non-condensing) | |
| Operating illuminance | 3–80,000 Lux | |
| Power supply | 5V±10%, 1.5A | |
| Power consumption | Average <1.6W | |
| Safety rating | Class 1 Laser safety | |
| Imaging performance | Depth data format | 16-bit Raw |
| Depth resolution/Frame rate | 640×400 @12fps (FOV: 74°×51°) | |
| RGB data format | NV12 | |
| RGB resolution /Frame rate | 640×400 @12fps (FOV: 74°×51°) | |
| IR data format | 8-bit Raw | |
| IR resolution /Frame rate | 640×400 @12fps (FOV: 74°×51°) | |
| Firmware capabilities | Firmware upgrade | Supports USB OTA Update |
| Hot start delay | <300ms | |
| System compatibility | OS compatibility | Linux / ARMv8 / ROS / Windows |
Wrapped Rear Tail Shell
It can effectively protect the PCB circuit and magnetic ring at the end of the motor from external influences, effectively improving the safety and service life of the motor.
Permanent Magnet Brushed Motor
The permanent magnet DC motor has fast starting response speed, large starting torque and smooth speed change.
High-precision Magnetic Encoder
The motor is equipped with a high-precision magnetic encoder, has strong horsepower, high precision, and strong anti-interference ability.
Adapt to Various Scenes
The low speed of 1:90 ratio and the high torque of 15kg.cm enable the motor to adapt to car chassis made of various materials.
Hall Encoder Geared Motor
520 motor comes with high-accuracy encoder, and features strong force and high performance. The built-in AB phase incremental Hall encoder stands out for its high accuracy and anti-interference ability.
Mecanum Wheel
The mecanum wheel has a compact structure and flexible movement, supports 360° all-round movement, and realizes the full lateral movement of the car.
Multi-functional Expansion Board
The expansion board has a built-in IMU sensor which can detects robot posture in real time. There are 2-channel PWM, two keys, a LED, a buzzer, 9-channel serial bus servo interface, two GPIO expansion ports and two llC interfaces on it.
Lithium battery parameters
The fuselage has a built-in 11.1V 6000mAh large-capacity lithium battery to improve the robot's endurance.
Specification Parameters
| Machine model | Mecanum chassis version (Standard) | Ackerman chassis version (Standard) | Four-Wheel Differential Chassis Version (Standard) |
|---|---|---|---|
| Product dimensions | 277x212x166mm | 277x204x169mm | 277x204x166mm |
| Product weight | 2.66kg | 2.15kg | 2.26kg |
| Motor | 520 metal gear reduction motor | ||
| Encoder | High-precision AB quadrature encoder | ||
| Chassis material | Full-metal aluminum alloy chassis with anodized finish | ||
| ROS controller | Jetson Nano / Jetson Orin Nano / Jetson Orin NX / Raspberry Pi 5 board | ||
| Multi-function expansion board | STM32 ROS robot controller, Jetson multi-function expansion board | ||
| Control method | App control, wireless controller control, PC software control | ||
| Depth camera | Aurora 930 Pro 3D depth camera | ||
| LiDAR | STL-19P D500 TOF LiDAR | ||
| Battery | 11.1V 6000mAh 3C lithium battery | ||
| Audio/pickup | WonderEcho Pro Al voice interaction box / 6-mic array module | ||
| Operating system | Ubuntu 18.04 LTS + ROS Melodic / Ubuntu 22.04 / ROS2 Humble | ||
| Software | iOS / Android app | ||
| Communication method | WiFi / Ethernet | ||
| Programming tools | Python / C / C++ / JavaScript | ||
| Storage | 64G TF card (Jetson Nano, Raspberry Pi 5) 128G SSD (Jetson Orin Nano, Jetson Orin NX) | ||
| Servo model | LD-1501MG | ||
| Map/Props | AI sandbox map for autonomous driving, traffic lights, road signs | ||
| Tutorials | Comprehensive tutorials, ROS source code, system image, supporting software | ||
| Mecanum chassis version (Standard) | Size: 277x212x166mm Weight: 2.66kg |
|---|---|
| Ackerman chassis version (Standard) | Size: 277x204x169mm Weight: 2.15kg |
| Four-Wheel Differential Chassis Version (Standard) | Size: 277x204x166mm Weight: 2.26kg |
| Motor | 520 metal gear reduction motor |
| Encoder | High-precision AB quadrature encoder |
| Chassis material | Full-metal aluminum alloy chassis with anodized finish |
| ROS controller | Jetson Nano / Jetson Orin Nano / Jetson Orin NX / Raspberry Pi 5 board |
| Multi-function expansion board | STM32 ROS robot controller, Jetson multi-function expansion board |
| Control method | App control, wireless controller control, PC software control |
| Depth camera | Aurora 930 Pro 3D depth camera |
| LiDAR | STL-19P D500 TOF LiDAR |
| Battery | 11.1V 6000mAh 3C lithium battery |
| Audio/pickup | WonderEcho Pro Al voice interaction box / 6-mic array module |
| Operating system | Ubuntu 18.04 LTS + ROS Melodic / Ubuntu 22.04 / ROS2 Humble |
| Software | iOS / Android app |
| Communication method | WiFi / Ethernet |
| Programming tools | Python / C / C++ / JavaScript |
| Storage | 64G TF card (Jetson Nano, Raspberry Pi 5) 128G SSD (Jetson Orin Nano, Jetson Orin NX) |
| Servo model | LD-1501MG |
| Map/Props | AI sandbox map for autonomous driving, traffic lights, road signs |
| Tutorials | Comprehensive tutorials, ROS source code, system image, supporting software |
Dimensional Diagram
Videos
What is ROSOrin? >>>
Blog
《YOLOv11: X Major Improvements Over YOLOv8》
《The ROSOrin All-in-One Design: Three Robots in One》
《How ROSOrin Is Pushing the Limits of ROS Education Robots with AI》
| Item | Specification |
| Mecanum chassis version (Standard) | Size: 277x212x166mm Weight: 2.66kg |
| Ackerman chassis version (Standard) | Size: 277x204x169mm Weight: 2.15kg |
| Four-Wheel Differential Chassis Version (Standard) | Size: 277x204x166mm Weight: 2.26kg |
| Motor | 520 metal gear reduction motor |
| Encoder | High-precision AB quadrature encoder |
| Chassis material | Full-metal aluminum alloy chassis with anodized finish |
| ROS controller | Jetson Nano / Jetson Orin Nano / Jetson Orin NX / Raspberry Pi 5 board |
| Multi-function expansion board | STM32 ROS robot controller, Jetson multi-function expansion board |
| Control method | App control, wireless controller control, PC software control |
| Depth camera | Aurora 930 Pro 3D depth camera |
| LiDAR | STL-19P D500 TOF LiDAR |
| Battery | 11.1V 6000mAh 3C lithium battery |
| Audio/pickup | WonderEcho Pro Al voice interaction box / 6-mic array module |
| Operating system | Ubuntu 18.04 LTS + ROS Melodic / Ubuntu 22.04 / ROS2 Humble |
| Software | iOS / Android app |
| Communication method | WiFi / Ethernet |
| Programming tools | Python / C / C++ / JavaScript |
| Storage | 64G TF card (Jetson Nano, Raspberry Pi 5) 128G SSD (Jetson Orin Nano, Jetson Orin NX) |
| Servo model | LD-1501MG |
| Map/Props | AI sandbox map for autonomous driving, traffic lights, road signs |
| Tutorials | Comprehensive tutorials, ROS source code, system image, supporting software |
Shipping & Delivery
1. Free Shipping Policy
Hiwonder offers free standard shipping on orders over US $499 (before tax and duties), delivered via trusted carriers such as DHL, UPS, or FedEx.
Most orders arrive within 6 working days after dispatch. You can check if your country is eligible for free shipping in the regions listed below.
Eligible Regions
- Americas: United States, Canada, Mexico, Chile
- Europe: Germany, United Kingdom, Netherlands, Switzerland, France, Austria, Italy, Belgium, Spain, Sweden, Finland, Norway, Portugal, Czech Republic, Poland, Greece, Slovenia, Romania, Denmark, Bulgaria, Luxembourg, Slovakia, Hungary, Ireland, Estonia, Lithuania
- Asia: China, Japan, Singapore, South Korea, United Arab Emirates, Malaysia, Thailand, Philippines, Indonesia
- Oceania: Australia, New Zealand
If your country or region is not listed above, the shipping cost will be automatically calculated at checkout once you enter your shipping address.
2. Customs & Import Fees
All international shipments are subject to local customs regulations. Import duties, taxes, or other fees may be charged by your country's customs authority upon arrival.
Our shipping term is FOB Shenzhen, which means these fees, if any, are the customer's responsibility. The shipping carriers will assist with the customs clearance process to ensure smooth delivery.
If a package is returned due to unpaid customs duties, any related return or reshipping fees will be deducted from your refund.
3. Order Processing Time
Orders are typically processed within 24 hours after placement and shipped the next business day.
Orders placed on weekends or public holidays will be shipped on the next working day.
Once shipped, you will receive a confirmation email with tracking information.
