ROSPug Quadruped Bionic Robot Dog Powered by Jetson Nano ROS Open Source Python Programming

• Driven by Jetson Nano and Intelligent Servos: ROSPug is a Jetson Nano-powered quadruped robot with 12 high-voltage intelligent servos for precise, fast, and powerful movement.
• AI Vision for Versatile Applications: Equipped with an HD camera and OpenCV, ROSPug supports AI tasks like object recognition, line following, obstacle avoidance, and face detection.
• Multiple Control Methods & Live Feed: Control via WonderROS app (Android/iOS), PC, or wireless PS2. Live camera feed provides a first-person view in the app.
• Adjustable Gait Planning: With inverse kinematics, you can fine-tune leg movements for various gaits, and access the source code for customizations.

Product Description:

ROSPug is smart quadruped robot dog built upon Robot Operating System (ROS). It is equipped with 12 high-voltage strong-magnetic serial bus servos and integrates a range of high-performance components, including NVIDIA Jetson Nano controller, TOF Lidar, HD camera, IMU sensor, OLED display, and more. Featuring self-developed dynamic balancing kinematics algorithm, it can switch seamlessly between multiple gaits.

ROSPug supports Gazebo simulation, providing users with a valuable platform to learn and validate quadruped kinematics algorithms and path planning. Thanks to the robust computing power of its controller, ROSPug can perform tasks such as SLAM mapping navigation, path planning, dynamic obstacle avoidance, climbing, obstacle bypassing, and many other applications. We also offer solutions for expansion of ROSPug's capabilities, including deep learning, machine vision, and other development projects to meet users' specific needs.

① 12 DOF Aluminium Alloy Structure

ROSPug employs 12 high-performance servos, distributed across its elbow, shoulder, and hip joints of each leg, closely mimicking the post of a real quadruped animal. Its entire body is crafted from aluminum alloy, with the calf joint reinforced by metal bearings, ensuring both low weight and high strength.

② High-performance Hardware

ROSPug is powered by Jetson Nano and features high-performance Lidar and an HD wide-angle camera, enabling the validation of various creative AI applications.

③ Link Structure Enhanced Joint Efficiency

ROSPug features a link structure design that enhances the speed of the calf joint and ensures smooth motion without interference, thereby extending the leg's rotation range.

④ High-performance Intelligent Bus Servo

ROSPug features high-performance intelligent serial bus servos with a torque of 30KG, providing exceptional accuracy, data feedback, easy wiring, and support for a robust 12V voltage power supply.

1) Dual-Controller Design for Efficient Collaboration

ROSPug features a dual-controller design that combines the advanced AI capabilities of the Jetson Nano with the high-frequency control functions of the MCU. This integration enhances operational accuracy, enabling the system to tackle more complex challenges and explore a wider range of possibilities.

2) Inverse KinematicsGait Planning

Support Walk, Amble and Trot Gaits. Walk, Amble, and Trot gaits can be achieved by fine-tuning the touchdown time, lifttime, lifted height of each leg and the timing of switching between front and hind legs.

① Link Inverse Kinematics

ROSPug comes with a visual PC software for action editing, allowing users to define end coordinates for each leg. The robot then employs an inverse kinematics algorithm to calculate angles ofach servocreatingamotion profile that enables its feet to rech the desired target locations.

② Adjustable Walking Height, Speed and Pose

ROSPug's height, inclination, pitch angle and roll angle can be freely adjusted. Users can clectively adjust these variables to achieve both waking and turning motions.

③ Self-balancing and Yaw Angle Correction with IMU

ROSPug integrates an IMU sensor capable of real-time pose monitoring and data coquisition for closed-loop control. Regardless of the plane's inclination, ROSPug can promptly just its joints to maintain balance.

3) Support Various Control Methods

① PC Software Control

Using graphical PC software, you can effortlessly control servos and customize actions by simply dragging sliders, without the need for programming.

② Support Python Programming

All intelligent Python code is open source, with detailed annotations for easy self-study.

③ Coordinate PC Software Control

To support users in exploring ROSPug's capabilities, we offer detailed quadruped kinematics analysis, ROS-based inverse kinematics functions, and parameter debugging software.

④ APP Control

Android and iOS mobile APP are available. Via the APP, you can remotely control the robot and view what the robot sees.

4) Machine Vision

① Line Following

ROSPug can recognize red lines, and then calculate the location of the line so as to adjust the walking gait and realize line following.

② Drifting in Circle

By utilizing visual recognition to identify the circular pile and measuring the distance between itself and the pile using Lidar, the robot can adjust its moving direction to perform circular drifting.

③ Up and Down Stairs

Through independent vision judgment, the position of the stairs in front is identified and the scene of autonomously going up and down the stairs is realized.

④ Ball Shooting

ROSPug utilizes OpenCV to determine ball's position and employs a PID algorithm for real-time ball tracking. This approach allows it to fine-tune its gait to achieve accurate ball shooting, considering the distance from the target and the target's location.

⑤ Vision Recognition Target Tracking

ROSPug has a built-in HD wide-angle camera on its head, which can recognize and locate different targets, thereby realizing creative Al gameplay such as Tag recogniticn, face recognition, color tracking, and KCF target trcking.

⑥ MediaPipe Development Action Capturing

Developed upon MediaPipe algorithm, ROSPug can identify human body features, facilitating interactions such as face detection, emotion recognition, gesture recognition, and human body recognition, among others.

5) Lidar Functions - ROSPug is equipped with a Oradar MS200 Lidar

① Lidar Mapping and Navigation

ROSPug can realize advanced SLAM functions by lidar, including localization, mapping and navigation, path planning, dynamic obstacle avoidance, Lidar tracking and Lidar guarding, etc.

② Various 2D Lidar Mapping Methods

TOF Lidar employs mapping algorithms such s Gmapping, Hector, Karto, ond others for mapping purposes. Additionaly, it supports fixed-point navigation, multi-point navigation, and TEB path planning.

③ Multi-Point Navigation

ROSPug is equipped with a high-accuracy Lidar that provides real-time environmental detection. It supports both fixed-point navigation and multi-point navigation, making it suitable for complex navigation scenarios.

④ Dynamic Obstacle Avoidance

Using TOF Lidar, ROSPug can detect obstacles during navigation and inteligently plan its path to effectively avoid them.

⑤ Lidar Tracking and Guarding

ROSPug can work wvith Lidar to scan and subsequently track a moving target ahead. ROSPug utilizes TOF Lidar to scan the secured area. Upon detecting cn intruder, it will automgticlly turn toward the intruder and activate an alarm.

6) Gazebo Simulation

ROSPug utilizes the ROS framework and offers Gazebo simulation support. Gazebo provides a novel approach for controlling ROSPug and validating algorithms within a simulated environment, reducing the need for physical experiments and enhancing efficiency.