The goal of this R&D project was to develop a fully autonomous robotic system capable of navigating complex environments and performing precise, automated imaging tasks. The primary challenge was to enable a mobile robot to intelligently map its surroundings, identify a target object (e.g., a vehicle), and execute a pre-defined path around it to capture images from optimal angles, all while dynamically avoiding obstacles.:

Solution: I engineered a sophisticated software stack using the Robot Operating System (ROS2) to bring intelligent autonomy to the HiWonder Jetacker robot. My solution was built on three core pillars:

Advanced Perception & Sensor Fusion: I fused data from a YDLidar G4 for 2D environmental mapping and an Orbbec3D Astra depth camera for 3D obstacle detection. This allowed the robot to build a rich, real-time understanding of its environment, reliably distinguishing between static obstacles and the target object.

Intelligent Navigation & Path Planning: Using the ROS2 Navigation Stack, I implemented robust algorithms for SLAM (Simultaneous Localization and Mapping) and dynamic obstacle avoidance. The robot could autonomously create a map of an unknown area and navigate it safely.

Custom Application Logic: I developed a custom C++/Python node that defined the robot's primary mission. This algorithm enabled the robot to:

Identify the target vehicle within its mapped environment

Calculate a precise geometric path (e.g., a circle or rectangle) around the vehicle

Execute the path while triggering the camera at specific waypoints to ensure optimal image capture

Framework: ROS2 (Robot Operating System)

Languages: C++, Python

Key Concepts: Sensor Fusion, SLAM, Autonomous Navigation, Obstacle Avoidance

Hardware: HiWonder Jetacker, YDLidar G4, Orbbec3D Astra Depth Camera

Libraries: ROS2 Navigation Stack, OpenCV

Project Outcome I successfully developed and demonstrated a fully autonomous robotic system. The robot could reliably map its environment, detect target objects, and execute complex imaging patterns with high precision and adaptability. This project served as a successful proof-of-concept for automated robotic inspection and data collection, showcasing my deep expertise in ROS2, sensor fusion, and the development of intelligent robotic behaviors.