LEAP ROS2 (low-cost open ROS 2 learning car: LEAP_ROS / LEAP_ROS_MAX) · A ROS 2 learning car built around an ESP32-S3 driver board with lidar, ultrasonic sensing and a Wi‑Fi camera, in two configurations: LEAP_ROS is a two-wheel differential base (65 mm wheels, 131.7 mm track, 4680 encoder counts per turn) and LEAP_ROS_MAX a four-wheel mecanum base. Overall size and weight are not given
A ROS 2 learning car built around an ESP32-S3 driver board with lidar, ultrasonic sensing and a Wi‑Fi camera, in two configurations: LEAP_ROS is a two-wheel di…
- Maker
- Chuyun (GitHub czu963889306-dev, Bilibili “Chuyun Tech 2001”)
- Form
- A ROS 2 learning car built around an ESP32-S3 driver board with lidar, ultrasonic sensing and a Wi‑Fi camera, in two configurations: LEAP_ROS is a two-wheel differential base (65 mm wheels, 131.7 mm track, 4680 encoder counts per turn) and LEAP_ROS_MAX a four-wheel mecanum base. Overall size and weight are not given
- For
- Students and beginners who want a low-cost start in ROS 2, students and teachers doing course projects, graduation projects, competition prototypes or teaching cases, and developers extending vision, navigation, SLAM or control on it (README)
- Sensors
- The driver board carries an LSM6DS3TR-C 6-axis IMU and two AT8236 motor drivers; the firmware reads the motor encoders, a Camsense lidar (115200 baud), an HC-SR04 ultrasonic sensor and battery voltage, and drives one servo and status lights; a second ESP32-S3 board (LEAP_CAM) streams MJPEG video. The published BOMs contain no microphone; the AI voice-assistant client asks for a USB microphone and speaker on a Linux main board
- Connectivity and app
- The ESP32-S3 firmware v2.1 has a Wi‑Fi provisioning page and switches between micro-ROS (Wi‑Fi UDP or UART at 921600 baud) and MAVLink (UDP or UART), and takes an I²C gamepad; the ROS 2 side runs in a virtual-machine image on a PC (two-wheel version) or on a Rockchip main-board image (MAX and the version with a main board), with mapping, Nav2 navigation and OpenCV samples; the camera streams on port 81; the creator's site offers in-browser firmware flashing. No phone app. The optional AI voice assistant calls Alibaba Cloud Bailian
- Subscription
- Open-source project, no subscription; the optional AI voice assistant needs the user's own Alibaba Cloud Bailian DASHSCOPE_API_KEY
- Openness
- The main repo leap_ros2 carries an Apache‑2.0 LICENSE and holds the schematics (PDF) of the driver and camera boards, BOMs, JLC EDA project files and 3D-printed shell STLs; but its README says commercial use is forbidden without permission and the project is for learning, exchange and research only, which conflicts with Apache‑2.0, so commercial use needs the creator's consent. The software lives in separate repos: the ROS 2 workspace leap_ros_ws is GPL‑2.0, while the firmware for LEAP_ROS and MAX, the camera firmware and the voice-assistant client declare no licence
- Privacy
- The camera streams to the ROS 2 host over local Wi‑Fi. The optional voice assistant sends microphone audio to Alibaba Cloud Bailian speech recognition (while it waits for the wake word, recordings that pass its VAD are uploaded too, and the wake word is matched after cloud recognition), text to Qwen, and pictures or camera frames to image understanding, with CosyVoice for speech.
- Not stated
- Overall size and weight, battery spec and runtime, the ROS 2 distribution, what each configuration's reference cost includes, assembly time, recommended age
Vendor claims
- A low-cost open-source ROS 2 learning robot: about CN¥200 for LEAP_ROS and about CN¥600 for the mecanum LEAP_ROS_MAX
- Step by step from hardware assembly, firmware flashing and ROS 2 setup to node communication, sensor integration and motion control
- Suited to getting started, teaching, competition preparation, course projects and graduation projects
- Schematics, BOMs, JLC EDA files and the printed shell are published, so you can order boards, solder, source parts and print the shell yourself
Parts
- ESP32-S3 · Provides Wi-Fi camera video transmission
- HC-SR04 · Measures and reports ultrasonic distance data
Sources
- github.com/czu963889306-dev/leap_ros2
- github.com/czu963889306-dev/leap_ros2/blob/main/LICENSE
- github.com/czu963889306-dev/leap_ros2/tree/main/%E7%A1%AC%E4%BB%B6
- github.com/czu963889306-dev/leap_ros2/blob/main/%E7%A1%AC%E4%BB%B6/7.BOM/BOM_LEAP_ROS.xlsx
- github.com/czu963889306-dev/leap_ros2/blob/main/%E7%A1%AC%E4%BB%B6/7.BOM/BOM_LEAP_CAM.xlsx
- github.com/czu963889306-dev/leap_ros_mcu_driver
- github.com/czu963889306-dev/leap_ros_max_mcu_driver
- github.com/czu963889306-dev/esp32_s3_wifi_camera
- github.com/czu963889306-dev/leap_ros_ws
- github.com/czu963889306-dev/voice_assistant_client
- maturo.cn/
- maturo.cn/about/
- maturo.cn/products/leap-ros/
- maturo.cn/site-content.json
- bilibili.com/video/BV1nNNb6cER7
- bilibili.com/video/BV1uxp46SEoU