Mobile and ROS robots

TurboPi

A Raspberry Pi mecanum robot car for AI vision in Python, with ROS2, voice and AI models in higher kits.

Raspberry Pi
Suggested for secondary
Suggested for university

TurboPi is an open-source AI vision car for beginners, built on a Raspberry Pi with a mecanum wheel chassis and a 2DOF HD wide-angle camera. It is programmed in Python, with OpenCV and YOLO26 for image processing and object detection, and handles colour recognition, object tracking, line following and model autonomous driving with road signs and traffic lights. The Standard Kit runs on a Raspberry Pi 5 or 4B. The Advanced and Ultimate kits run ROS2 on a Raspberry Pi 5 and add the WonderEcho Pro AI voice interaction box, multimodal large AI models and the OpenClaw agent; the Ultimate Kit also adds a 2DOF robotic arm with a metal gripper.

Zeta42 is an authorised Hiwonder partner in the Middle East.

Configurations

Choose how it ships.

Model: Standard Kit · Advanced Kit · Ultimate Kit
Controller: Without Raspberry Pi 4B · With Raspberry Pi 4B 4GB · Without Raspberry Pi 5 · With Raspberry Pi 5 2GB · With Raspberry Pi 5 4GB · With Raspberry Pi 5 8GB · With Raspberry Pi 5 16GB

What it does

Built for.

Mecanum chassis
Four mecanum wheels move TurboPi forwards, sideways, diagonally and on the spot, and a hard aluminium alloy chassis protects the control board.
Pan-tilt camera
Two anti-blocking servos turn the HD wide-angle camera through 180° horizontally, and with the chassis TurboPi can see all the way round.
Vision in Python
YOLO26, MediaPipe and OpenCV handle object detection, colour recognition and tracking, face recognition and gesture control, with no complex setup.
Model autonomous driving
A 4-channel infrared line follower on I2C reads lines from 0.5 cm to 6 cm wide through right angles, T-junctions and crossroads without using the Raspberry Pi's CPU. YOLO26 recognises road signs, and OpenCV reads traffic lights.
Voice and large AI models
In the Advanced and Ultimate kits, ChatGPT or Gemini turns spoken commands into actions, describes the scene and drives tracking and patrolling. OpenClaw takes commands from a PC or the app and breaks them into tasks; in the Ultimate Kit it can also direct the arm to pick up and carry objects.
Control and expansion
Control it from the WonderPi app on iOS or Android or from a PC over VNC, and extend it with electronic modules and LEGO-compatible blocks.

Where it is used

Applications.

A first course in Python and computer vision on Raspberry Pi
Model autonomous driving with line following, road signs and traffic lights
Introductory ROS2 and large AI model projects (Advanced and Ultimate Kit)
Voice-driven pick-and-carry tasks (Ultimate Kit)

Videos

See it run.

Specifications

The full sheet.

Product dimension: 187 × 162 × 139 mm
Weight: 0.8kg
Material: Metal bracket
Camera resolution: 480P
Camera pan-tilt: 2DOF, horizontal rotation angle 180°
Hardware: Raspberry Pi 4B / 5 and Raspberry Pi expansion board
Controller and system by kit: Standard Kit: Raspberry Pi 5 or 4B, Debian; Advanced and Ultimate Kit: Raspberry Pi 5, Ubuntu, ROS2 Humble (Docker)
Power supply: 2 × 18650 LiPo battery 1800mAh (Standard Kit); 2 × 18650 LiPo battery 2200mAh (Advanced and Ultimate Kit)
Working hours: About 60min
Voice: WonderEcho Pro AI voice interaction box (Advanced and Ultimate Kit)
Robotic arm: 2DOF robotic arm with metal gripper (Ultimate Kit)
Software: VNC software (PC) + WonderPi app (iOS / Android)
Communication method: Wi-Fi, Ethernet
Servo: LFD-01 anti-blocking servo
Control method: PC / phone control
Package size: 27 × 23 × 7 cm
Package weight: About 1.2kg