Alexander Kirillov's Yozh Is an Expandable, Compact, Open Source Robot for Education and More

Maker and educator Alexander Kirillov has put together an open source tracked robot design for a summer camp's robotics class, inspired by the Pololu Zumo — but with, he claims, many improvements: the Yozh.

"Sure, there are dozens of existing robots. However, when I started choosing a robot for my summer camp, none of them seemed to be quite what I wanted," Kirillov explains, highlighting requirements including a compact size, support for Python, at least five on-board sensors, and expansion capabilities. "I looked at Pololu's Zumo and 3 Pi+, DFRobot's Maqueen Plus, and Pimoroni's Trilobot. None of them satisfied all my requirements, so I built my own. It took a while — designing the boards, writing the library, and more, but at the end it was worth it."

The Yozh itself is a short robot running on two rubber tracks, measuring just 4.8cm (around 1.9") at its highest point. The Pololu motors are driven from a custom electronics board, which also forms the chassis, with an Adafruit ESP32-S3 Feather as the primary microcontroller running CircuitPython and a Microchip SAM D21 as a coprocessor for low-level operations including interfacing with the motors' encoders and running a closed-loop proportional–integral–derivative (PID) control system for maintaining a stable speed.

Its standard sensor package includes a seven-sensor infrared (IR) reflectance array for line-following tasks, two front-facing STMicro VL53L0X time-of-flight (TOF) distance sensors, and a six degrees of freedom (6DOF) inertial measurement unit (IMU) with sensor fusion in firmware. Other features include two RGB LEDs and a buzzer for feedback, four RGBW LEDs as headlights, and a user interface based around a 240×135 color TFT display and three buttons.

For expansion, the Yozh chassis — named for the Russian for "hedgehog" includes two servo connectors, two Qwiic/STEMMA QT compatible I2C headers, and a range of general-purpose input/output (GPIO) pins. "Yozh is compatible with mechanical attachments (grabber, forklift…) by DFRobot," Kirillov adds, while also demonstrating an on-device computer vision application — the recognition of a container marked with an AprilTag — using an add-on DFRobot Huskylens camera.

Full details on the Yozh are available on the project website, while the design files and source code are published to GitHub under the permissive MIT license; more information can be found on Kirillov's Hackaday.io page.