Our Robots Colleagues

Individual Robots

Unitree Go2

Unitree Go2

The Unitree Go2 is a research-grade quadruped robot capable of reaching speeds up to 3.7 m/s, climbing ramps up to 40°, and stepping over 16 cm obstacles. It integrates 4D LiDAR, high-definition cameras, and foot-end force sensors for real-time perception. The robot includes an 8-core processor and support for NVIDIA Orin NX 16GB modules. The robot operates for up to 4 hours on a 15,000 mAh smart battery and includes standard connectivity options such as Ethernet, USB, and SBUS. Go2 provides full SDK access with ROS 2 compatibility and is suitable for learning-based control, especially RL.

Dobot Nova 5

Dobot Nova 5

The Dobot Nova 5 is a compact 6-DoF collaborative robotic arm with a working radius of 850 mm and a maximum payload of 5 kg. It offers ±0.05 mm repeatability and a TCP speed of up to 2 m/s, making it well-suited for precision tasks in research and industry. Nova 5 supports full-angle installation and features built-in I/O, RS485, and Modbus communication, enabling seamless integration with ROS-based control and intelligent systems. It is commonly used in interactive robotics, light-duty automation, and AI-enhanced service scenarios.

myAgv

myAgv from Elephant Robotics

myAGV is a ROS-compatible mobile robot developed by Elephant Robotics. It features omnidirectional movement with Mecanum wheels, a built-in Lidar for SLAM (gmapping and cartographer), and a 5-MP camera for vision-based tasks. The platform supports modular extensions, including robotic arms, and is suitable for research in navigation, mapping, and collaborative robotics.

myCobot 280 Pi

myCobot 280 Pi

myCobot 280 Pi is a lightweight 6-axis robotic arm designed for AI education and early-stage research. Powered by a Raspberry Pi 4B and running Ubuntu Mate with built-in support for Python, ROS1, and ROS2, it enables hands-on learning in kinematics, machine vision, and algorithm development. Despite its compact size (860 g), it supports various end-effectors and modular accessories, making it ideal for use in smart homes, labs, and embedded AI applications. Programming is supported through Python, graphical tools like myBlockly, and full access to the ROS ecosystem.

DJI Tello

DJI Tello

The DJI Tello is a compact, programmable drone ideal for teaching and experimenting with aerial robotics. It supports Python, Scratch, and ROS, making it a popular tool for education and research in perception, control, and autonomous flight. With a 5-MP camera, 13 minutes of flight time, and stable indoor hovering, Tello is a practical platform for exploring vision-based navigation and RL tasks at low cost.

Robots in Action

Different possible real-life scenarios that our robots might be able to perform.

We used Sora AI that generated those images to better visualise/represent those scenarios that are still under research/development.

Robot dog rescue mission after earthquake
Rescue scenario — Autonomous Unitree Go2 quadruped robot searching for survivors after a powerful earthquake
Dobot Nova5 robotic arm and myAGV mobile robot performing a package pickup
Warehouse scenario — Autonomous Dobot Nova5 robotic arm carrying a package with the help of the autonomous myAGV mobile robot
F1 single-sitter style RC cars performing autonomous driving
Autonomous driving scenario — RC cars, F1 style, having a cooperative/non-egocentric behaviour towards each other