Unitree Go2 Pro Robot Dog Tested for AI, Balance, and Autonomous Features

Unitree’s Go2 Pro robot dog was put through a series of challenging tests within its first 24 hours to evaluate its AI capabilities, balance, and autonomous functions. The testing took place in various environments, including an obstacle course built inside the CNET office, outdoor stairs, rocky hills, and a park setting.

The obstacle course was designed to test the robot’s AI obstacle avoidance mode, which is critical for autonomous navigation in complex environments. The Go2 Pro demonstrated its ability to detect and maneuver around obstacles, showcasing the effectiveness of its AI-driven sensors and algorithms.

Balance was tested by running the Go2 Pro up and down stairs as well as across uneven, rocky hills. These tests are essential for assessing the robot’s stability and mobility in real-world terrains, which are often unpredictable and challenging for quadruped robots.

The autonomous follow mode was evaluated during a stroll through a park, where the robot dog was tasked with following a human subject without manual control. This feature highlights the robot’s potential for applications such as personal assistance, security, and interactive robotics.

To gauge social interaction and public reaction, the Go2 Pro was brought to a dog park. Observers compared its behavior and appearance to real dogs, providing insights into the robot’s acceptance and potential for companionship or service roles.

Unitree’s Go2 product line includes three main models: the Go2 AIR, which is the most affordable option; the Go2 Pro, which offers enhanced features like AI follow mode; and the Go2 EDU, designed for developers and educational institutions to program and teach new capabilities to the robot.

The testing revealed both the strengths and challenges of the Go2 Pro, providing valuable feedback for future improvements. The unboxing and initial trials, documented in accompanying videos, offer a transparent look at the robot’s capabilities and real-world performance.

Unitree’s advancements in quadruped robotics demonstrate significant progress in AI-driven mobility and autonomous functions, paving the way for applications in security, research, education, and personal assistance. These developments highlight the growing intersection of robotics and AI technologies in practical, everyday contexts.