humanoid.guide introduces Sharpa Robotics’ wheeled humanoid North

humanoid.guide reported on September 28, 2026, that it had introduced Sharpa Robotics’ wheeled humanoid North. We have not been able to verify the primary source, so readers should consult the original for details.

Sharpa Robotics presented North as a wheeled humanoid featuring Jetson Thor, an OS compatible with ROS 2 and tactile hands. This shows that robots are being evaluated not as standalone machines, but as integrated designs combining computing, control and touch. The company’s claims are based on humanoid.guide’s coverage, so the robot’s actual performance requires separate verification.

The central fact here is that humanoid.guide introduced Sharpa Robotics’ North. Based on the information available, the coverage does not point to a new plan or business shift beyond describing the robot and its features. It is best read as an account of the machine’s components, including a wheeled base paired with a humanoid upper body, NVIDIA Jetson Thor and a Linux-based OS compatible with ROS 2. Readers should consult the original for details. In a related article, we reported on September 28, 2026, that humanoid.guide introduced D01, another Sharpa Robotics machine described as 162 centimeters tall and weighing 40 kilograms. North is 160 centimeters tall and weighs 90 kilograms, and differs in its wheeled mode of mobility. This suggests that the company may be presenting machines with different configurations for different uses, rather than focusing on a single model. From an industry perspective, North is presented as a system that includes not just the robot itself, but also its computing platform and software layers: NVIDIA Jetson Thor, Ethernet, Wi-Fi, an OS compatible with ROS 2, and CraftNet, a model connecting vision, touch, language and action. This suggests that robot differentiation is shifting beyond mechanical design to how perception, touch and action are implemented together. However, the article alone does not establish the robot’s reliability in real-world operations, the scope of its training data or the tasks it can perform autonomously. Questions to examine next include how long North can sustain tasks in commercial settings, which of its 67 degrees of freedom matter in practical use, and which tasks CraftNet performs well. For now, these points remain outside the scope of the introductory article.

For robotics developers in Japan, North’s combination of a wheeled base, humanoid upper body, tactile hands and ROS 2 compatibility could serve as a point of comparison for how to design mobility and manipulation systems. However, the article makes no direct mention of Japanese companies or the Japanese market, so its relevance is limited to comparing design approaches.