AVIC Unveils Fengxing-001 Humanoid Prototype for Aviation Manufacturing

humanoid.guide reported on Nov. 3, 2026, that Aviation Industry Corporation of China (AVIC) had developed the Fengxing-001 humanoid prototype. It is said to stand 180 centimeters tall, weigh 80 kilograms, carry 20 kilograms while walking and have a top speed of 3 kilometers per hour. The robot was designed for aviation manufacturing, aircraft maintenance and support work, and was reportedly unveiled at the 2026 World Robot Conference.

Although the robot was designed for aviation manufacturing and maintenance, its reported manipulation score was 1 out of 5, and no test conditions were provided to substantiate its practical ability to assemble products. To assess its suitability for aviation applications, key questions include the conditions under which its carrying capacity was measured and the specific evaluation results for maintenance and assembly tasks.

The prototype combines an application focus on aviation with leg motors derived from aviation engine technology and an electrohydraulic hybrid drive. At the same time, its reported skills assessment totaled 3 out of 10, with a score of 2 out of 5 for navigation and 1 out of 5 for manipulation. What can be gleaned at this stage is the coexistence of a design intended for aviation work and an assessment that describes its mobility and manipulation capabilities as still under development. The reported information does not demonstrate deployment in the field or the replacement of workers. In terms of dimensions, carrying capacity and speed, the Fengxing-001 is reported to stand 180 centimeters tall, weigh 80 kilograms, carry 20 kilograms while walking and have a top speed of 3 kilometers per hour. These figures indicate the robot’s size and movement conditions, but do not show under what loads, shapes or environmental conditions its performance was verified when handling actual aircraft parts or tools. Obstacle avoidance, jumping and landing, and product assembly are listed as capabilities, but their respective test conditions and reproducibility are also unknown. Listing capabilities should not be equated with achieving continuous work or safe operation in the field. From an industrial perspective, a key issue is the link between the stated design intent for aviation manufacturing and maintenance and the robot’s drive system and the origins of its leg motors. The reported specifications alone are insufficient to assess how using aviation engine technology in the robot’s legs contributes to its performance. The difference between its navigation and manipulation scores also indicates the need to distinguish between moving through environments with obstacles and interacting with objects. Even if mobility functions are listed, the low manipulation score does not provide evidence that the full task, including product assembly, can be performed. The report describes the robot as a prototype but does not provide information on mass-production plans, operational results, work time, safety, evaluation methods or specific tasks. The next points to verify are the walking conditions under which its 20-kilogram carrying capacity was measured, which processes its assembly capability covers, and the criteria used to calculate its evaluation scores. Only once this information is available will it be possible to distinguish the design intent for aviation applications from its suitability for actual work. For now, it is appropriate to treat the reported figures and assessments as prototype-stage information.