2026 Global Developer Pioneer Conference Embodied Intelligence Skills Competition: Huachuang Zhiqu, Banxing Technology and other supply chain companies appear

For a humanoid robot, the joint modules alone account for nearly half the cost. The dancing robot you see is backed by dozens of reducers, motors and lead screws working together at their limits. Every bit of progress in components is crucial to the entire industry. The 2026 Global Developer Pioneer Conference and International Embodied Intelligence Skills Competition is underway. At the competition site, besides nearly a thousand teams competing "for real" in real-world scenarios, nearly 200 supply chain companies and other supporting enterprises have gathered. A single competition has indeed brought out some "hidden skills." The few-tooth-difference reducer is "maxed out on hardcore specs," yet it came to the quality inspection track to show off its "embroidery skills." This morning, beside the quality inspection arena, the reporter met a contestant from Huachuang Zhiqu who had just finished competing. Competing for the first time, they brought a robot equipped with their self-developed "few-tooth-difference reducer." A reducer is a precision transmission device connecting the servo motor and the actuator, and is the core of a robot's joints. Currently, robots mainly use harmonic reducers, RV reducers and planetary reducers. Among them, harmonic reducers are small in size and high in precision, making them the mainstream choice for joints such as humanoid robot wrists, forearms and waists; RV reducers have good rigidity and high torque, and are mainly used for heavy-load joints such as industrial robot bases, upper arms and shoulders; planetary reducers have a mature structure and the highest cost-effectiveness, making them suitable for parts with lower precision requirements. Contestant Mr. Yang said that Huachuang Zhiqu's few-tooth-difference reducer reaches more than five times that of harmonic reducers in parameters such as stiffness, end load capacity and service life. Targeting the different needs of industrial robots, humanoid (bionic) robots, high-end machine tools, medical robots and other fields, the company has developed corresponding joint transmission solutions. Interestingly, the scenarios that best leverage the advantages of the few-tooth-difference reducer are those with high demands on a robot's ability to withstand instantaneous impacts and load capacity, such as handling or rescue; yet Huachuang Zhiqu, competing for the first time, chose the quality inspection track that tests a robot's "dexterity." Huachuang Zhiqu is not that concerned about winning a place. "Taking part is what matters," Mr. Yang said with a smile. Their company is right next to the Zhangjiang Science Hall; after this site was converted into a permanent embodied intelligence scenario training space, they have the advantage of proximity and can make good use of "real-scenario" data. Judging from the on-site competition, the team performed well, and the entire operation was very smooth. In the future, Huachuang Zhiqu is very likely to enter the robot body market with the few-tooth-difference reducer as its core. If you can't buy it, build it yourself: an 800 N·m joint motor + 1000 Hz electronic control, "full-stack cheating" At the competition site, the Banxing Technology booth was always surrounded by people. This "low-key" robot ODM (original design manufacturer) has been "seen." Banxing Embodied was founded in 2022, but the team began research and development of humanoid robots as early as 2020. With no high-torque motors and high-frequency electronic controls available on the market that could meet the needs of humanoid robots for backflips, Banxing began self-developing from joint motor modules, tried cycloidal and planetary solutions, produced a planetary joint motor with a peak torque of 800 N·m, and also developed an in-motor water-cooling solution; combined with its self-developed 1000 Hz electronic control system, it produced a first-generation prototype. Banxing's core advantage lies in "full-stack self-development." It is one of the very few ODM solution providers that has participated in and completed a robot marathon, and one of the teams that simultaneously possesses whole-machine design, core motors, motion control algorithms and scenario software delivery capabilities. At present, the company has provided general-purpose humanoid robot ODM services to twenty or thirty B-end users and individual developers, with users in the automotive industry accounting for the highest proportion; its solutions cover white-label whole machines, joint motor modules, motion control algorithms and application software development. Because Banxing also serves individual developers, its booth is often "surrounded" by groups of children accompanied by their parents, and the children can learn a lot about robots on site. Teaching robots to learn a doctor's "first-person view" and sending experts to distant places Today's robots have reached a stage where technical details matter more and more. Linyan Medical is a company that relentlessly pursues technology. Its core product is an AR spine surgery navigation system. In August this year, the company completed the world's first humanoid robot spinal animal laminectomy experiment on pigs. The real-time information through the skin surface, two-dimensional tomographic information and surgical instrument movement information that doctors see through AR glasses are restored 1:1 to the robot for training, allowing it to see the same first-person view as the doctor, learn the doctor's hand instrument movements, and understand the real-time positional relationship between instruments and surgical operations. Humanoid robots take the doctor's place in receiving radiation and, through remote teleoperation, allow expert resources to be used even in remote areas. Brand manager Ren Xu said that experiments have currently proven that the robot can complete the step of removing the lamina on a patient's spinal segment, and subsequent iterations will have higher precision and lower latency, ultimately to solve the problem of universal access to healthcare and enable more people to use good medical technology. Being low-key does not mean having no special skills. This robot skills competition, with no script and real on-the-job work, has brought a group of "hidden players" in the robot industry chain to the forefront.