On December 30th, the award ceremony for the 2025 Shenzhen Intelligent Robot Dexterous Hand Competition was grandly held in Nanshan District, Shenzhen. Leveraging the VLA (Vision-Language-Action) model, high-degree-of-freedom dexterous hands, and the innovative fusion sensor system Active Camera provided by ROBOSENSE (02498), Shenzhen Gesong Technology Co., Ltd. (Gesong Technology) broke the completion limits for long-range instant delivery tasks for robots at the competition, securing the championship. This victory validates ROBOSENSE's industry-leading full-stack technological capabilities in robot hand-eye coordination, while also demonstrating its commercial strength in achieving efficient deployment for real-world application scenarios and creating industrial value alongside partners. Zhang Lin, Secretary of the Party Leadership Group and Director of the Shenzhen Science, Technology and Innovation Commission, presented the award medal to the champion, Gesong Technology.
On December 24th, ROBOSENSE quietly released a major technical preview through its official WeChat account via a video filled with holiday spirit. In the video, a robot accurately performed a series of连贯 actions—receiving a gift from Santa Claus, autonomously operating an elevator, delivering the gift upstairs, and finally hanging it on a Christmas tree—all smoothly and naturally, demonstrating highly sensitive environmental perception and fine manipulation capabilities. Industry analysis indicates that the hand-eye coordination solution previewed by ROBOSENSE will deeply integrate its core self-developed technologies, including VLA, Active Camera, and dexterous hands, targeting scenarios requiring highly flexible automated operations, such as last-100-meter delivery, smart manufacturing, warehouse logistics, and commercial services. ROBOSENSE revealed that this complete technical solution will be exhibited at the International Consumer Electronics Show (CES 2026) in January 2026.
As the first professional competition in China focused on the R&D and real-world deployment of dexterous hand technology, this year's event coincided with the critical juncture of the "first year of robot mass production." Since its launch, it attracted enthusiastic participation from 53 high-level teams from universities, research institutes, and key enterprises in Shenzhen, Hong Kong, Hangzhou, and elsewhere, sparking a wave of technical competition centered on "the last centimeter" of operational capability. Ultimately, with the support of ROBOSENSE's dexterous hands, the robotic "eye" Active Camera, and VLA model technology, Gesong Technology stood out from the numerous competing teams to win the championship. During the competition, complex variables such as pedestrian traffic, shared elevator use, and changing lighting conditions thoroughly tested the robots' stability, environmental adaptability, and generalization capabilities. In particular, the task of "placing a folded takeaway box into a recycling slot with a height of only 3 centimeters" was widely recognized as the most technically challenging aspect of the event, described by participants as a "challenge to the limits of human-like hand operations." Empowered by ROBOSENSE's AI robot technology, the Gesong robot autonomously completed the entire task sequence—from unboxing the delivery, folding the packaging, and navigating, to riding the elevator and finally making the delivery—with fluid movements and a high completion rate, ultimately winning the competition championship.
Gesong Technology's championship victory showcases that ROBOSENSE's AI robot technology already possesses core capabilities including "multi-modal data collection, efficient edge-side deployment, small-sample generalization, one-brain-multiple-forms, and continuous evolution." These capabilities can support robots in autonomously completing long-range, complex mobility and manipulation tasks, forming a full-chain technological barrier from underlying technology to upper-layer applications. Currently, the unmanned delivery industry is in a period of rapid development, with mature robot applications already deployed in segments like warehousing and trunk transportation. However, in the "last 100 meters" scenario from regional sorting centers to end-users, robot deployment still faces significant challenges due to factors like complex terrain and human traffic interference. This dexterous hand competition was designed around the practical needs of the final stage of food delivery and intentionally incorporated real-world challenges such as navigating through pedestrian flow and sharing elevators with humans, giving it high application-oriented value. The victory of ROBOSENSE's full-stack robot technology solution demonstrates the potential for innovative, leading hand-eye coordination technology to enable seamless, end-to-end unmanned delivery applications from the starting point to the end-user. As the technology continues to generalize and mature, its application boundaries will extend from instant delivery to broader life service scenarios such as smart elderly care and smart homes, transforming robots into intelligent partners that truly empower daily life and create social value.
Comments