Recently, Shenzhen Gesong Technology Co., Ltd. (Gesong Tech) secured the championship at the 2025 Shenzhen Intelligent Robot Dexterous Hand Competition (Dexterous Hand Competition), leveraging ROBOSENSE's (02498) advanced dexterous hand technology, integrated sensor system Active Camera, and VLA model. This victory underscores ROBOSENSE's industry-leading full-stack hand-eye coordination capabilities and its ability to efficiently implement solutions in real-world scenarios while creating industrial value with partners.
The Dexterous Hand Competition is China's first professional event focused on robotic end-effector technology. Centered on real-time delivery scenarios, the competition tested robots in complex tasks such as parcel retrieval from lockers, food package unboxing, folding and recycling packaging, and building deliveries—evaluating precision and stability in dynamic environments. The event featured dual tracks, attracting top universities, research institutions, and enterprises from Shenzhen, Hong Kong, Hangzhou, and beyond.
With ROBOSENSE's dexterous hand, Active Camera vision system, and VLA model, Gesong Tech outperformed 28 competing teams to claim the title. Gesong Tech, an emerging innovator in robotics, benefited from ROBOSENSE's full-stack AI robotics technology, which integrates perception, decision-making, and control into a seamless loop, ensuring accuracy and fluidity in navigation, mapping, and hand-eye coordination.
During the competition, despite a narrow 7cm recycling slot and wind-induced package movement, ROBOSENSE's dexterous hand successfully folded and retrieved the food container, demonstrating exceptional coordination. Empowered by ROBOSENSE's AI robotics, Gesong's robot autonomously completed the entire workflow—from unboxing and folding to navigation, elevator use, and final delivery—with high precision, earning the championship.
Gesong Tech's win highlights ROBOSENSE's core competencies in multimodal data collection, edge deployment, small-sample generalization, multi-model control, and continuous evolution, enabling robots to execute long-range, complex tasks and establishing a full-stack technological barrier from foundational to application layers.
While unmanned delivery is rapidly advancing in warehousing and logistics, the "last 100 meters" from distribution centers to end-users remains challenging due to complex terrains and human interference. The competition simulated real-world hurdles like pedestrian flow and shared elevator use, emphasizing practical applicability. ROBOSENSE's victory demonstrates the potential of its hand-eye coordination technology to enable end-to-end unmanned delivery solutions.
As the technology matures, its applications will expand beyond delivery into elderly care, smart homes, and other service scenarios, transforming robots into intelligent partners that enhance daily life. ROBOSENSE will continue driving AI robotics from technical validation to commercialization, collaborating with global partners to build a thriving physical AI ecosystem.
Comments