Live Updates > Live Update Details
2026-07-18 20:26:13
At the World Artificial Intelligence Conference held from July 17th to 20th, 2026, Qianjue Robotics officially unveiled its first VTLA embodied haptic model. At the exhibition, Qianjue Robotics not only showcased this vision-touch-language-motion integrated embodied haptic model, but also exhibited a series of related achievements, including a multimodal visual-touch dataset, core data acquisition equipment, and a haptic sensor hardware matrix. In a media interview, Ma Daolin, founder of Qianjue Robotics, explained that unlike traditional single-vision VLA models, the VTLA model innovatively integrates visual and tactile dual-modal perception information, endowing the robot with dual capabilities of visual recognition and tactile perception. This allows it to capture physical attributes of objects that are difficult to obtain with a single visual device, significantly improving both the dimensionality of recognition and the accuracy of detection. The application of this type of embodied intelligence technology is expected to provide more reliable perception support for various scenarios such as industrial precision operations, home service robots, and medical assistance, further enhancing the robot's task completion capabilities in complex environments.
