Geek+ and Intel have launched the Vision Only Robot Solution, using Intel’s Visual Navigation Modules. This robotics solution, designed for the logistics sector integrates Intel’s RealSense camera in autonomous mobile robots (AMRs). The system has several features, including depth-based visual perception, navigation, and obstacle avoidance.
The Intel RealSense camera, central to the Vision Only Robot Solution, operates independently of specific platforms. It conducts depth calculations within the device itself. The module, equipped with the Intel Core i7-1270P processor, also powers the Robotic Vision Hub, which supports algorithms, such as V-SLAM (Visual Simultaneous Localization and Mapping), composite detection networks, and robot following.
This enables mapping, positioning, and obstacle detection in dynamic settings. V-SLAM leverages multi-sensor fusion to generate various map types, facilitating reliable navigation. The composite detection network reduces detection errors by combining multiple data validation approaches, and the robot following feature enables efficient human-robot interaction for tasks such as warehouse guidance and transportation. This interaction is controlled by RoboGo, Geek+’s proprietary system, which coordinates local planning and obstacle avoidance for smooth navigation.