Adopting 3D vision + 3D laser hybrid sensors based on natural features of the environment, rich in information and high accuracy as the core sensors for navigation, it achieves highly robust and high-precision positioning in a variety of scenarios, indoor and outdoor, all-weather and highly dynamic, without the need to modify the site environment and existing business processes.
Intelligent visual perception technology. Using the perception scheme of 3D laser vision, the perception distance, accuracy and space are greatly improved, and the environment semantic map (object-based SLAM) is generated. The learning of fork targets, obstacles, and the observation of the pick-and-place space can judge the status of the object in real time. , which provides data reference for adaptive operation and autonomous obstacle avoidance.
Based on vision perception technology, the unmanned forklift can perform Max 6 degrees of freedom (6 DOF) movement optimization in real time to achieve smooth multi-axis combined movement, ensuring accurate and efficient execution of fork end operations such as loading and unloading, pick and place, stacking, etc.
Large-scale multi-level scheduling technology. Taking into account the communication delay and motion uncertainty of unmanned forklifts, the scheduling algorithm and multi-vehicle interaction strategy are optimized to realize long time, incremental and cluster scheduling of unmanned forklifts and related logistics robots, and the system has a single cluster scheduling capacity of several hundred