Kudan Inc. announced that Kudan and ADLINK Technologies Inc. launches AMR Visual SLAM Solution for autonomous mobile robots (AMRs) and auto guided vehicles (AGVs) to help OEMs accelerate developing AMRs more reliable and robust in broader applications. The initial version consists of ADLINK's RQP-T37 and Kudan Visual SLAM software optimized for Intel XPU on ROS2 platform. It is also planned to release the version which supports ROSCube-X that is based on NVIDIA Jetson platform. The AMR Visual SLAM Solution is a "Solution Ready Platform (SRP)" using the well-validated ADLINK's RQP-T37 controller and Kudan Visual SLAM (KdVisual), which is compatible with ROS and ROS2. This ensures prompt integration of KdVisual into the customer's overall architecture with boosted performance, especially in challenging environments of AMR applications such as dynamic objects, scenery changes, and
indoor/outdoor mixed applications. This is a perfect platform for the development of industrial use robotic applications such as autonomous mobile robots (AMR) and au onomous mobile industrial robots (AMIR). The most distinctive benefit of this offering is the ease of integrating Visual SLAM into the AMR system and high-performing AMR navigation. ADLINK offers its Neuron SDK together with ROSCube, which contains various software modules required in AMR development and Kudan Visual SLAM package also offers several supporting software to enable seamless integration, especially for those who use 2D-Lidar SLAM for their localization approach. 3D Visual SLAM is quite different from 2D-Lidar SLAM in terms of mechanism and characteristics. Therefore, many companies cannot see its proper performance before they give up using it. Kudan has very extensive knowledge on how to deploy visual SLAM properly
through numerous projects. This Solution Ready Platform includes not only Kudan's Visual SLAM software but also other supporting documents and modules to accelerate integration and prototyping. Have some examples of developing an initial prototype of AMRs with visual SLAM package in 2 weeks. Another challenge robotics OEMs face when they try to adapt visual SLAM is that it's quite challenging to have reliable performance by building from scratch or based on open-source software. The following points are some of Kudan Visual SLAM's distinctive advantages, which result in a productivity increase of robots, and application expansion. 5x - 10x faster process speed, 25%, 50% memory usage, Average <1cm repeatability/repeatable accuracy, Better absolute accuracy than 2D-Lidar SLAM and Robust against 95% scenery changes.