Kudan provides localization and mapping software solutions leveraging visual cameras, LiDAR and other sensor data (e.g. GPS) as input. Kudan’s technology is based on SLAM (Simultaneous Localization And Mapping). Kudan typically works as a value-added software provider, supporting customers who integrate Kudan’s software into their solution/system. Kudan is public in Japan and has ~600 Million USD valuation as of January 2020.
Improve productivity or lower sensor & processor cost through more robust and accurate localization & mapping by tightly coupled visual+lidar SLAM. Expand use-case scope by significantly compressed lidar map for localization (>300x lighter than lidar raw data), zero-lag localization with IMU, no accelerator required (or accelerate with them), no pre-/ post-processing required.
USE CASE/CUSTOMER PROFILE
Autonomous driving: working with one of the global top5 OEM for an autonomous driving project in Europe including both highways and city streets. We offer localization & mapping software. HD map: working with a Japanese leading mapping company for developing HD map with Kudan’s software Robots: working with global top5 AGV (auto guided vehicle) manufacturer for improving its autonomous driving capabilities Surveying/ mapping: working with mapping companies for mobile mapping solutions utilizing our software on top of GNSS for more robust localization.
Lidar stand-alone SLAM:
Kudan lidar SLAM leverages our deep expertise in visual SLAM and has a distinctive performance. Still lidar localization and mapping on city streets and handling large scale lidar-based maps are challenging topics for many use-cases and Kudan lidar SLAM provides solutions to them.
Tightly-coupled visual+lidar SLAM:
“Sensor fusion” is something many companies do. However, multiple approaches exist and their resulting performance totally varies. Kudan offers Tightly-coupled visual+lidar SLAM while others use the “loose coupling” of sensors. Tight coupling requires a more complicated and sophisticated way to deal with multiple sensor data in one SLAM system. On the other hand, loose coupling is easy and it can be done by applying filters without deep expertise in SLAM. This brings better performance as a whole system. Therefore, this improves the safety and accuracy of the system and you can have the same performance with more affordable sensor setups