Kudan has successfully completed a field trial in partnership with NTT Infrastructure Network Corporation (NTT InfraNet), showcasing an innovative approach to generating high-precision 3D maps in areas where dense buildings block GNSS signals. Kudan’s advanced simultaneous localization and mapping (SLAM) technology was combined with NTT InfraNet’s infrastructure data, demonstrating a simple and highly efficient method for mapping urban canyon environments: spaces where traditional satellite-based positioning fails. This development marks a significant step forwards in the creation of smart cities, offering a practical solution for urban infrastructure management and spatial data collection in challenging environments.
High-precision point cloud maps are essential for the realization of smart cities, enabling the safe and accurate operation of autonomous mobile robots and self-driving vehicles. If map accuracy is insufficient, errors in vehicle localization and navigation can increase, significantly impacting safety and operational efficiency.
Ideally, GNSS data would be used as correction data to create high-precision point cloud maps. However, in urban environments, the ‘multipath problem’ occurs, where signal reflections from high-rise buildings cause inaccuracies in GNSS positioning. This makes it difficult to use high-precision GNSS data as correction data for map creation. As a result, the implementation of measurement systems equipped with expensive sensors and the integration of advanced technologies for complex data acquisition and processing have posed significant technical and cost barriers.
To address these challenges, there is a growing demand for methods that enable the efficient generation of high-precision 3D point cloud maps in urban areas by utilizing simpler and more cost-effective sensor configurations and positioning data, while minimizing complex setup and data processing. Through such initiatives, the advancement of digital infrastructure is expected to accelerate, contributing to the widespread adoption of smart cities while reducing associated costs.
Kudan and NTT InfraNet have been working together to establish high-precision 3D mapping technology. As part of this initiative, Kudan conducted a proof-of-concept experiment to generate high-precision 3D maps in the densely built-up Konan area of Shinagawa, Tokyo. In this experiment, Kudan utilized its proprietary artificial perception technology, Kudan 3D-Lidar SLAM (KdLidar), and applied map correction techniques leveraging high-precision 3D spatial information (manhole location data) provided by NTT InfraNet. The results demonstrated the ability to efficiently generate 3D point cloud maps with a horizontal positional standard deviation within 12cm and an elevation point standard deviation within 25cm without relying on GNSS, even in urban canyon environments.
Kudan’s latest mapping technology is opening up new possibilities for smart city development. In a recent trial, the company showed how high-precision 3D maps can now be created even in dense urban areas where GNSS signals typically fail. This advancement paves the way for better traffic management, streamlined infrastructure maintenance and stronger disaster response. It also supports greener urban planning, with potential to cut emissions and boost carbon neutrality efforts.
Looking ahead, Kudan plans to scale up testing and move toward commercial rollout, with an eye on powering next-gen robotics and digital twin solutions across industries.
