RollingDot

RollingDot, founded in Toulouse in 2024, is a Deeptech startup leveraging disruptive technology stemming from French scientific research. RollingDot offers software initially developed over several years within IRIT (Toulouse Institute of Computer Science Research) and the CNRS (French National Centre for Scientific Research).

The company is notably supported by Toulouse Tech Transfer, which is also a shareholder, as well as by BPI as part of the France 2030 program. The founding team consists of two entrepreneurs from the publishing and software development sector, 3 world-renowned research professors (who initiated the project at the University of Toulouse and the CNRS) and complemented by a team of engineers in Machine Learning and 3D geometry.

RollingDot offers a software solution that combines frugal artificial intelligence, geometry and 3D visualization to automate the very large-scale processing of 3D data (LiDAR, Photoneo and photogrammetry).

Our solution accelerates the creation and use of digital twins: 3D modeling of infrastructure, territories and industrial parts in the robotics industry 4.0.

RollingDot's innovation stems from scientific research (IRIT and CNRS) on the processing and visualization of point clouds (LiDAR or photogrammetry data). By combining advanced 3D geometry processing approaches with frugal artificial intelligence, RollingDot deploys automated processing pipelines to extract information relevant to specific use cases: – local geometric properties: location of curved areas, edges, contours, etc. – structures and elements: pylons, cables, roofs, pipes, ridges, ground, vegetation, etc. This automated processing structures the data to build and update digital twins. Unlike traditional manual and semi-automated approaches, our processing is performed using minimal training data and consuming few resources (time and energy): RollingDot's software solution positions it as a significant player in the frugal processing of point clouds. RollingDot's innovation stems from scientific research (IRIT and CNRS) on the processing and visualization of point clouds (LiDAR or photogrammetry data). By combining advanced 3D geometry processing approaches with frugal artificial intelligence, RollingDot deploys automated processing pipelines to extract information relevant to the use case: – local geometric properties: location of curved areas, edges, contours, etc. – structures and elements: pylons, cables, roofs, pipes, ridges, ground, vegetation, etc. This automated processing structures the data to build and update digital twins. Unlike traditional manual and semi-automated approaches, our processing is carried out using little training data and consuming few resources (in time and energy): RollingDot thus positions itself with its software solution as a significant player for the frugal processing of point clouds.