Miratlas’ Machine Learning R&D project “LASER” on atmospheric data receives “Deep Tech” label from Bpifrance
“LASER” consists in the deployment of 50 units of Miratlas Sky Monitor stations around the world, gathering a unique and proprietary atmospheric database of all parameters impacting light propagation through the atmosphere such as cloud cover, absorption, and turbulence, night and day for several years. This data will then be processed by Machine Learning allowing operational short term forecasting.
Miratlas helps operators to deploy terrestrial Optical Ground Stations supporting Direct To Earth Laser communication for future satellite constellations in the most cost-effective way by maximizing the optical communication link availability and throughput thought site diversity and expert knowledge laser channel atmospheric propagation.
“LASER” will allow Miratlas to accelerate its development in providing strategic data required by Satcom and Telecom operators to ensure the deployment and operations of their network.
Jean-Edouard COMMUNAL, co-founder and CEO of Miratlas said: “The prestigious Deeptech label granted by Bpifrance acknowledges years of development on our Sky Monitor which we are now ready to produce and deploy in volume and complete our pivot from hardware supplier to atmospheric Data as a Service provider. Bpifrance had already helped us to lift key milestones, preparing our instrument and database for large scale Machine Learning. We are grateful to them for their continuing support. “LASER” is a structuring project for Miratlas which will significantly leverage our revenues and the structuring of our medium/long term sales through the launch of a unique “Forecasting as a Service” offer.”
The investment in the deployment of our own network of stations will allow Miratlas to respond to market demand by increasing the number and diversity of our geographical coverage. Miratlas’ data is unparalleled in its specificity and density. Each station delivering 11 million atmospheric data per year over several years makes possible the implementation of powerful machine learning algorithms.