Weather forecasting startup Tomorrow.io has unveiled the DeepSky satellite constellation, designed to improve atmospheric forecasts by collecting large amounts of data, necessary to feed artificial intelligence models.
The individual satellites of the DeepSky group will be much larger than six-unit cubesats, which make up the initial grouping of the Gen1 company. According to Tomorrow.io's Director of Strategy, DeepSky satellites will be equipped with "devices of a completely different class: very highly effective, co-located sensors". Advances in sensors and fast refresh rates, powered by DeepSky, creates a new class of more adaptive and localized decision-making systems based on artificial intelligence. According to the startup, this evolution will define the future of the world's largest operations.
Artificial intelligence promises to radically improve the accuracy of weather forecasts, however, the models require huge amounts of data. DeepSky was created to provide a large amount of high-precision data, needed by Tomorrow.io to power a new generation of its own weather models. Previously, weather models were often limited by available computing power and physical understanding of the atmosphere, and now the bottleneck has become the data itself.
DeepSky's constellation of satellites is a significant step toward defining a new category of capability, when agentic artificial intelligence becomes a critical part of planning and creating a single operational picture. DeepSky will complement existing satellite systems, including geostationary and low-orbit groups, which feed public weather models, пропонуючи покращене прогнозування швидко змінюваних та екстремальних погодних явищ.
Source: https://spacenews.com