On November 11, 2023, at 10:47 a.m. local time, a Polish observation satellite – Intuition–1 – was launched into space from the United States Space Force base in Vandenberg, California. In the process of transport into orbit around the Earth, a two–stage Falcon–9 carrier rocket was involved¹ , owned by SpaceX. The whole endeavor was organized as part of the Transporter–9 space mission.²

Intuition–1 is a device classified as an artificial remote sensing satellite, i.e. one capable of performing imaging observations and reconnaissance using adequate remote sensing techniques and specialized equipment, most often hyperspectral.³ It was constructed by engineers of the aerospace company KP Labs, based in Gliwice. The satellite was launched into space with dedicated hardware, namely the Antelope on–board computer with high computing power. By performing calculations and data analysis while still aboard the satellite in space, it will significantly reduce incurring operational costs. In practice, this means that huge amounts of collected data will not go to ground centers, but only the conclusions drawn from them.⁴

The satellite is designed to support the development of numerous fields, but it is to be used first for agriculture, environmental research and deforestation analysis. Accordingly, Intuition–1 has been retrofitted with a powerful data processing unit – the Leopard DPU – which will make it possible to interpret the current state of plants and forests, create soil maps, forecast crop yields, plan reforestation, and monitor environmental changes through the development of maps of air, water, soil pollution, among other things.⁵

The goal of the satellite mission involving the Intuition–1 satellite is to conduct coordinated observation of the Earth using a hyperspectral instrument and an onboard computing unit capable of processing data using neural networks (artificial intelligence) deployed in orbit. In addition, Intuition–1 has the computing power to segment the acquired hyperspectral images in orbit. Segmentation means the ability to automatically determine the features of the images, i.e. to find similar patterns. Such patterns could be, for example, crop diseases or climatic anomalies such as droughts.⁶ The size of the transmitted data has been reduced to the necessary minimum (for recording imaging results of a 40 km x 40 km area with a spatial resolution of 25 m/pixel for spectral measurements using 150 bands/wavelength ranges, about 7 GB of data). After processing, it is possible to reduce the data by more than 100 times to retain only relevant information.⁷ The hyperspectral camera will use up to 192 spectral bands ranging from 465 nm to 940 nm. Each scene captured by the camera will contain an image divided into multiple frames, while each frame will capture a different spectral band. This means that in order to obtain a hyperspectral image of a specific area, the frames from all spectral bands will have to be assembled and processed by a data processing unit – the Leopard DPU – which will also store the images in non-volatile memory, the so–called ‘cloud’.⁸

The Intuition–1 satellite is a technology demonstrator that is expected to prove that the use of artificial intelligence to process hyperspectral data already in orbit positively affects the efficiency of the remote sensing process, and hyperspectral instruments based on optics can observe phenomena previously impossible to detect by standard panchromatic, multispectral or radar satellites. The satellite has an intrinsic mass of 12 kg and dimensions of 30x20x10 cm and belongs to the group of microsatellites.

 

Sources: