Artificial Intelligence is increasingly relevant in space-related projects. In these projects, the solution is as important as the quality of the data. The current potential of the automatic learning solutions make it possible to use it to complement or even replace in some cases the classic techniques for resolving tasks such as processing signals or detecting anomalies. In addition, the more data is available, the better the performance, so it makes sense that different entities collaborate on a common solution. However, this can cause a problem in terms of privacy and it is not always possible to share the data among the different parties.

To tackle this problem, GMV Data Scientist Juan Miguel Auñón presented uTile PET at the “Big Data from Space 2021 (BIDS21)” event, a solution for the collaborative development of Artificial Intelligence algorithms without having to compromise the privacy of each of the parties. During his presentation, he also offered the example of secure k-means, a clustering algorithm used by organizations to collaborate for a common goods, safeguarding privacy at all times.

uTile PET is a solution developed by GMV that leverages confidential and private data to improve the automatic learning algorithms and analytical models, always in compliance with the requirements of the organization, guaranteeing data privacy and the applicable legislation. With this technology, it is not necessary to choose between data privacy and usability, as it uses advanced cryptographic methods that keep the data encrypted while all the necessary calculations are completed. Thus, uTile PET makes it possible to keep the organization’s sensitive data from being exposed or transferred between departments, organizations or different countries.