Applying AI to a new challenging problem

BUBBLES project aims to formulate a new U-Space Separation Management service to define ‘safety bubbles’ around the unmanned air systems (UAS) operating in some area. Procedures are developed to prevent collision and manage bubbles’ overlap. 

30 Oct 2020

Within this Exploratory Research project, Sapienza University is exploiting its know-how in several fields of artificial intelligence (AI) to contribute in defining and managing the separation minima and separation methods between UAS, guaranteeing the suitable level of safety for operations.

The methodology integrates several AI techniques, including machine learning and multi-agent systems, focusing in particular on distributed learning and distributed reasoning. AI and UAS are dynamic fields of high interest for the research community, because they are very promising technologies to improve human activities. BUBBLES project will indeed investigate suitable use of AI techniques to improve safety and efficiency of UAS performing useful tasks in public environments.

As a matter of fact, AI techniques, such as Machine Learning and Multi-Agent Systems, have been successfully used in many applications domains to support decisions of expert human operators, ranging from medical diagnosis to financial and commercial services.

In BUBBLES, such technologies will be applied to support decisions within an Air Traffic Management system that will manage the operations of thousands of UAS above our cities. New algorithms and new models based on AI will enable the computation of crucial parameters for a safe and effective management of UAS operations.

Involvement in BUBBLES project will bring several benefits to Sapienza University of Rome, Italy, including the possibility of involving young researchers and students in a very interesting research and application field, thus giving them enormous opportunities for their future careers. Moreover, from an AI perspective, it is very interesting to evaluate the effectiveness of these techniques to solve problems arising from UAS management, thus enlarging its scope.