PROACTIVE AUTOMATIC IMAGERY INTELLIGENCE POWERED BY ARTIFICIAL INTELLIGENCE EXPLOITING EUROPEAN SPACE ASSETS.

ended

Imagery Intelligence (IMINT), and Earth Observation (EO) IMINT in particular, is an essential source of intelligence for the defence and security community. Traditionally, IMINT workflows have been designed to deal with scarcity of data due to the high cost of imagery, process capacity and IMINT analysts. IntSen2 aims at researching a disruptive change, using Artificial Intelligence to develop a concept of application for an IMINT service, from a reactive service design to a proactive service design. And it does so by basing the research and proposed solutions exclusively on European assets in order to support EU operational autonomy. This proposal, therefore, will design a system to fully automatize the IMINT workflow, consisting of: i) Automatic Acquisition, ii) Automatic Pre-processing; iii) (Key Challenge) Automatic Extraction, iv) Automatic report generation; v) Automatic Dashboard management. Two technologies using Artificial Intelligence lie at the core of the solution that IntSen2 proposes: - Super-resolution: Improving the resolution of imagery by a factor of 4 allows imagery from Sentinel satellites to provide value in security scenarios. - Automatic Extraction: Applying AI based extraction techniques to super-resolved Sentinel imagery can unlock the use of all operational potential of Sentinel for security scenarios. Under this context, IntSen2 appears as the first urgent step for creating the infrastructures that the EC needs for obtaining an autonomous and independent intelligence that can face the rise of capabilities of major third countries and contribute to a Common Information Superiority of the Member States. To do so, IntSen2 has assembled a european multi-stakeholder consortium, built to embrace and exceed the requirements for crossborder cooperation. IntSen2 mustered a European consortium of 11 entities from 7 different countries and engage the support of part of the countries’ MoDs present in the project.