Using Artificial Intelligence Imaging Systems to Detect Signs of Genocide from Low Earth Orbit
Faculty Mentor Name
Davide Conte, Richard Magnum, Erika Podest
Format Preference
Poster
Abstract
Genocide is an under-reported issue across the world, partly due to the difficulty regarding reporting and collecting evidence. Reconnaissance and Documentation (RAD) proposes deploying a satellite, Leza, in low- Earth orbit (LEO) that will capture images of countries on the United Nations (UN) genocide watchlist to look for signs of genocide remotely. RAD will utilize an Artificial Intelligence Image-Recognition Software (AIIS, pronounced “eyes”) to determine if the observed countries display signs of past, ongoing, or future genocide. Once AIIS is confident about an observed region displaying signs of genocide, the images will be reviewed to ensure AIIS is working as intended and distribute any images of potential genocide out to media. The primary mission objectives for RAD 2.1 were to develop AIIS along with an orbit to support Leza’s launch as a secondary payload. Additional mission objectives included the capability to continue training AIIS after Leza’s launch. RAD 2.1 split AIIS into two components AIIS-1 and AIIS-2 to facilitate a realistic execution of this project. AIIS-1 is intended to filter out images that do not contain signs of human activity while AIIS-2 will perform the analysis on images filtered by AIIS-1 to determine whether the images display signs of genocide. By developing and testing the technologies and concepts needed to develop and launch Leza, RAD 2.1 enabled students to gain research and development experience for making a remote-sensing satellite. Overall, RAD aims to promote global awareness of genocide propagating across the world throughout its development and project execution stages.
Using Artificial Intelligence Imaging Systems to Detect Signs of Genocide from Low Earth Orbit
Genocide is an under-reported issue across the world, partly due to the difficulty regarding reporting and collecting evidence. Reconnaissance and Documentation (RAD) proposes deploying a satellite, Leza, in low- Earth orbit (LEO) that will capture images of countries on the United Nations (UN) genocide watchlist to look for signs of genocide remotely. RAD will utilize an Artificial Intelligence Image-Recognition Software (AIIS, pronounced “eyes”) to determine if the observed countries display signs of past, ongoing, or future genocide. Once AIIS is confident about an observed region displaying signs of genocide, the images will be reviewed to ensure AIIS is working as intended and distribute any images of potential genocide out to media. The primary mission objectives for RAD 2.1 were to develop AIIS along with an orbit to support Leza’s launch as a secondary payload. Additional mission objectives included the capability to continue training AIIS after Leza’s launch. RAD 2.1 split AIIS into two components AIIS-1 and AIIS-2 to facilitate a realistic execution of this project. AIIS-1 is intended to filter out images that do not contain signs of human activity while AIIS-2 will perform the analysis on images filtered by AIIS-1 to determine whether the images display signs of genocide. By developing and testing the technologies and concepts needed to develop and launch Leza, RAD 2.1 enabled students to gain research and development experience for making a remote-sensing satellite. Overall, RAD aims to promote global awareness of genocide propagating across the world throughout its development and project execution stages.