Gravitational Replicating Autonomous Space Hopper
Faculty Mentor Name
Davide Conte, Richard Magnum
Format Preference
Poster
Abstract
Gravitational Replicating Autonomous Space Hopper (GRASHopper)’s purpose is to autonomously identify suitable landing areas, map terrain, and stabilize itself during the decent using Artificial Intelligence (AI). Completing these tasks will contribute to safe and consistent asteroid landings without human intervention. GRASHopper’s purpose aligns with the RASC-AL prompt and aims to advance the possibilities of deep space exploration through the integration of AI. By implementing components like a depth camera, GRASHopper will be able to choose an appropriate landing site with material necessary in the greater self replication process. The chosen landing site must also have minimal debris and low peaks and valley to ensure a stable landing position. Determining and controlling the attitude of the spacecraft with an IMU will allow for a soft landing and keep the spacecraft intact for subsequent landings. GRASHopper’s incorporation of the Generalizable Episodic Memory (GEM) AI Model serves to improve the overall accuracy and speed of each landing using thousands of computer simulated tests and physical tests on rocky terrain on Earth. GRASHopper’s physical testing method uses an overhead three-axis motorized structure which will communicate with the onboard computer to land in Prescott, AZ and mimic landing on an asteroid.
Gravitational Replicating Autonomous Space Hopper
Gravitational Replicating Autonomous Space Hopper (GRASHopper)’s purpose is to autonomously identify suitable landing areas, map terrain, and stabilize itself during the decent using Artificial Intelligence (AI). Completing these tasks will contribute to safe and consistent asteroid landings without human intervention. GRASHopper’s purpose aligns with the RASC-AL prompt and aims to advance the possibilities of deep space exploration through the integration of AI. By implementing components like a depth camera, GRASHopper will be able to choose an appropriate landing site with material necessary in the greater self replication process. The chosen landing site must also have minimal debris and low peaks and valley to ensure a stable landing position. Determining and controlling the attitude of the spacecraft with an IMU will allow for a soft landing and keep the spacecraft intact for subsequent landings. GRASHopper’s incorporation of the Generalizable Episodic Memory (GEM) AI Model serves to improve the overall accuracy and speed of each landing using thousands of computer simulated tests and physical tests on rocky terrain on Earth. GRASHopper’s physical testing method uses an overhead three-axis motorized structure which will communicate with the onboard computer to land in Prescott, AZ and mimic landing on an asteroid.