ATHENA - Autonomous Decision Making and Control of Autonomous Systems using AI Methods
Abstract
This proposal considers the design of algorithms for a multitude of autonomous agents to execute different kind of tasks in a specific area (scenario-operation area). For this purpose, a variety of state-of-the-art techniques is going to be examined, including Machine Learning, Game Theory, Evolutionary Algorithms allowing the agents to perform dynamically changing tasks. The main technological challenge is to develop a cooperative system which can keep up with the needs of invader chasing applications, forming specific coalitions, area monitoring even rearranging formations using fundamental algorithms with the implementation of advanced decision-making theory. Therefore, it consists of a dynamic decision-making approach for autonomous agent-based systems so that they will complete the required tasks, recover safely from faults and emergencies, and respond predictably to instructions at all time. Especially, this issue becomes more significant as the size of the unmanned autonomous agent based systems increases. This project will use and compare various artificial intelligence methodologies, focusing on game theory, machine learning and evolutionary algorithms which are to show a strong promise, for specific UAV autonomous agents and space (satellite controls-formation flying) applications.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Feb 22, 2024
- Source ID
- FA86552317218
Entities
People
- Vaios Lappas
Organizations
- Air Force Office of Scientific Research
- National and Kapodistrian University of Athens
- United States Air Force