Enhancing Decision Speed and Resilience of Multi-Agent SituationalAwareness and Inference through Physics-informed AI

Abstract

In networked multi-node Cyber-Physical Systems (CPS) and Dynamic Data Driven Applications Systems (DDDAS), each node can carry out computation or communication responsibilities. We consider a collaborative setup where system nodes pursue an orchestrated effort to execute system-level tasks (such as situational awareness). These setups abound in practice, e.g., autonomous vehicle networks, Unmanned Aerial Vehicles (UAVs), clusters of satellites, and sensor networks monitoring a power system. As proposed by the Air Force researchers, we assume that the intelligence and autonomy of each node depends on its control diffusion, information fusion, and operator infusion agents. The network performance inherently depends on iterative cooperation enabled by nodal computations and communications. The computation and communication costs often do not support the desired decision speed and resilience and hinder real-world implementation. To address this gap, our proposal focuses on improving system-level performance by leveraging node-level intelligence and autonomy and poses three fundamental questions to increase the decision speed and resilience- 1. How can we adjust node-level information fusion rates to speed up system-level decision-making (convergence rate) and improve resilience (though critical information sharing). 2. How can nodes optimize starting points for their update functions. and how can nodes leverage the operator s input to deal with data access issues.3. How can the underlying systems physical laws be leveraged to speed up the convergence process. and how can a node predict neighbors behavior to fill data gaps. Hence integrating Artificial Intelligence (AI) with multi-node information processing paradigms can play a critical role in answering these questions, in turn increasing the decision speed and enhancing resilience in contested setups.

Document Details

Document Type
DoD Grant Award
Publication Date
Feb 06, 2025
Source ID
FA95502410099

Entities

People

  • Javad Mohammadi

Organizations

  • Air Force Office of Scientific Research
  • United States Air Force
  • University of Texas at Austin

Tags

Fields of Study

  • Computer science

Readers

  • Computer Networking
  • Distributed Systems and Data Platform Development
  • Systems Analysis and Design

Technology Areas

  • AI & ML
  • AI & ML - Autonomous Systems
  • AI & ML - DoD AI Strategy
  • Autonomy
  • Autonomy - Autonomous System Control
  • Autonomy - Human-Robot Interaction
  • Cyber
  • Space