Can Intelligent Agent Improve Human-Machine Team Performance Under Cyberattacks?

Abstract

This paper presents a preliminary work of a simulation study for the evaluation of human-machine team (HMT) performance with assistance of a sentinel system under cyberattacks. Sentinel system is an intelligent agent of cyberattack-detection kit, whose user interface (UI) was designed and implemented as a proof-of-concept in the simulation. The goal for this sentinel system is to improve the pilots’ Situation Awareness (SA) of the system vulnerability and resiliency, such that the HMT performance can be improved against disruptive events. Based on the literature of cyberattacks on unmanned aerial vehicle (UAV) control systems, realistic mission operation and cyberattack scenarios were identified and implemented on a simulated UAV ground control station (GCS). In the follow-up experiment, up to twenty-four Air Force pilots will be instructed to supervise the UAVs under specific mission scenarios, and to respond for system recovery solutions generated on Sentinel UI after detection of cyberattacks. Understanding the interactive behaviors of the pilot and Sentinel under mission contexts and cyberattacks is expected to help improve HMT performance.