Human Performance in Highly Automated, Cyber Vulnerable Unmanned Aerial Systems: Effects of Operators’ Background and Task Load

Abstract

This article evaluated the effects of operator background and task load on operator ability to detect and respond to cyber events while controlling multiple unmanned aerial systems (UAS). Cyber physical systems (CPS) are susceptible to cyber threats and associated events via connections critical to their operations. The effects of cyber intrusions on UAS operator performance as a function of background and task load have been understudied. Fifty-one operators with different backgrounds (Aviators, Gamers and Non-Gamers; n = 17 per group) were evaluated across two levels of task load (in-the-loop (ITL) and out-of-the-loop (OTL)) while controlling a simulation of multiple UAS involving cyber events. Cyber events represented spoofed information that could not be corroborated with available secondary information/displays. Multivariate analysis of variance was used to analyze the effect of background and task load on operator performance, subjective perception of workload (assessed using the NASA Task Load Index), and trust. Nongamers responded to cyber events 62.8% and 67.7% slower than aviators and gamers, respectively. Mean response times across task load conditions were substantially longer for cyber events (128.9 s) than for noncyber, “real” events (42.8 s). Scenario type (ITL/OTL) was statistically significant for workload, cyber classification, and cyber action response. The findings may inform recruitment criteria for future UAS operators. Individuals with gaming experience may be as adept at responding to cyber events while operating multiple UAS as aviators. Cyber events are likely to affect the performance of the overall CPS due to their impact on the human operator.