Unmanned Air Vehicles: a New Age in Human Factors Evaluations

John J. Spravka, Edwards AFB, CA, USA
Deborah A. Moisio, Edwards AFB, CA, USA
Mary G. Payton, Edwards AFB, CA, USA

Abstract

As the role of the aircraft pilot transitioned from a nearly total manual controller in early manned aircraft to one of supervisory control and/or cooperative functioning in “unmanned” aircraft, the human factors flight test approach and the associated test methodologies have necessarily changed. Piloting air vehicles evolved from using cockpit instruments and manual controls to fly the aircraft, to monitoring the cockpit instruments which fly the aircraft nearly automatically, to using ground station instruments to fly the aircraft remotely. While most, if not all, of the physical stressors of the cockpit are absent from the typical ground control station, many of the cockpit stimuli that provide invaluable aircraft health and status information are also absent. Increased levels of automation have induced new types of failures. These include failure to monitor, vigilance decrement, over reliance on standard values, automation-induced complacency, and increased latency in detecting problems. Consequently, these failures often lead to reduced operator performance due to information shortfall.

As the pilot-aircraft relationship evolve, the focus of human factors evaluations moves from what the pilot physically perceives and processes in the cockpit, to what the pilot mentally perceives and processes on the ground. Physical information from the aircraft, such as vibration and sound cues, must be transformed into usable information on the ground station displays. This includes keeping the level of automation appropriate so the pilot on the ground can be aware of and adequately handle emergency situations. Since pilot workload for an Unmanned Air Vehicle (UAV) is mostly mental, maintaining situation awareness is paramount.

This paper identifies some critical components of new human factors approaches for evaluating UAV humansystem interfaces and compares them with approaches traditionally used to evaluate manned air vehicles.