Examination of Human Factors in Networked Sensors in Live and Virtual Environments

摘要

A demonstration was held with Soldiers operating robotic networked sensors in an actual mock-up of a reconnaissance and surveillance vehicle (HMMWV high mobility, multipurpose wheeled vehicle)) in the field and in virtual reality simulators. The crew in the live re-supply vehicle (RSV) and the two crews in the virtual simulators consisted of three Soldiers each. The allocation of personnel in both environments was the same: (a) one Soldier served as the vehicle commander and operated the unmanned ground sensors (UGS); (b) one Soldier operated unmanned aerial vehicles (UAVs), and (c) one Soldier operated an unmanned ground vehicle (UGV). Data from surveys and interviews revealed insights into the interface design, sensor and software capabilities, workload, and skills and abilities required of robotic operators. The interface (maneuver command and control MC2)) was judged as very good for mission planning, but re-tasking sensors to different routes was difficult and time consuming, which resulted in the inability to identify previously detected targets. The imagery on the infrared (IR) UAV sensor, unlike the daylight camera, was not good enough for target identification. The imagery and interface in the virtual simulators was not as those on the live vehicle, resulting in higher workload and stress in the virtual simulators. Situational awareness was equal but low in both the live and virtual environments, probably because of technical difficulties, especially in the live RSV. Skills needed in both environments tended to be those necessary to receive information from and send information to higher (communication), identify and solve problems concerning sensor coverage (conceptual), and recognize targets quickly (speed loaded). Visual skills were also needed to interpret sensor feeds and track the sensor on the MC2, especially on the virtual simulators.