Human Factors Issues in Using Micro-Uninhabited Vehicles in Urban Disasters

Technical advances in remotely operated vehicles (drones) have seen an increase in their use in a variety of work settings, including remote searches and damage assessment. In the case of disaster response and management, uninhabited aerial vehicles (UAVs) could be extremely useful. For example in floods UAVs could be used to search for people trapped on roofs or floating on debris, in bushfires they could be used to improve the view of the fire and provided response managers with improved situational awareness, in marine environments UAVs could be used to search for lost people, and finally, in earthquakes UAVs could be used, if small enough, to search within buildings for damage assessment and searches. In the later case, this has already been attempted in response to the Christchurch February 22nd earthquakes, in which a Parrot A.R quadrocopter drone (commercial off-the-shelf drone) was flown into the Christchurch Cathedral in order to assess for damage. This idea radically reduces the risk to search and damage assessment personnel. However, actually flying the micro-UAV presents a significant challenge in these kinds of environments.
In the present talk, we will discuss recent research at the University of Canterbury on human factors issues of using micro-UAVs in urban disaster scenarios. One of the challenges in flying micro-UAVs is that the pilot has limited sensory information. The pilot is flying with a limited field of view, a 2-d camera image instead of a 3-d naturalistic scene (out of a cockpit window), and the pilot receives no tactile or vestibular cues regarding the UAV’s orientation. One significant challenge in using micro-UAVs, therefore, in urban settings is operating the drone in constrained environments, for example, inside buildings. The turbulence and tight spacing of these environments presents a significant challenge to the pilot. We have been examining the issues involved in using micro-UAVs in these constrained environments. We will discuss training programs for pilots, motor control aspects of the piloting task, the relative costs and benefits of different control interfaces, and assessments of the cognitive workload of the piloting task. In the later research we have added a secondary mental workload to the pilot. This technique is called dual-tasking and it enables a sensitive assessment of the actual mental workload of a primary task. We had pilots engage in a piloting task and a secondary communication-like task. The secondary task represented communication with a co-worker and required the pilot to recall verbally transmitted words presented to the pilot during the flight operation. The pilots navigated and controlled the drone in simulated search operations (flying an obstacle course). We used the secondary task to assess the cognitive load of the flight task. In comparison to other simple motor tasks such as using a joystick to select targets on a computer screen the UAV piloting task is extremely demanding. We will also discuss future planned research.