Search for: All records

Unmanned systems are becoming increasingly engaged in disaster response. Human error in these applications can have severe consequences and emergency managers appear reluctant to adopt robots. This paper presents a taxonomy of normal and off-normal scenarios that, when combined with a model of impacts on cognitive and attentional resources, specify sources of human error in field robotics. In an emergency, a human is under time and consequences pressure, regardless of whether the mission is routine or whether the event requires a change in the robot, the mission, the robot’s work envelope, the interaction of the humans engaged with the robot, or their work envelope. For example, at Hurricane Michael, unmanned aerial systems were used for standard visual survey missions with minor human errors but the same systems were used at the Kilauea volcanic eruption for novel missions with more notable human errors. An examination of two cases studies suggests the physiological and psychological effects of an emergency may be the primary source of human error. 

A team from the Center for Robot-Assisted Search and Rescue flew 28 day and 16 night sorties with four models of small unmanned aerial systems (sUAS) rotorcraft from eight locations to assist with the tactical response to the 2018 Kilauea Volcano Lower East Rift Zone event. The sorties had significant flight outcomes for responders, added value to the responders and public, and add use cases to the use of sUAS. Most notably, this was the first known use of s UAS for emergency response to an eruption, first use of rotary sUAS to sample air quality, and first live streaming of video from sUAS over the new FirstNet cellular network.