 |
 |
 |
 |
| |
|||||
|
|
|
|
||||||
Human-Unmanned System Interaction By Jill Drury
Despite being called "unmanned vehicles," none of the systems discussed in this issue could operate effectively without an experienced team of operators to monitor them. But human operators of unmanned vehicle systems can only be as effective as the interfaces that connect them to their machines. An interface needs to provide the operator with "situation awareness"—the operator's perception, comprehension, and predictions of the state of the unmanned vehicle and the environment surrounding it. In addition, an operator needs the interface to provide information on what other vehicles and their associated team members are doing to ensure smooth coordination. MITRE has found that the most effective interfaces provide their operators with situation awareness by presenting data based on interaction metaphors that make sense to the intended users. To further our research into innovative data display methods, we're observing unmanned aircraft system operators in action and developing analysis tools to better understand how our observations point to improvements in enhanced interfaces. The Apprentices A technique from the social science realm called Contextual Inquiry has been helping us get the most out of observation sessions in the field. The analyst becomes a temporary apprentice rather than a passive observer, asking questions about what's being done and why (when doing so would not disrupt operations). During our observation sessions, the unmanned aircraft operators often referred to their levels of situation awareness. We noted in particular that operators complained about a suboptimal level of situation awareness in regard to the spatial relationships between their aircraft and the terrain. Operators lacked sufficient context to understand what the video feed from their aircraft was depicting. They needed a more expansive field of visual data. A Matter of Perspective To provide operators with better visual context, our interface places the aircraft's video display into an inset over a background of preloaded map data. The interface ties the aircraft's position to the background map. The interface window then changes its perspective as the aircraft changes its position. The preloaded maps widen the operator's field of view, easing the feeling of trying to direct the aircraft by looking through "paper towel tubes," as one operator put it. In a traditional aircraft, the pilot can detect whether the aircraft is turning, climbing, or descending by feeling the physical forces exerted on his or her body. To help replace those cues, we used an exocentric, or "chase-plane," view of the aircraft as represented by a transparent silhouette. Thus, the operator's view of the aircraft's attitude can help in interpreting the live video shown on the interface. Experimentation with this approach has shown promise, so we have extended it to multi-aircraft control. When the operator is zoomed in on one aircraft, he or she sees much the same interface as in our single-aircraft design. When taking a broader view, the operator will see multiple aircraft icons and colored lines representing their planned path between pre-set waypoints. The idea is to smoothly change the view of the flight paths—and the rest of the environment—from any angle. Eyes on the Team 
Our latest innovations to the interface involve adding collaboration mechanisms so that a team of aircraft operators can be aware of each other's actions. First, we are providing an "overview in miniature" view that shows all the operators' aircraft, which operator is in charge of which aircraft, the territory that each operator can see, and the aircrafts' sensor footprints. Second, we are turning the entire map-based display area into a giant collaborative whiteboard. Operators will be able to mark interesting points on the video or map for the other operators to see. Except for markings deliberately chosen to be temporary, these points will persist even when aircraft move on to other locations. Finally, we are incorporating mechanisms to help operators determine quickly which other team member should be asked to take control of an aircraft. This is particularly important when team members are scattered throughout the area of operations. Such mechanisms would enhance the smooth transfer of control and reduce stress on operators. We are now in the process of sharing our results with MITRE's customers
so that our findings can influence the next generation of unmanned vehicle
interfaces. Our hope is that in the not-too-distant future, humans and
machines will be able to operate effectively as a team.
|
|
|||||
| For more information, please contact Jill Drury using the employee directory. Page last updated: August 20, 2007 | Top of page |
||||||
Serving as Architects of Information Advantage.™ |
 |
|
Unmanned
aircraft systems (UAS) require significant amounts of electromagnetic
spectrum for their missions. Line-of-sight and/or satellite
links are typically employed to control the aircraft, as well
as download high-bandwidth telemetry and payload sensor data
such as full-motion video. Spectrum availability and management
have emerged as significant concerns for both mission effectiveness
and safety of flight reasons. 
Download this issue [1.1MB]