Alia Fisher

Effective Human to Human Communication of Information Provided by an Unmanned Vehicle

Much of the research on unmanned-vehicles (UVs) focuses on technology or interface design. This study however, investigated how to best support effective communication between the operator monitoring a UV and the Soldier in the field using that information to complete a mission. Several questions arise: Does the operator need to be co-located with Soldiers in the field or can he or she be in a more secure rearward location? Does the team need the capability to transmit visual images or is radio communication adequate? Is information from one type of UV better than others? Do real time mapping and tracking technologies increase situation awareness (SA)? To begin to answer these questions, military teams conducted rescue missions using the video game Raven Shield as a simulated battlefield. The analysis of performance data, self reports, and observations provide some valuable insight to these questions.

Advantages of Co-Location for Effective Human to Human Communication of Information Provided by an Unmanned Vehicle

Much of the research on unmanned-vehicles (UVs) focuses on technology or interface design. This study investigated how to best support effective communication between the UV operator and the Soldier in the field using UV-provided information to complete a mission. In a previous study investigating the impact of different team configurations and the utility of supporting communication technologies, our team found co-location of team members to be beneficial (McDermott et al., 2005). In this experiment we investigate what aspects of co-location are key to successful team performance: Is face-to-face communication vital compared to voice-only when team members are distributed? Is the ability of the UV operator to see what the Soldier performing the mission can see critical? We also seek additional insight to inconclusive results from the first study regarding the utility of image transmission and access to an electronic map displaying both the UV and Soldier locations.

Bandwidth allocation in a military teleoperation task

The implications of bandwidth allocation are described for teleoperation in a military task that involved navigation, target detection, and target identification. Color versus grayscale imagery was manipulated. Participants themselves traded off resolution and frame rate settings. Participants minimized switching between resolution/frame rate settings and tended to use settings with high resolution/low frame rate. Courses completed with the highest resolution (and lowest frame rate) had the fastest target identification times, but no other differences were observed between settings. Color imagery offered advantages for overall course time and the time to identify a tank as friendly or enemy.

Team Communication with and without Aids for Transmitting Remote Information

The introduction of unmanned vehicles (UVs) and real-time electronic information presentation to military teams is intended to keep Soldiers safe and enable more effective performance. Two game-based experiments were conducted in which teams conducted a time-limited “Black Hawk Down” rescue mission with and without UV-provided information. When UV information was available, it was relayed between team members in face-to-face communication or remotely, with or without electronic maps, and with or without video images from the UV. Three types of analyses were performed. The communication content of the verbal protocol was analyzed and the use of information display technologies and employment of strategies was tabulated. The verbal protocol analyses revealed the push/pull of team communication and common confusions. Results of the technology use and strategy assessment form the basis for recommendations for display design and recommended practice for the use of such technologies in the field.

The Tradeoff of Frame Rate and Resolution in a Route Clearing Task: Implications for Human-Robot Interaction

The implications of bandwidth allocation are described for teleoperation in a military task that involved navigation, target detection, and target identification. Color versus grayscale imagery was manipulated. Participants themselves traded off resolution and frame rate settings. Participants minimized switching between resolution/frame rate settings and tended to use settings with high resolution/low frame rate. Courses completed with the highest resolution (and lowest frame rate) had the fastest target identification times, but no other differences were observed between settings. Color imagery offered advantages for overall course time and the time to identify a tank as friendly or enemy.

Cost Effective Approaches to Training Reconnaissance Tasks on Unmanned Systems

This study investigated the cost effectiveness of part task training for teaching operators to control an unmanned vehicle. It was assumed that initial training would be done via simulation and that learning would then be transferred to a live environment with a physical robot. Different combinations of part and whole task training in the simulation and live environment were tested. Findings supported the efficiency and transfer effectiveness of conducting part task training in the live environment prior to attempting the whole task in the live environment. Trainees spent time using relatively low-cost part task trainers. This translated to less time being used to achieve proficiency on the physical robot resulting in potential cost savings.

The Transmission of Spatial Route Information in Distributed Unmanned Vehicle Teams

The purpose was to understand the utility of different methods of communicating spatial information between distributed team members. An unmanned vehicle (UV) operator used a UV to scope out safe routes and enemy locations and used that information to help a Soldier conduct a rescue mission. There were four communication methods: visual, verbal, real-time verbal plus visual (VPV), and delayed VPV. We manipulated who was in command to see the impact on both pushed and pulled information. Results showed advantages of both visual and real-time VPV. We also found that with delayed VPV, the Soldier communicated infrequently which resulted in poorer performance, especially when the Soldier was in command. Results inform how beneficial UV information should be transmitted to Soldiers who need it in the field.

Methodologies for Assessing Situation Awareness of Unmanned System Operators

Situation awareness (SA) in human robot interaction is an important issue – because it is so essential to reconnaissance tasks and because remote viewing of the environment makes it so difficult. Two query-based techniques for assessing SA are presented and contrasted. Participants provided information regarding detection and identification of potential threats during a scenario. After the scenario, operators reconstructed where they navigated, the targets they encountered, and the identification of target type (e.g., friendly, neutral, or enemy). The first SA coding method (perception-based) compared post-experiment recollections of targets (location and type) to information the subject provided during the scenario, using the operator’s perceptions during the scenario as a baseline. The second SA coding method (reality-based) compared the same postexperiment recollections to actual target locations and type, using objective reality as a baseline. The SA reality metric was significantly correlated with task performance, but the SA perception metric was not correlated with objective performance. The SA reality errors were consistently greater than the SA perception errors, possibly because the SA reality measure includes both recall and task accuracy errors. The discussion explores why and when it would be beneficial to employ each method.

Error Training and Adaptive Remediation The Impact on Transfer Performance in a Complex Planning Task

An experiment investigated the relative effectiveness of three training strategies for improving transfer performance in a cognitively complex decision-making task. Error Management is an exploratory training approach in which trainees learn by exploring and making errors. Error Prevention training includes structured training on likely errors. Adaptive Remediation is error management training supplemented with practice scenarios tailored to the types of errors made. The study is an important addition to the literature because it examines training and transfer performance in a cognitively complex militarily-relevant task. Participants used a digital system to plan, execute, and monitor unmanned vehicles in reconnaissance missions. Transfer scenarios tested the ability to adapt what was learned when attempting new tasks and in an error prone environment. The results showed that transfer performance was better when participants did not complete adaptive remediation scenarios. It is believed that the adaptive remediation tasks encouraged participants to focus on single tasks and as a result they spent less time exploring the whole task. The implication is that the cost and effort required to create an adaptive remediation training program may not be justified in a cognitively complex task like the planning task used in the experiment.

Transfer from a simulation environment to a live robotic environment: are certain demographics better?

The ability to remotely operate an unmanned vehicle while simultaneously looking for suspicious targets and then classifying those targets is not a trivial skill. This study looked at different training approaches to make better use of simulation as a first training step. When transferring to a live environment, the operators could be grouped into two categories according to whether they passed live training criteria or not. There were clear performance differences between these groups. The group that failed to pass criteria had poorer performance overall, more SA errors, and spent more time in training. Post-hoc analysis showed differences in the demographics between those who passed and those that did not. Male participants and younger participants were more likely to achieve criteria. There were no differences in gaming experience and perceived sense of direction.