Kristin S. Moore

Perception of Robot Passability With Direct Line of Sight and Teleoperation

Objective: To examine participants’ abilities to judge the passability of robots through apertures in direct-line-of-sight (DLS) and teleoperation (TO) conditions, two experiments were conducted. Background: Past work has demonstrated that operators find it difficult to perceive aspects of remote environments during TO. For example, urban search-and-rescue operators have experienced difficulty judging whether a robot could pass through openings or over obstacles. Although previous research has discussed perceptual difficulties in TO, the differences between DLS and TO have not been quantified. Method: In the first experiment, participants judged the smallest passable aperture widths for three robot sizes for both DLS and TO conditions. In the second experiment, aperture widths were judged for three camera heights and two robot distances during TO. Results: In the DLS condition, participants produced similar judgments for the three robot sizes using dimensionless measurements. In the TO condition, participants’ judgments were more variable and they judged smaller apertures as passable. Conclusion: Overall, participants judged apertures that were too small for the robot to pass as passable. This tendency was more pronounced in four instances: as robot size increased, during TO, when the camera was at its lowest height, and as distance between the robot and the aperture increased. Application: Judgments of passability help to quantify differences in perception between DLS and TO. These results will be useful in the design of training regimes for TO tasks. Increasing operator understanding of performance differences under varying conditions will lead them to be more accurate when making critical decisions in remote environments.

Investigating the Relationship between Visual Spatial Abilities and Robot Operation during Direct Line of Sight and Teleoperation

Objective: To determine how scores on standard spatial measures correlate with the ability to operate a robot under different teleoperation conditions. Background: Past work has demonstrated that there is a relationship between visual spatial ability and teleoperation performance. Method: In this experiment participants completed a spatial visualization (VZ-2) and spatial relation (S-2) measure, and teleoperated a robot through both low and high difficulty courses under direct line of sight (DLS) and teleoperation (TO) conditions. Performance was determined by course completion time and the total number of collisions made during navigation. Results and Conclusion: Aggregate visual spatial ability was inversely correlated with operator performance under each of the experimental conditions. Analyzed independently, only spatial relations ability correlated with TO performance, while both measures correlated with DLS operation. Application: Better understanding of the relationship between spatial abilities and teleoperation performance can assist in the selection and training of future operators, as well as the design of superior interfaces.

Human Factors Evaluation of the Advanced Search Boolean Design of the Dynamic Enterprise Integration Platform (DEIP)

Introduction. Due to recent defense information superiority initiatives, intelligence assets have exponentially increased collection. Under the Maritime Domain Awareness (MDA) program, the Dynamic Enterprise Integration Platform (DEIP) has been created to mitigate this data overload and fuse several systems into a common interface. Practice Innovation. A recent human factors evaluation revealed concerns with the design of the DEIP Boolean Advanced Search Functionality. Users of DEIP appear to have difficulty constructing accurate queries. The goals of this study were to assess the accuracy of DEIP Advanced Search users and to determine the effectiveness of the addition of visible Boolean equations in the Advanced Search graphical user interface (GUI) header. Findings. Overall, expert users were 67% accurate using DEIP Advanced Search. Further, participants answered 74% of queries correctly without Boolean equations visible but only 57% correctly with equations visible. Discussion. Recommendations include additional usability testing, software design changes, improved feedback, and improved Boolean training.

A Call to Classify and Quantify Hospital Medication Errors: Results from a Meta-Analysis

In 2000, the Institute of Medicine published To Err Is Human: Building a Safer Health System which stated that between 44,000 to 98,000 patient deaths occur annually due to largely preventable medical errors. While this number is staggering, the number of errors adversely affecting patients in ways other than death must be much greater. The goal of the current study was to determine the types of in-hospital drug-related medical errors that occur, as well as to quantify the proportion and severity of those errors. A lack of consistency across studies in hospital error reporting prevented an accurate analysis of drug-related errors and their severity. The authors recommend future studies (and hospitals alike) adhere to, at minimum, five guidelines in error reporting. It is hoped that with standardized reporting methods a better understanding of medication-related errors can be gained, thus resulting in the design and implementation of error-reducing measures.

VRML use for Iterative Usability Testing of user Interfaces: A Novel Continuum Robotic Limb Example

Our team has developed a teleoperated robotic arm based on the biological model of an octopus. This “OctArm” has several advantages over traditional robotic designs, including the ability to grasp fragile objects with soft compliance, distributed pressure and the ability to pick up large or irregularly shaped objects. The development of an operator interface using a Virtual Reality Modeling Language (VRML) simulation has been a major focus. The simulation is used for both portable usability testing and as a visual display aid during the operation of the arm. Usability tests have been completed in the laboratory and the interface was tested in the field during an event sponsored by the Alliance for Robot Assisted Crisis Assessment and Response (ARACAR). VRML has proved to be a very useful tool for the development of the operator interface, robot control strategies, and training regimes. The use of VRML as a usability tool will be demonstrated.