THIS IS MY ROBOT: THERE ARE MANY LIKE IT, BUT THIS ONE IS MINE
Loading...
Authors
Yurkovich, Daniel M.
Subjects
trust
transfer of trust
trust in automation
manned unmanned teaming
interactive machine learning
human-automation interaction
machine learning
human-systems integration
explainable AI
explainable artificial intelligence
transfer of trust
trust in automation
manned unmanned teaming
interactive machine learning
human-automation interaction
machine learning
human-systems integration
explainable AI
explainable artificial intelligence
Advisors
Fitzpatrick, Christian R.
McGuire, Mollie R.
Date of Issue
2020-06
Date
Publisher
Monterey, CA; Naval Postgraduate School
Language
Abstract
This study explored the interactions of machine learning (ML) and serious gaming on trust in the context of a manned-unmanned team. While the government commits immense capital to develop autonomous systems for our warfighters, they often go unused due to skepticism of their performance and reasoning. Complexity and cost of the systems create an atmosphere that is prohibitive to daily training. These factors foster mistrust in valuable systems that could otherwise aid the warfighter.
In our experiment, the influence of serious gaming and autonomous behavior development was field tested with 40 participants in a two-group dual-task paradigm design to measure choice, trust indicators, and secondary task performance (STP). In a serious game, the control group learned the capabilities of an autonomous ground vehicle (AGV), while the experimental group “trained” the behaviors of the AGV. The experimental group invested significantly more time in the serious game. During execution of a live AGV task, no significant differences of trust indicators or STP occurred between groups. Time in the serious game in combination with trends in the choice of autonomous or teleoperated control of the AGV may imply that users prefer a user-trained AGV over an off-the-shelf solution. All data points to the need for further studies into the use of serious gaming to develop autonomous behaviors through an interactive ML approach.
Type
Thesis
Description
Series/Report No
Department
Computer Science (CS)
Organization
Identifiers
NPS Report Number
Sponsors
Funder
Format
Citation
Distribution Statement
Approved for public release. distribution is unlimited
Rights
This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States.