Human-Robot Teams Informed by Human Performance Moderator Functions

Abstract

Standard Form 298 (Rev. 8/98) REPORT DOCUMENTATION PAGEPrescribed by ANSI Std. Z39.18 Form Approved OMB No. 0704-0188The public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing the burden, to the Department of Defense, Executive Services and Communications Directorate (0704-0188). Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ORGANIZATION. 1. REPORT DATE (DD-MM-YYYY)2 This research resulted in several key results. These results are highlighted here and discussed in more detail in the following sections. Teaming scenarios were modeled using IMPRINT Pro in order to predict the impact of various HPMFs on the teaming relationship. The developed human-human and human-robot models represented teaming relationships in which a) one team member was guided through a series of task steps and b) collaborative teams participate in tasks incorporating joint decision making for environments involving higher levels of uncertainty. These models provide predictions of human performance when working with a human or robot partner, and how performance differs based on the particular teaming partner. Model evaluations were conducted with single and multiple HPMFs and informed the validation evaluation designs. Additionally, the modeling results can inform human-robot teaming allocations.

Assessing physical workload for human–robot peer-based teams

“…The objective physical workload metrics provide the ability to assess this important aspect of peer-based human-robot interaction. While the previous analysis indicated that overall workload was lower in the H-R condition (Harriott et al, 2011a;Adams et al, 2012), it was determined that physical workload, an individual component of overall workload, does not follow the same trend, indicating that physical workload may be worth monitoring separately. Specifically, overall workload may not provide an accurate assessment of all team task activities involving physical movement.…”

“…Previous analysis of of the Collaborative evaluation Adams et al, 2012) and the Guided evaluation (Harriott et al, 2011a(Harriott et al, , 2011b indicated that overall workload was lower in the H-R condition. The Collaborative evaluation's investigation into physical workload also showed similar results to those produced by the Guided evaluation, even though the Guided evaluation's results were potentially influenced by the lack of a collocated human partner in the H-H condition.…”

“…An analysis of overall workload for the H-H and H-R teams found that the H-R team produced significantly lower workload in the Guided (Harriott et al, 2011a(Harriott et al, , 2011b and the Collaborative evaluations Adams et al, 2012). Due to observations of seemingly higher amounts of movement in the H-R conditions of both evaluations, a specific investigation into the impact of partnership on physical workload was conducted.…”

“…The developed models focused on estimating workload and time spent on the tasks. The evaluation analysis for both scenarios found that the overall workload was lower for the human-robot condition (Harriott et al, 2011a;Adams et al, 2012). The comparison between the Guided evaluation's overall workload results and its model predictions showed that while both the human-human and human-robot models did a good job of estimating the resulting overall workload and time spent triaging each victim, the human-robot team model more closely predicted overall workload (Harriott et al, 2011b).…”

Modeling Human Performance for Human–Robot Systems

“…Secondary tasks are separate from the primary task and can include, for example, responding to verbal memory questions or completing math problems. Response time and accuracy of responses to recall and recognition questions have provided insight into human workload levels and spare mental capacity (Adams et al, 2012;Harriott et al, 2011aHarriott et al, , 2011b. Metrics including correctness and speed to respond can indicate levels of participant performance and workload.…”