Search

A trust model IMPACTS (intention, measurability, performance, adaptivity, communication, transparency, and security) has been conceptualized to build human trust in autonomous systems. A system must exhibit the seven critical characteristics to gain and maintain its human partner’s trust towards an effective and collaborative team in achieving common goals. The IMPACTS model guided a design of an intelligent adaptive decision aid for dynamic target engagement processes in a human-autonomy interaction context. Positive feedback from subject matter experts who participated in a large-scale exercise controlling multiple unmanned assets indicated the decision aid’s effectiveness. It also demonstrated the IMPACTS model’s utility as a design principle for enabling trust between a human-autonomy team.

Background In 2014, the Accreditation Council for Graduate Medical Education (ACGME) formally mandated trainee (resident and fellow) participation in health care quality improvement (QI) projects as one of the Clinical Learning Environment Review (CLER) Pathways to Excellence. Subsequent national reviews showed large variations in how QI education is conducted, as well as a significant mismatch between educational and organizational goals. Objective We developed a web-based platform to engage trainees in QI that better aligned with best practice methodology and matched identified institutional priorities. Methods A needs assessment survey was distributed to trainees to understand the obstacles to compliance with ACGME QI requirements. Based on the results, a web-based clearinghouse, called the QI Platform, was developed and launched in July 2016, and utilization was analyzed in February 2019. Results A total of 196 of 440 needs assessment surveys (45%) were completed. Themes extracted from surveys to identify barriers in QI participation included difficulties designing projects, lack of mentorship or expert support, and difficulty engaging an interprofessional team. Over 2.5 years, 151 projects were registered on the platform. Of these, 17 (11%) were collaborative entries. At the time of analysis, 166 of 437 trainees (38%) were listed as participants in active QI projects. A total of 22 projects were archived as complete, and 68 incomplete projects were reassigned to the “Ideas” section as works in progress after lead trainee graduation. Conclusions An institutional QI Platform clearinghouse for GME QI projects was feasible to develop and maintain, and it appeared acceptable to most GME programs and trainees for recording and tracking QI projects, and linking these to hospital QI priorities.

An experiment was conducted to investigate the effects of automation and role allocation on performance in a simulated picture compilation task with fourteen two-person student teams. In the absence of automation support, the system integrated sensor information. In the presence of automation support, the system both integrated sensor information and identified contacts. Roles were assigned either based on warfare domain or geographical sectors. Results showed that human-automation system performance was similar in two automation conditions, but participants were slower in classifying tracks and overall classified fewer tracks when the automation was present. We conclude that working with automation may lead to degraded team performance due to complacency and additional task complexity.

A study was conducted to examine the operational use of multi-sensor data fusion for military airborne intelligence, surveillance and reconnaissance. Participants performed a simulated picture compilation task wherein they had to identify all ships and planes in their area of operation using various sensors. One group performed the task using only native sensor data. The second group had imperfect data fusion to help them resolve the kinematic information, but they still had to identify each contact. The results indicated that data fusion automation improved the identification of ships and planes over the native sensor group. However, over time, map clutter continually increased for the group with fusion automation, surpassing the clutter for the native sensor group. The results suggest that while multi-sensor data fusion has benefits for picture compilation, dealing with plot clutter from false and spurious tracks is a key concern. Interface suggestions are provided to mitigate the effects.

We conducted two experiments to examine the effect of providing automation support and communicating the limitations of the automation on team performance in a simulated navy environment. Two-person teams engaged with a picture compilation task with or without automation support. In the first experiment, there was no explicit explanations provided regarding the automation’s limitations. In the second experiment, limitations of the automation were communicated to the participants verbally. A comparison of two experiments revealed that participants classified fewer contacts when the automation support was present. Moreover, communicating the limitations of automation resulted in even fewer classifications than when no information was provided. Possible reasons for these results include confusion created by the additional information and reprioritization. These results highlight the complexity of delivering automation transparency to operators in safety-critical environments. We conclude that automation transparency should be carefully designed and delivered to avoid negative impacts on performance.

Introduction The COVID-19 pandemic has resulted in an increased use of Powered Air Purifying Respirators (PAPRs), by health care providers to mitigate the risk of viral transmission, especially for aerosol-generating procedures. In this study, we evaluate communication devices that could be used concurrently with PAPRs to promote improved communication. Methods We tested two devices, a Bluetooth earpiece and a throat microphone that operated over mobile networks, against a control scenario in a simulated operating room environment with participants donning PAPRs. Participants read a short paragraph to each other, transcribed short phrases, and evaluated the scenarios according to speech intelligibility, ease of use, and comfort. Results There were 30 participants of varying PAPR experience. The Bluetooth headset had the most accurate transcriptions, followed by control, and lastly the neckpiece (94.7%vs 88.4%vs 76%, p Conclusion Communication devices have the potential to bridge but also worsen communications barriers between providers donning PAPRs.

Ecological interface design (EID) is an approach to designing user interfaces that is based on the objective of providing functional system relationships to users in ways that reduce perceptual loa...

In this paper we introduce MIFresh, a grocery store system consisting of a large display and individual kiosks that aims to increase the demand of local produce. In cities like Detroit, where poverty and health are major concerns, increasing consumption of local produce can help create local jobs, sustain the environment, and improve health. We used rapid contextual design to analyze Detroit's existing food system and, based on our findings, iteratively design a solution. MIFresh uses proven techniques such as coupons and rewards points programs and delivers education and awareness about the importance of consuming local produce. User testing results among the target population are promising.

Military signal operators listen, transcribe, and respond to audio traffic over multiple audio channels, in high-level noise from vehicles and weapons. The messages typically overlap in time and may be difficult to disentangle. Two studies were carried out to determine the benefit of supplemental texting. Normal-hearing participants were tested in a mock up of a military command post. Brief messages were played simultaneously over a communications headset (dichotic) and a loudspeaker array, either in quiet or in a background of vehicle noise. The at-ear speech-to-noise ratio was 5 dB. Only those messages beginning with a pre-assigned call sign were encoded. Mean scores of 84% or better were observed for messages presented over the headset, although there was a clear right ear advantage in noise. Messages coming over the loudspeakers were more difficult to understand but a visual cue directing attention to the source of an incoming targeted message resulted in a significant improvement of 7%. Replacing audio messages over the loudspeakers in noise with visual or audiovisual presentations resulted in an improvement from 71% to 96% that did not negatively affect headset performance. The data suggest that texting is a viable option for communication in cases of degraded audio.