notesum.ai

Published at December 10

Human-Computer Interaction and Human-AI Collaboration in Advanced Air Mobility: A Comprehensive Review

cs.HC

cs.AI

cs.LG

Released Date: December 10, 2024

Authors: Fatma Yamac Sagirli¹, Xiaopeng Zhao¹, Zhenbo Wang¹

Aff.: ¹University of Tennessee, Knoxville, TN 37996, USA

Arxiv: http://arxiv.org/pdf/2412.07241v1

Refer to caption

References	Key Focus	Methodology	Findings
[51]	Development of urban aerial networks using digital twin technology	Application of contextual data (population, job locations, building types)	Identified suitable vertiport locations and enabled real-time adjustments to aerial networks based on traffic volumes.
[52]	Visualization and interaction design of digital twins	Tangible digital twin framework combining 3D-printed models and holographic representations	Enhanced air traffic controllers’ (ATCOs) situational awareness and interaction through mixed reality headsets.
[53, 54]	Simulator-based MR approach for eVTOL pilot training	Immersive training environments for realistic flight simulations	Improved pilot proficiency and training methodologies for emerging aviation technologies.
[42]	MR training for eVTOL emergency scenarios	AI-powered Cognitive Agent assisting pilots	Evaluated AI performance metrics and integrated dialogue into multi-agent reinforcement learning for air traffic control.
[55]	Human-embodied drone interfaces	Integration of AR/VR and haptics into aerial manipulation	Enhanced capabilities for workers in dangerous tasks, applicable to future AAM cargo delivery.
[68]	Vertiport infrastructure for UAM	Introduction of the Vertiport Human Automation Teaming Toolbox	Facilitated real-time human-in-the-loop simulations for arrival, surface, and departure operations.
[69]	Cognitive human-machine interfaces for UAS operations	Development of a system using neurophysiological sensors and machine learning	Tested real-time adaptation capabilities in a bushfire detection and UAV coordination scenario, showing potential for enhanced performance.
[70]	Autonomous Decision Support System (DSS) for UTM	Analysis of human and system roles for effective Human-Machine Interfaces (HMI)	Introduced a Cognitive HMI concept to support closed-loop interactions and enhance system integrity in urban airspace.
[71]	Dynamics of human-AI collaboration in AAM	Reinforcement learning framework for adaptive interactions	Well-designed AI systems enhance situational awareness and support human operators in managing complex tasks compared to fully human or fully autonomous operations.
[77]	Trust in HCI within AAM	Investigation of trust levels across different flight phases and visibility conditions	Adaptive AR interfaces can enhance trust by providing real-time information and contextual awareness, although the interface is video-based and not interactive.
[78]	Impact of HMI design on passenger trust in eVTOL vehicles	Immersive simulator-based experiment with 34 participants testing four HMI conditions	Movement and hazard detection information significantly enhance passenger trust; gaze behavior correlates with trust levels, indicating potential for real-time trust indicators.