Paper Number
ICIS2025-2350
Paper Type
Complete
Abstract
This study examines the impact of different Augmented Reality (AR) interaction modes—human-driven, AI-driven, and AI-guided—on training efficiency and learning effectiveness in industrial training contexts. In collaboration with a major airline, we conducted a large-scale online experiment involving over 3,000 maintenance mechanics and a field experiment with 122 participants. Drawing on sustained attention theory, we find that while AI-driven interaction enhances training efficiency compared to human-driven interaction, it diminishes learning effectiveness due to limited user cognitive engagement. In contrast, AI-guided interaction, where the system provides intelligent suggestions but requires user decisions, strikes a balance between efficiency and cognitive absorption, leading to superior learning outcomes, particularly for less experienced users and tasks requiring problem-solving skills. These findings provide valuable insights into the design of AR systems that enhance both the speed and depth of learning, with important implications for the future of AI-supported industrial training.
Recommended Citation
Zhu, Runge; Yi, Cheng; and Li, Ting, "Human-driven vs. AI-driven AR Interactions: An Empirical Study in Aircraft Maintenance Training" (2025). ICIS 2025 Proceedings. 9.
https://aisel.aisnet.org/icis2025/imm_tech/imm_tech/9
Human-driven vs. AI-driven AR Interactions: An Empirical Study in Aircraft Maintenance Training
This study examines the impact of different Augmented Reality (AR) interaction modes—human-driven, AI-driven, and AI-guided—on training efficiency and learning effectiveness in industrial training contexts. In collaboration with a major airline, we conducted a large-scale online experiment involving over 3,000 maintenance mechanics and a field experiment with 122 participants. Drawing on sustained attention theory, we find that while AI-driven interaction enhances training efficiency compared to human-driven interaction, it diminishes learning effectiveness due to limited user cognitive engagement. In contrast, AI-guided interaction, where the system provides intelligent suggestions but requires user decisions, strikes a balance between efficiency and cognitive absorption, leading to superior learning outcomes, particularly for less experienced users and tasks requiring problem-solving skills. These findings provide valuable insights into the design of AR systems that enhance both the speed and depth of learning, with important implications for the future of AI-supported industrial training.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
08-ImmersiveTech