AI-assisted Work Services Utilizing human-AI Collaboration Levels: The Case of Ward Nurses; [인간-AI 협업 수준을 활용한 AI 서비스 연구: 병동 간호사 사례를 중심으로]

초록

Background The present study proposes a method for designing artificial intelligence(AI) services by gradually integrating the level of human-AI collaboration when introducing AI services in work environments. The research focuses on the role of ward nurses, which has been relatively overlooked in patient-centered healthcare systems, among various work environments where AI services can be utilized. Through a service design approach, the study aims to identify the specific issues that they face and proposes practical improvement measures. Methods First, based on the characteristics of the two entities—humans and AI—the study theoretically explored the concept of human-AI collaboration and the service factors that need to be considered in collaborative situations. Second, through in-depth interviews, the work processes and environments of the study participants were comprehensively understood, and problems were diagnosed. Third, based on this understanding, the study analyzed opportunity factors for human-AI collaboration, developed service scenarios according to levels of AI collaboration, and assessed preferences from the participants’ perspectives to derive key improvement measures. Fourth, service evaluation criteria specialized for AI services were established, and the proposed improvements were validated through evaluations and interviews on prototypes created based on the scenarios. Finally, the evaluation results were analyzed and visualized in a system map, and the study proposed the necessity of innovation toward a desirable future through a service blueprint based on levels of human-AI collaboration. Results The study found that nurses experience inefficiencies due to factors such as spending unnecessary time during shift handovers, receiving requests from multiple patients and their families simultaneously, leading to communication fatigue, and performing double work by manually recording tasks and later digitizing them. To address these issues, the study proposed the introduction of a high-level AI service, based on the preferences of the participants. The validation results showed positive evaluations in terms of usability, interaction, and human-AI collaboration, but relatively lower ratings in terms of trust. Conclusions The study identifies that excessive automation of AI services during the research process can lead to reliability issues, emphasizing the need for explainable, data-driven evidence. The study concludes that gradual changes are necessary considering the traditional workflow processes and unique characteristics of the medical field, highlighting the need to explore measures that ensure human intervention and judgment. This is an Open Access article distributed under the terms of the Creative Commons Attribution NonCommercial License (http://creativecommons.org/licenses/by-nc/3.0/), which permits unrestricted educational and non-commercial use, provided the original work is properly cited.

키워드