A systematic review of research issues with Levels of Automation (LOA)

This systematic review critically analyzes the contexts and challenges of Levels of Automation (LOA), particularly in light of advancements by Artificial Intelligence (AI). Human–automation interaction (HAI) seeks to understand how humans interact with automated systems, particularly as increasing automation places operators increasingly out of the loop. A foundational framework within this field is the Levels of Automation (LOA) framework, which characterizes the respective roles and capabilities of human operators and automated systems. As AI becomes increasingly integrated into traditional HAI applications, a renewed examination of the current state of LOA is warranted. This systematic review examines the following research question: What research demographics, contexts, and challenges associated with LOA are reported in the literature? Data analysis, extraction, and open and axial coding were employed, and consensus was reached by iterative discussion between the authors. The review ultimately included 118 studies. The review identified six contextual categories and 26 challenge codes. The findings provide a comprehensive analysis of the evolving landscape of human–automation interaction. The contribution of this work is to outline weaknesses embedded in research with LOA and suggest a reform with prospects of enhanced HAI with AI.