Challenges of Bringing Augmented Reality into Nursing Medication Preparation

Augmented reality (AR) is frequently proposed as a support technology for nursing medication preparation, yet its integration into routine practice remains challenging. This study examines these challenges by situating pill recognition within the constraints of everyday ward work. An exploratory design combined a targeted literature review, workflow analysis, a market scan of wearable devices, and a proof of concept for image-based recognition of oral medication. While classification was feasible under constrained conditions, performance declined under realistic variability, particularly when distinguishing visually similar medications. The findings indicate that limitations arise less from model performance alone than from the difficulty of obtaining sufficiently informative visual input during routine tasks. Approaches that support sequential evidence acquisition and verification appear more viable than single-image recognition, emphasizing the need to align sensing, interaction, and workflow conditions.