Assisting Visual Search Task with Augmented Reality: An Exploratory Study in an Industrial Workshop

Łysakowski, Mikołaj; Żywanowski, Kamil; Banaszczyk, Adam; Nowicki, Michał; Skrzypczyński, Piotr; Bohné, Thomas; Tadeja, Sławomir · 2024 · OpenAlex-citations

DOI: 10.1007/s11042-026-21516-y

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the challenge of visual search tasks in cluttered industrial environments, where operators must rapidly and accurately identify specific tools or objects amidst complex backgrounds. The authors highlight that this task is particularly difficult when target items are obscured or visually similar to nearby objects, a common scenario in workshops where diverse tools are grouped closely together. Efficient identification is critical for facilitating manual tasks such as assembly and device repair. To mitigate these difficulties, the study proposes leveraging Augmented Reality (AR) head-mounted displays (HMDs) equipped with advanced object detection capabilities to assist human operators. The research was conducted as an exploratory study involving eleven participants within a highly ecologically valid experimental environment designed to mimic real-world industrial conditions. The experimental setup utilized AR HMDs integrated with cutting-edge computer vision models, specifically combining the Segment Anything Model with the YOLOv8 object detection model. This technical integration was intended to enhance the system’s ability to locate and highlight target objects within the chaotic backdrop of an industrial workshop. The study aimed to substantiate the proposal that AR-assisted visual search could improve human performance in these demanding settings. The findings underscore the transformative potential of AR HMDs in efficiently locating objects within industrial settings. The results indicate that the explored approach significantly enhances human capabilities in visual search tasks, allowing for faster and more accurate identification of targets compared to unassisted methods. The integration of the Segment Anything Model and YOLOv8 was found to amplify the effectiveness of the visual search process, demonstrating the practical utility of combining segmentation and detection technologies in AR applications. The study provides evidence that AR assistance can reduce the cognitive load and time required for operators to find specific items in cluttered environments. The significance of this work lies in its demonstration of how AR technology can be effectively applied to improve operational efficiency in industrial workshops. By validating the use of specific AI models within an ecologically valid setting, the paper sheds light on novel experimental designs for human-computer interaction research. The conclusions suggest that integrating advanced object detection with AR HMDs offers a viable solution for enhancing manual tasks, potentially leading to broader adoption of such technologies in industrial automation and maintenance workflows. This study contributes to the field by providing empirical evidence for the benefits of AR-assisted visual search and highlighting the importance of robust object detection algorithms in real-world applications.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-25; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• visual search