Embedded Files
Assessing the learning curve of human operators under verbal distraction
Assessing the learning curve of human operators under verbal distraction
This study investigates how verbal distraction influences the learning curve and performance of human operators during a repetitive manual assembly task. Motivated by challenges in Industry 4.0 and 5.0, such as skill gaps, increasing cognitive demands, and human-centered manufacturing, the research focuses on understanding how learning efficiency, task completion time, and work quality evolve under distraction. While learning curves are well established in manufacturing and training research, the combined investigation of learning dynamics, cognitive distraction, and objective quality assessment remains limited. This study addresses this gap by proposing an integrated experimental and evaluation framework.
Ethical approval number: KEB_MK_RIT_2023_02
Seventeen participants completed a pattern-building task using eight blocks, repeatedly assembling a predefined spatial configuration. The experiment consisted of two phases. In the first, participants performed at least 20 repetitions without any distraction until a stable, “learned” performance level was reached, identified by a flattened learning curve. In the second phase, participants repeated the task 10 times while answering standardized verbal questions, introducing controlled verbal distraction similar to real workplace interactions. Task completion time was recorded for each repetition, while work quality was assessed using a camera-based system with ArUco markers attached to each block.
A video-based evaluation method was developed to objectively assess assembly accuracy. Marker detection enabled the calculation of angular and distance deviations between the assembled pattern and a reference configuration. These deviations were combined into a weighted quality metric, allowing subtle spatial inaccuracies to be quantified automatically rather than relying on binary or subjective assessments. This approach supports scalable and repeatable quality evaluation in human performance experiments.
The results demonstrate a clear learning effect during the undisturbed phase, with completion times decreasing significantly as participants gained experience. When verbal distraction was introduced, completion times increased for most participants, with average slowdowns ranging from 6% to 46%. Statistical analysis using paired t-tests confirmed that this increase in completion time was significant, supporting the hypothesis that distraction negatively affects efficiency. In contrast, work quality showed only minor variation between the undisturbed and distracted phases, and no statistically significant degradation was observed. As a result, the hypothesis that distraction reduces quality was rejected, while the hypothesis regarding increased completion time was confirmed.
These findings suggest that scheduled verbal distractions primarily affect efficiency rather than accuracy in relatively simple, well-learned tasks. The study highlights the importance of considering cognitive load and distraction when analyzing human performance and learning curves. By combining learning curve analysis, verbal distraction, a realistic manual task, and automated visual quality assessment, this research contributes a comprehensive framework for evaluating human performance in industrial contexts. The results have practical implications for the design of adaptive work instructions, training systems, and human-centered production environments, where managing distraction is critical to maintaining productivity without compromising quality.
Page updated
Google Sites
Report abuse