School of Measurement and Communication Engineering, Harbin University of Science and Technology, Harbin 150080, China.
Sensors (Basel). 2022 Jul 7;22(14):5115. doi: 10.3390/s22145115.
Poor electrical hazard recognition is a widespread issue in the production industry. Hazard perception has impacts workers' hazard recognition, causing them to experience unanticipated hazard exposure and suffer catastrophic injuries. To improve the factors of affecting hazard perception, the current study examined hazard recognition as an everyday visual search task. A comparative test was carried out combining the advantages and disadvantages of the two test methods. It was confirmed that the virtual image test data can replace the real image test data and demonstrate superior flexible settings performance, so the virtual image test method is used. A hazard perception test method based on wearable eye tracking technology was proposed to analyze the eye-tracking data (i.e., fixation, count, search duration, mean fixation duration, eye tracking, and hazard recognition performance feedback) were compared between experts in the field of electrical safety: skilled workers with at least five years of work experience and workers who had been on the job for less than a year. It was found that experts had a better hazard recognition accuracy and missed detection rate than other workers. Experts' hazards research track was more concised and paid less attention time. This advantage is most obvious in complex risk environments. The findings also suggest that workers who have different working years was not obvious visual search patterns other than the search duration. As can be seen the work experience is not an absolute factor in improving hazard perception. The present research will be useful to understand the influence of working years on hazard perception and provide a theoretical basis for corresponding training.
在生产行业中,工人普遍存在电气危险识别能力差的问题。危险感知会影响工人对危险的识别,使他们遭遇意外的危险暴露,并遭受灾难性的伤害。为了提高影响危险感知的因素,本研究将危险识别作为一项日常视觉搜索任务进行研究。该研究结合了两种测试方法的优缺点进行了对比测试,证实了虚拟图像测试数据可以替代真实图像测试数据,并具有优越的灵活设置性能,因此采用了虚拟图像测试方法。提出了一种基于可穿戴眼动跟踪技术的危险感知测试方法,分析了眼动跟踪数据(即注视、计数、搜索持续时间、平均注视持续时间、眼动和危险识别性能反馈),比较了电气安全领域的专家(至少具有五年工作经验的熟练工人和工作不满一年的工人)之间的差异。结果发现,专家的危险识别准确性和漏检率均高于其他工人。专家的危险研究轨迹更简洁,关注时间更少。这种优势在复杂的风险环境中最为明显。研究结果还表明,工作年限不同的工人除了搜索持续时间外,视觉搜索模式并不明显。可见,工作经验并不是提高危险感知能力的绝对因素。本研究将有助于了解工作年限对危险感知的影响,为相应的培训提供理论依据。