This study aimed to construct a model that proactively predicts each nurse’s workload five minutes ahead in real time by utilizing information obtainable within the ICU, through simulation experiments simulating nursing tasks in the ICU. An experiment was conducted where 30 teams of five members each performed simulated ICU nursing tasks for two hours, collecting NASA-TLX data every five minutes along with various data concerning the nurses （experimental participants） and patients （simulators）. We applied five machine learning methods to predict five workload levels based on the 5-minute NASA-TLX WWL score, using 130 features derived from aggregated nurse and patient data, and compared their accuracy. While the prediction accuracy for the five levels did not exceed 40% for any model, it was found that the GBDT （gradient-boosting decision tree） model could achieve nearly 80% accuracy if a ±1 difference was tolerated. Analysis of feature importance revealed that incorporating data not only from the nurse being predicted for, but also from other nurses and patients not under their care, was effective for predicting each nurse’s workload level.

Technology to assess the state of VDT （Visual Display Terminal） workers is useful for improving work efficiency and health management. We developed a measurement method using an RGB camera, which imposes no burden on the user, with the aim of utilizing physiological responses that can capture the worker’s condition and unconscious reactions. While the first report focused on autonomic nervous system measures, this report centers on blinks and gaze estimation, which serve as indicators of concentration and cognitive load. As in the first report, data from 13 healthy adult males for 8 min of VDT work were included in the analysis. Facial feature points were extracted from facial video data using image processing, enabling the estimation of blink waveforms and gaze positions on the display. Blink detection evaluated using electrooculogram as the reference achieved better results than existing studies. For gaze estimation, waveforms showing the point and direction of gaze point were obtained, however the target accuracy for gazing region estimation was not achieved. The proposed method’s advantages include real-time processing using only a CPU without requiring specialized hardware and the ability to estimate blink waveforms from low-resolution videos. These results indicate that this method has high practicality for on-site use.

This study aims to comprehensively identify resilience competencies in the aviation domain using a large volume of near-miss event data collected through the Aviation Safety Reporting System （ASRS）. The analysis was conducted using BERT （Bidirectional Encoder Representations from Transformers）, a natural language processing method that captures context by learning word meanings through pretraining. Using 575 free-text narrative reports from ASRS, a model was constructed to extract resilience competencies while considering the four potentials of resilience （anticipating, monitoring, responding, and learning）. This model was then applied to analyze 42,938 reports. Sentences identified as containing each resilience potential （anticipating, monitoring, responding, and learning） were vectorized using Sentence-BERT, and k-means clustering was applied. By interpreting the meaning of each cluster, 31 resilience competencies utilized by aviation personnel to prevent near-miss events from escalating into accidents were identified. The comprehensiveness of these resilience competencies was confirmed through comparisons with related prior studies. Additionally, this study discusses practical examples of how these identified resilience competencies can be applied in operational settings.

The design of labels on infusion fluids manufactured and sold by Terumo Co., Ltd. was examined to prepare labels that are easy to check and reduce the risk of misidentification. To verify how easy the labels were for nurses to check, 10 infusion fluids in relatively frequent use were selected, and the labels in current use （current label） and the newly designed labels （new label） for each infusion fluid were compared. Eleven nurses examined the instruction sheets and infusion fluids. The presence/absence of misidentification incidents, gaze information, confirmation time, and subjective assessments were determined. One incident of misidentification occurred using the current label. Subjective assessment revealed that the new labels were easier to read and understand in more than 90％ of cases, presumably due to the layout of each new label, in which the most essential information for identifying the product were concentrated at the center of the label, enhancing identification efficacy. We intend to continue our efforts to achieve better label designs by more extensively examining opinions regarding these labels for actual use.

We report the results of a psychological evaluation of high water-retention floor wet sheets, developed according to consumer needs. Participants performed floor cleaning using both high water-retention floor wet sheets and typical water-retention floor wet sheets. We measured changes in positive and negative moods before and after cleaning, sheet usability, perceived satisfaction, perceived effort, perceived accomplishment with cleaning, and cleaning time. The results showed that floor cleaning with the high water-retention sheets significantly reduced negative mood compared to floor cleaning with the typical water-retention sheets. Additionally, the usability of the sheets, overall impression of the cleaning （such as perceived satisfaction） improved, and the cleaning time was reduced using the high water-retention sheets compared to the typical sheets. The high water-retention of the floor wet sheets, enabling effective removal of floor dirt, possibly contributed to these results.