Delft University of Technology, Department of Cognitive Robotics, Mekelweg 2, Delft, 2628CD, the Netherlands.
Cruden, Pedro de Medinalaan 25, Amsterdam, 1086XP, the Netherlands.
Appl Ergon. 2023 Jan;106:103897. doi: 10.1016/j.apergo.2022.103897. Epub 2022 Oct 4.
Increasing levels of vehicle automation are envisioned to allow drivers to engage in other activities but are also likely to increase the incidence of Carsickness or Motion Sickness (MS). Ideally, MS is studied in a safe and controlled environment, such as a driving simulator. However, only few studies address the suitability of driving simulators to assess MS. In this study, we validate a moving base driving simulator for MS research by comparing the symptoms and time course of MS between a real-road driving scenario and a rendition of this scenario in a driving simulator, using a within-subjects design. 25 participants took part as passengers in an experiment with alternating sections (slaloming, stop-and-go) with normal and provocative driving styles. Participants performed Sudoku puzzles (eyes-off-road) during both scenarios and reported MIsery SCale (MISC) scores at 30 s intervals. Motion Sickness Assessment Questionnaire (MSAQ) scores were collected upon completion of either scenario. Overall, the results indicate that MS was more severe in the car than in the simulator. Nevertheless, significant correlations were found between individual MS in the car and simulator for 3 out of 4 MSAQ symptom categories (0.48 < r < 0.73, p < 0.02), with a strong overall correlation (r = 0.57, p = 0.004). MS onset times were similar between the car and the simulator, and sickness fluctuations as a result of driving style showed a similar pattern between scenarios, albeit more pronounced in the car. Based on observed similarities in MS, we conclude these simulator results to have relative validity. We attribute the observed reduction of MS severity in the simulator to the downscaling of the motion by the Motion Cueing Algorithm (MCA). These results suggest that, at least in eyes-off-road conditions, findings on MS from simulator studies may generalize to real vehicles after application of a conversion factor. This conversion factor is likely to depend on simulator and MCA characteristics.
车辆自动化程度的提高预计将使驾驶员能够从事其他活动，但也可能增加晕车或运动病（MS）的发生率。理想情况下，MS 是在安全和受控的环境中研究的，例如驾驶模拟器。然而，只有少数研究涉及驾驶模拟器评估 MS 的适宜性。在这项研究中，我们通过比较真实道路驾驶场景和驾驶模拟器中的该场景的呈现之间的 MS 症状和时间过程，使用受试者内设计验证了用于 MS 研究的移动基座驾驶模拟器的适用性。25 名参与者作为乘客参加了一项实验，该实验交替进行了（蛇行，走走停停）正常和挑衅性的驾驶方式。参与者在两个场景中都进行了数独谜题（视线离开道路），并在 30 秒间隔报告 MISC 评分。在完成任何一个场景后，都会收集运动病评估问卷（MSAQ）的分数。总体而言，结果表明 MS 在车内比在模拟器中更严重。尽管如此，在车内和模拟器中，4 个 MSAQ 症状类别中的 3 个（0.48<r<0.73，p<0.02）发现了 MS 个体的显著相关性，整体相关性很强（r=0.57，p=0.004）。车和模拟器之间的 MS 发作时间相似，由于驾驶方式引起的疾病波动在两个场景中表现出相似的模式，尽管在车内更为明显。基于观察到的 MS 相似性，我们得出结论，这些模拟器结果具有相对有效性。我们将模拟器中观察到的 MS 严重程度降低归因于运动提示算法（MCA）对运动的缩小。这些结果表明，至少在视线离开道路的情况下，从模拟器研究中得出的 MS 发现可以在应用转换因子后推广到真实车辆。该转换因子可能取决于模拟器和 MCA 的特点。
