Miscommunication detection strategies for safe collaboration

本研究试图通过分析眼动仪和动作捕捉设备等传感器数据来检测容易发生错误沟通的情况，以避免操作错误。结果表明，沟通不畅的概率变得更高时，每个合作者平行工作，不看的合作伙伴，虽然这种情况并不总是导致沟通不畅。这表明，相互凝视数据是有用的，以检测团队，往往犯错误，虽然该方法很难被应用在真实的时间错误检测。

This research attempts to detect situations where miscommunication tend to happen by analyzing sensor data such as eye-trackers and motion capture devices for the purpose of avoiding operational errors. The results suggest that the probability of miscommunication becomes higher when each collaborator works parallel and does not look at the partners, although such situations do not always lead to miscommunication. This shows that mutual gaze data is useful to detect teams that tend to make errors, although the method is difficult to be applied in the real time error detection.