Degraded Situation Awareness Research Articles

An Online Method to Study Remote Operation of Automated Vehicles

Remote operation is an emerging technical component of vehicle automation. One scenario could be a remote operator monitors the functioning of automation and take over vehicle control whenever needed. As the operator may face challenges such as degraded situation awareness or vigilance decrement, it is important to understand operator performance in a variety of takeover scenarios. This study presents an effort to create an online method for studying remote takeover performance. In this study, participants were presented with simulated driving videos and were required to indicate when they felt a takeover was needed. To ensure quality data collection in an online procedure, we made specific considerations when implementing the study. In this paper, we address these considerations and present preliminary data. Our findings suggest that operators may be less able to detect certain automation failures such as disobedience of road signs, and the online method could be useful although with limitations.

Degraded situation awareness in a robotic workspace: accident report analysis

We adopted an approach based on interactions in a robotic workspace, coupled with Endsley’s model of Situation Awareness (SA), to identify factors that potentially degrade operators’ SA: the eight SA demons. We propose here to study the implications of these demons in 45 industrial accidents involving robotic. These demons were associated with possible interactions a human may have in a robotic workspace. Our results showed that five of the eight SA demons appeared in the accident reports examined. The results showed that the identified SA demons were mainly associated with the first level of SA (perception level). We have also been able to describe five patterns of SA demon occurence in these accidents. SA demons are generally involved in human-robot interactions rather than human-environment interactions. Thanks to this analysis, we found that an operator’s SA is mainly impacted by the Errant Mental Model and Out-of-the-Loop SA demons in human-robot and human-environment interactions. Finally, we propose recommendations regarding workspace design.

Effects of remote and in-person verbal interactions on verbalization rates and attention to dynamic spatial scenes

This study focused on how teams allocated attention between a driving-related spatial task and a verbal task, and how different kinds of verbal interactions affected performance of the driving-related task. In Experiment 1, 29 two-person teams performed an interactive verbal task while one team member also performed a simulated driving task. Of the team members performing only the verbal task, half could see their partner’s spatial situation, as a car passenger can (in-person condition), and half were remotely located, similar to someone speaking to a driver using a cell-phone. Teams interacted verbally at an overall slower rate during remote than in-person interactions, suggesting that remote verbal interactions are more difficult than in-person interactions. Verbal interactions degraded situation awareness for driving-related information while performing the spatial task; and this degradation was not greater during remote than in-person interactions. Experiment 2 used a faster-paced verbal task and found greater degradation of situation awareness due to the verbal task. These findings are potentially relevant to the issue of how passenger and cell-phone conversations affect driving performance.

Differences in Remote versus in-Person Communications While Performing a Driving Task

In Experiment 1, 29 participants performed a simulated driving task both alone and while talking with another participant. Half of the non-driving participants could see the driving scene (in-person communication group) and half could not (remote communication group). When participants performed the driving task while talking with a partner, their situation awareness was significantly less than when they performed only the driving task. Thus, concurrent verbal interactions degraded situation awareness for the driving task. However, the amount of degradation in situation awareness during in-person and remote interactions did not differ significantly. The pace of the in-person and remote verbal interactions differed, suggesting that remote verbal interactions may be more difficult for drivers. Also, drivers talking with remote partners generated more long pauses than drivers talking with in-person partners, suggesting that drivers engaged in remote verbal interactions were modulating their verbalizations in order to maintain adequate driving performance. Experiment 2 replicated the finding that both in-person and remote verbal interactions degraded driving situation awareness, with no significant difference in the amount of degradation for the two types of verbal interaction.