Enhancing the Teleoperator’s Situation Awareness by Human-Centered HMI Design

Kurzfassung

Teleoperation of vehicles has the potential to serve as an approach to reap off the benefits of automated driving already in the near future when fully automated vehicles (SAE level 5 [1]) will not yet be fully operable. In teleoperated driving, a human op-erator in a remote control center boosts passenger safety and service reliability by monitoring the vehicle and taking over control when disturbances occur that the vehicle automation’s skills cannot cope with by itself. A user-centered human-machine interface (HMI) for the teleoperation of highly automated shuttles (SAE 4) by a public transport control center was generated to integrate the vehicle’s automation and hu-man remote-operation. The HMI prototype was evaluated regarding its usability, sit-uation awareness (SA) [2], acceptance, and perceived workload by transport control center professionals as an online interview study [3, 4]. Even though results supported our HMI overall, especially regarding usability, acceptance, and workload, SA ratings did not differ significantly from the scale mean, suggesting a merely average degree of SA perceived by participants. SA was measured using the Situation Awareness Rating Technique Starting (SART), a subjective self-report measure. Setting out from this particular result, the following questions arose: (1) Why did the current HMI not provide sufficient SA to the participants? What improvements need to be made in order to do so? Specifically, how did the remote setup as an online interview study influence factors such as involvedness and immersion? (2) Is the SART questionnaire-based method suitable to measure SA in this setting? If not, are there more suitable approaches to measure SA in teleoperated driving? In order to answer the first cluster of questions, a systematic analysis of the first HMI prototype’s evaluation was conducted. Both improvement suggestions from participants and additional approaches to enhance key outcome variables regarding the HMI were compiled. These include theoretical frameworks such as the Out-of-the-Loop concept in automated driving [5] as well as empirical findings on the interplay of teleoperator, environment, and operation interface characteristics on the teleopera-tor’s SA [6]. This compilation will subsequently be used to refine the HMI and then re-validate it by conducting a second evaluation study, preferably on-site using a lab-based simulator teleoperation workstation with scenarios that typically occur in tele-operated driving. To address the second cluster of questions, methods to measure SA will be systematically reviewed with a focus on objective approaches, such as the query-based Situation Awareness Global Assessment Technique (SAGAT) [7] that Endsley specifically suggested for measuring SA in teleoperation [8]. Subsequently, the SA measurement method that suits the requirements of teleoperated driving best will be selected to evaluate the refined HMI. The objectives for the focus group at the German Human Factors Summer School derive from the two central questions stated above: (1) Soliciting advise for designing the second evaluation study at the simulator teleoperation workstation and (2) discuss-ing the construct situation awareness, its usefulness in the context my research, how to measure it and what other related constructs and approaches could be considered.