Teleoperation stands as a potential solution for hybrid assembly, combining the strengths of both the technical system and the human operator. With this tool for remote assembly, operators oftentimes do not have a direct view of the remote site. Thus, a visualization method is required to enhance the operator’s performance while posing minimal workload. While most research focuses on technological advancements of teleoperated systems and their visualization, there is a gap in understanding the human factors. Flat screens and head-mounted displays (HMD) are the most common visualization methods. However, current research presents contradictory results regarding performance and workload for operators while using flat screens or HMDs as visualization during teleoperation. Specifically, it remains unclear whether effects on the performance and workload of operators can be attributed to different dimensionality (2D vs. 3D) or to variations in the device (flat screen vs. HMD).
A study is conducted in the scope of this thesis to investigate the effects of different visualization methods on the performance and workload of an operator during teleoperated assembly. The investigation particularly focuses on differences between expert and non-expert operators. Both groups have distinct needs and preferences regarding visualization. Therefore, different visualization methods potentially affect the performance and workload of expert and non-expert operators differently.
For the study, gamers are considered expert operators due to their spatial abilities, skills with input devices, and their trust in technology. Both gamers and non-gamers perform an easy manipulation task of moving different shapes into a box with a teleoperated system. The system consists of two connected Franka Emika Robots by Franka Robotics GmbH with a Franka Hand gripper. Each study participant performs the task with four different visualization methods: direct view, 3D HMD, 2D HMD, and 2D screen. The distinction between the 2D and 3D HMD is made to distinguish between the effects of depth information (2D and 3D) and the device (HMD and screen). A ZED mini camera is used for stereoscopic video recording and the Oculus Rift as the HMD. The performance is evaluated by analyzing the Task Completion Time, whereas the subjective workload is evaluated via the standardized NASA Task Load Index questionnaire.
A between-within ANOVA statistical analysis for each performance and the workload did not find a significant difference between gamers and non-gamers. This finding might be attributed to all participants having a technical background and qualifying as expert users regardless of their gaming abilities. Thus, gamers are not an accurate representation of experts in teleoperation. Alternatively, an intuitive system design compensates for differences between gamers and non-gamers in performance and workload. However, a significant difference could be found between the visualization methods. Participants performed better and experienced less workload with visualization with stereoscopic depth information. The different devices only had a marginal influence on performance and workload. In view of these findings, future research should focus its efforts on enhancing depth perception (e.g. through improving secondary visual cues) to improve visualization during teleoperation with any device.
«
Teleoperation stands as a potential solution for hybrid assembly, combining the strengths of both the technical system and the human operator. With this tool for remote assembly, operators oftentimes do not have a direct view of the remote site. Thus, a visualization method is required to enhance the operator’s performance while posing minimal workload. While most research focuses on technological advancements of teleoperated systems and their visualization, there is a gap in understanding the h...
»