This study expands an existing Transparency-Based Action (TBA) model to the domain of variable autonomy for teleoperated robot navigation. A simulation of a teleoperated robot equipped with two Levels of Automation (LOAs): Low teleoperation and High-conditional navigation, was developed and evaluated for a dynamic office environment. Participants engaged in a navigation task under two switch control modes: robot-initiated (RI) and human-initiated (HI). The study implemented two TBA levels, High and Low, to assess their impact on task performance and participant perception. The findings revealed that the High TBA level improved task performance and user acceptance in the RI control mode, whereas minor differences were observed in the HI mode. Results suggest that higher transparency levels enhance performance and perception in robot-initiated control situations but are less critical in human-initiated control.