This project explores how augmented reality can make robot behaviour more transparent and easier to understand in human-robot interaction. Robots often move, pause, or react based on sensor information that is invisible to people, such as what the robot is seeing, where it is looking, or whether it detects nearby obstacles. This can make it difficult for audiences, operators, or collaborators to interpret the robot’s actions.
The project investigates whether AR can bridge this gap by placing the robot’s internal perception back into the physical space around it. Using the I-do robot as a case study, the research visualises sensor information such as camera view, field of view, bumper states, LiDAR data, and robot pose as spatial cues in augmented reality. In this way, the robot is not only seen as a moving machine but as an agent with visible perception.
The research is situated in a stage-based human-robot interaction context, where clear communication between robots and people is especially important. By combining ROS 2, Unity, and the Meta Quest 3, the project examines both the potential and the limitations of AR as a tool for robot transparency, including challenges such as real-time data communication, spatial alignment, occlusion, and user understanding.
