Assisted Teleoperation and Telepresence

Project Description

This project explores the development of advanced shared control systems and remote telepresence robotics to enhance task performance, manage cognitive load, and facilitate social interactions for teleoperators. Remote telepresence robots enable human operators to extend their presence to remote locations, fostering collaboration in critical applications such as healthcare, education, and industrial operations. Our research seeks to address the challenges of reduced situational awareness, heavy cognitive workload, and communication latency in telerobotics.

The integration of dialogue-driven navigation systems is a pivotal focus, leveraging natural language interactions to enable robots to interpret conversational cues and adapt their navigation goals dynamically. This fosters a more intuitive and human-centric approach to teleoperation, particularly in social and collaborative environments. Additionally, by employing benchmarking tools such as Fitts’ Law, we aim to establish systematic methods to evaluate shared control quality and its impact on user performance, workload, and trust.

Our overarching goal is to develop semi-autonomous teleoperation frameworks that seamlessly blend human intuition and robotic precision, improving efficiency, adaptability, and user satisfaction across diverse real-world scenarios.

Recent Progress

Recent advances within the project include the following:

• Cognitive Load and Trust Analysis in Teleoperation: A comprehensive user study examined how varying levels of robot autonomy influence cognitive load and trust during trajectory tracking tasks. Results showed that autonomy significantly impacts perceived cognitive load and trust but revealed no direct interaction between these factors. This emphasizes the need for considering cognitive load and trust as independent but interrelated metrics in shared control design​. (Pan et al., 2024) ​

• Adapted Fitts’ Law for Shared Control Benchmarking: The project introduced an adapted version of Fitts’ Law to quantify task difficulty in target-reaching scenarios. This model demonstrated how task complexity and robot autonomy affect human performance and cognitive load, providing a foundation for systematically assessing shared control systems. The findings suggest that task difficulty can be used as a proxy to estimate cognitive load, enabling real-time adaptability in shared control​ (Pan et al., 2025)

• Remote Telepresence Robotics for Social Interactions**: Introduced a dialogue-driven navigation system that interprets conversational cues between individuals to guide robot actions naturally and contextually. This approach reduces the need for explicit instructions, allowing teleoperators to focus on the broader dynamics of the remote environment.