Pragathi Praveena
Carnegie Mellon University
pragathi@cmu.edu
Bio
Pragathi Praveena is a Postdoctoral Fellow at the Robotics Institute at Carnegie Mellon University, where she contributes to the NSF AI-CARING Institute. Her research focuses on developing novel interactive systems that facilitate human-to-human communication and collaboration. Pragathi earned her Ph.D. in Computer Science from the University of Wisconsin-Madison, with her work supported by funding from the NSF and NASA. Her dissertation introduced a new paradigm for collaborative work through the development and evaluation of Periscope, a robotic camera system designed to enable remote, real-time collaboration on physical tasks between two people. Her work was nominated for Best Paper at HRI, and she was selected to be an HRI and RSS Pioneer. Prior to her doctoral studies, Pragathi was a researcher at the Xerox Research Center India and earned a B.Tech. in Electrical Engineering from the Indian Institute of Technology Madras.
Areas of Research
- Human-Computer Interaction
Robots for Teams: Facilitating Human-Human Collaboration through Intelligent Robotic Systems
I am interested in developing novel collaborative systems that leverage advancements in robotics and AI to facilitate human-human interactions. In my prior work, I developed and evaluated Periscope, a robotic camera system designed to enable an expert to provide real-time and remote guidance to a less experienced worker for procedural, physical tasks (e.g., wiring a car stereo). The remote expert views the workspace through a camera mounted on a robotic arm that is co-located with the worker. The dynamic robotic camera provides users with diverse and task-relevant views to facilitate shared awareness between the expert and the worker during the collaboration process. In collaboration with Boeing, I explored the application of the system in facilitating remote workforce training in manufacturing environments. A key innovation in Periscope is the shared camera control approach, where control of the camera is dynamically distributed among the worker, the expert, and the robot based on task needs. This approach allows users to maintain their desired level of control without overwhelming them with the complexities associated with the advanced capabilities of robotic arm platforms. Through my work on Periscope, I gained insights into the unique challenges of developing collaborative systems, particularly due to the complex nature of group interactions. These systems must be designed to enable multiple users to achieve collective goals while accommodating their diverse and sometimes conflicting individual goals. Additionally, the integration of artificially intelligent agents into such systems introduces a new layer of complexity to human-human interactions. The evaluations of Periscope highlighted the potential of shared camera control to address these challenges but also revealed the need for more nuanced policies to accommodate various social dynamics, such as familiarity and hierarchical structures within the group, which may evolve over time and may be reshaped by the use of the system.