Emanuele De Pellegrin (cohort student representative)
This project will explore the problem of learning epistemic action models suitable for planning from interactions with physical or simulated robots acting in complex environments.
The project will investigate a range of learning techniques and data from sources like robot vision and situated action effects, Ranging from the application of computer vision for human activity recognition to the high-level automated planning to achieve autonomy in a human collaborative environment.
Applications to real-world robot scenarios will also be explored, in particular in the Robotic Assisted Living Testbed (RALT) at Heriot-Watt University.
I'm a PhD student in the Edinburgh Centre for Robotics (ECR) at Heriot-Watt University. I received a BSc in Chemistry from La Sapienza University of Rome in 2016, and an MSc in Artificial Intelligence from Heriot-Watt University in 2020 with a scholarship from The Data Lab.
I entered the ECR’s PhD program in Robotics and Autonomous Systems in 2020/2021 with research interests in plan visualisation, epistemic planning, multiagent planning, and human-robot interaction. My current PhD work is focused on learning epistemic action effects for planning in a human collaborative environment.
With interest in robotics research, I have had internships on projects in the Computer Vision Lab at Heriot-Watt University and on the ORCA Hub project. I'm also interested in applying AI to video game development, particularly the use of automated planning techniques for games.