New Project: Principles of Learning from UnStructured Human-Robot Interactions (PLUS-HRI)

L-CAS is delighted to announce our latest research project in collaboration with the University of Nottingham

The Lincoln Centre for Autonomous Systems (L-CAS) at the University of Lincoln is embarking on an exciting new research initiative aimed at revolutionising how robots learn from everyday interactions with humans.

Project Overview

The Principles of Learning from UnStructured Human-Robot Interactions (PLUS-HRI) project addresses one of the fundamental challenges in robotics today: enabling robots to learn effectively from unstructured data generated through interactions with non-expert users in real-world environments.

Funded by the UK-RAS Network Low TRL Research Activities scheme, this collaborative effort between the University of Lincoln and the University of Nottingham will run throughout 2025.

The Challenge

As robots become increasingly prevalent in our daily lives, their ability to adapt and learn from unstructured interactions becomes crucial. While robots can be pre-programmed with specific behaviours, they inevitably encounter situations in real-world deployments where these capabilities fall short.

The PLUS-HRI project recognises that failures in dynamic human environments are unavoidable and aims to develop novel theoretical frameworks to enable robot learning and adaptation from in-situ interactions.

Research Goals

The project will:

  • Develop a new theoretical framework for uncertainty quantification, intention inference, and transfer learning in unconstrained settings
  • Build upon previous work with mobile robots in public spaces, particularly the “Lindsey: the tour guide robot” project
  • Use autonomous navigation as a case study to develop methods for learning from suboptimal demonstrations
  • Evaluate the theoretical framework through a new real-world Human-Robot Interaction (HRI) deployment

Broader Impact

The successful development of these methods will have far-reaching implications for:

  • Increasing autonomy in real human environments
  • Enhancing interpretability and usability of robotic systems
  • Democratising access to robot control

The interdisciplinary nature of this work means its impact will extend beyond Robotics and AI, with planned follow-on activities incorporating multi-disciplinary perspectives to explore the theme more widely.

For more information about this project, please contact Dr Francesco Del Duchetto.


This project is funded by the UK-RAS Network Low TRL Research Activities scheme.