Matthijs Biondina

Radboud University Nijmegen, Ghent University, Ghent University Hospital

Papers

Total Citations

H-Index

About

Matthijs Biondina is a robotics researcher whose work spans the frontiers of deformable object manipulation, social human-robot interaction, and evolutionary robotics. His most impactful contribution, a video dataset of human clothing-folding demonstrations (22 citations), addresses a fundamental challenge in robotics: the manipulation of non-rigid textiles. By providing a standardized benchmark for learning-based approaches, this work aims to accelerate progress toward general-purpose laundry-folding robots. Biondina also explores the integration of Large Language Models into social robotics (18 citations), investigating how LLMs like ChatGPT can enable more natural, open-ended spoken interaction when carefully designed for embodied, multimodal contexts. His research further extends to evolutionary robotics, where he has analyzed genome pleiotropy in soft robots (7 citations), examining how conflicting mutations between morphology and control can lead to premature convergence during co-optimization. This comparative analysis contributes to understanding how to better evolve both body and brain simultaneously—a challenge inspired by biological evolution. Biondina’s work demonstrates a rare breadth, tackling core problems in manipulation, social AI, and embodied evolution, each with implications for building more capable and adaptive robotic systems.