Giuseppe Cotugno

Giuseppe Cotugno
PhD Student

Email: giuseppe.cotugno (at ) kcl.ac.uk

 

 

 

Biography

Giuseppe obtained his bachelor degree in Computer Science Engineering at University of Calabria in December 2007. While studying at UniCal, he had the opportunity to join the Erasmus project visiting Vrije Universiteit Brussels. Afterwards, he moved to University of Rome, La Sapienza for his master degree. While at La Sapienza, he joined the SPQR Nao Team and took the opportunity to participate in Robocup 2009 held in Graz, Austria and other RoboCup competitions. He obtained his master degree in January 2010.

After graduation, he was visiting researcher at Humboldt Universitaet Berlin, focusing on grasping on the Nao robot, research assistant at Technical University Berlin, focusing on active segmentation and perception, and research assistant at University of Calabria, focusing on SLAM.

Since April 2012 he is working towards his PhD degree in Robotics within the Centre for Robotics Research, King’s College London. In 2013, he supervised the KCL DRC manipulation student team helping the team qualifying for the simulation finals and he was employed as Research Assistant on the DARWIN project until January 2015.

Since February 2015 Giuseppe is working full time as Software Developer Engineer at Buhler Sortex UK in the R&D department, while terminating his PhD studies.

Description of Research

Giuseppe’s research focus is on Affordance Learning for manipulation on the humanoid robots such as the iCub. Affordances encapsulate possible ways of using an object (Object Affordances) and possible ways of grasping it (Grasp Affordances). For example, a stick can be used to reach a far object or to mix a liquid in a cup and, typically, in both cases a different grasping posture is required. The main topic of Giuseppe’s research is to understand how different Object Affordances can be extracted from objects by observing how the grasping posture changes for each different affordance.

Giuseppe is also exploring novel grasping techniques inspired from neuroscience and physiology of human grasping. He is also evaluating how the functionality of human hands can be better translated into robotics. Another area of research is learning based on demonstration, in particular, how demonstrations can speed up or improve the learning process for manipulation, and which teaching technique might be more appropriate for learning a skill.

Giuseppe is also working on analyzing how humans are selecting objects in the context of building a stack of simple objects. Are people always rationally selecting objects in order to build the best stack, or other factors, such as previous experience, perception and rush are biasing the final result? This study, carried on with the help of KCL students, could help us understand is the shape of objects could bias the selection on a shelf or strong emotion, such as rush, might influence the best choices.

Finally, Giuseppe is strengthening his knowledge of biology and human neuroscience to understand which principles are worth translating into robotics to make those devices more autonomous and better scalable over different tasks. He is also interested in robust software design for robotics.

Research Interests

  • Object and Grasp Affordance Learning for Tool Use and Grasping
  • Humanoid Robots
  • Robotic Grasping and Manipulation
  • Grasping Synergies
  • Learning by demonstration
  • Software Engineering for Robotics
  • Translation of Neuroscience and Cognitives Sciences findings into Robotics
  • Translation of Biological Principles into Robotics

Publications

Journal Papers

The Role of the Thumb: Study of Finger Motion in Grasping and Reachability Space in Human and Robotic Hands
Giuseppe Cotugno, Kaspar Althoefer, Thrishantha Nanayakkara
IEEE Transactions in Systems, Man and Cybernetics: Part B, To Appear

Dynamic Motion Control: Adaptive Bimanual Grasping for a Humanoid Robot,
Heinrich Mellmann, Giuseppe Cotugno,
Fundamenta Informaticae, 112, 1, pp. 89-101, 2011
[PDF] [BibTex]

Conference Papers

Arousal decreases conservativeness in a random dot motion decision making task
W. Torgerud, D. Mussack, T. Lee, G. Maffei, G. Cotugno, P. Schrater,
Neuroscience 2015 – Society for Neuroscience

Affective decision making: effects of arousal on a random dot motion task
W. Torgerud, D. Mussack, T. Lee, G. Maffei, G. Cotugno, P. Schrater,
Computational and Systems Neuroscience (Cosyne) 2015
[PDF] [Link]

On the Dexterity of Robotic Manipulation: Are Robotic Hands Ill Designed?,
G. Cotugno, J. Konstantinova, K. Althoefer, T. Nanayakkara,
Proceedings of the Sixth International Conference on Cognitive Science 2014
[PDF]

Human Affordances of Stacking: Best Placement or Best Outlook?,
G.Cotugno, A. Ibrahim, K. Althoefer, T. Nanayakkara,
Proceedings of the Sixth International Conference on Cognitive Science 2014
[PDF]

Simplifying Grasping Complexity through Generalization of Kinaesthetically Learned Synergies,
Giuseppe Cotugno, Vishawanathan Mohan, Kaspar Althoefer, Thrishantha Nanayakkara,
Proceedings of 2014 IEEE International Conference on Robotics and Automation (ICRA 2014),
[PDF]

Extended and Unscented Kalman Filters for mobile robot localization and environment reconstruction
Giuseppe Cotugno, Luigi D’Alfonso, Walter Lucia, Pietro Muraca, Paolo Pugliese
21st Mediterranean Conference on Control and Automation (MED 2013)
[PDF] [BibTex]

A New Extended Kalman Filter Based on Actual Local Information for Mobile Robots,
Giuseppe Cotugno, Luigi D’Alfonso, Pietro Muraca and Paolo Pugliese,
9th European Workshop on Advanced Control and Diagnosis (ACD 2011)
[PDF] [BiBTex]

Particle-based SPQR Vision System for SPL Soccer Robots,
Giuseppe Cotugno,
Proceedings of the 5th Austrian RoboCup Workshop, 2010
[PDF] [BibTeX]

Thesis

Riconoscimento di Oggetti per Robot Umanoidi Calciatori,
Master Thesis (Italian only), 2010
Giuseppe Cotugno, Supervisor: Prof. Luca Iocchi
[PDF] [BibTeX]

Mechanism Design: L’Economista Ingegnere,
Bachelor Thesis (Italian only), 2007
Giuseppe Cotugno, Supervisor: Prof. Luigi Palopoli
[PDF] [BibTeX]

Reports

EU FP7 DARWIN Project Deliverable D5.2: Internal models and affordance maps of objects,
G. Cotugno, K. De Meyer, T. Nanayakkara

EU FP7 DARWIN Project Deliverable D5.3: Complete Observational Learning Architecture,
G. Cotugno, K. De Meyer, V. Mohan, T. Nanayakkara

Currently Supervised Students

Supervised Students