Information : cette page n'est pas traduite en français

Workshop on Managing Deformation: a Step Towards Higher Robot Autonomy

Objectives

The ability to interact with deformable objects of any scale is a prerequisite for advanced robot autonomy. Depending on the size of the object, the development of these skills may involve all the degrees of freedom of one or more robotics systems, and even the mobility of an active platform. These skills are necessary in multiple fields, ranging from grasping and handling of everyday objects (food, clothing, etc.) in consumer robotics, surgical robotic procedures (manipulating tissues, guiding flexible needles, etc), picking up of agricultural products, or movement of large flexible objects (such as cables, ropes, and tents).

However, taking into account deformation introduces new challenges, in particular:

  • The complication of sensing deformation
  • The curse of infinite dimensionality of the deformation configuration
  • The complexity of high nonlinearity in modeling the deformation

New paradigms are needed to address these challenges. Therefore, the aim of this workshop is to discuss new research prospects considering the object’s deformation in various robotic applications.

We can sense deformation with different modalities, e.g. force/tactile measurements provide crucial information for grasping and manipulation, but, this feedback is only locally available. Vision, on the other hand, offers a global picture of shape deformation but is subjected to noise and occlusion. We seek 1) methods to robustly sense deformation. Then, to address the problem of infinite dimensionality, we are interested in 2) compact feedback features for characterizing deformation. Another possible direction is 3) exploiting environmental contact for reducing the deformation dimensions. As for the high nonlinearity problem, we are interested in 4) data-driven/model-based/hybrid approaches to model deformation for control.

We propose a workshop to present and discuss new perspectives in managing deformation as a step towards higher robot autonomy. The purpose of the workshop is to bring together researchers who have common interests in working with deformable objects in robotic applications. The previously-presented challenges and possible directions of research are not exhaustive. The organizers encourage participants to propose new challenges and solutions in the scope of managing deformation. The workshop will also be open to video/abstract submission, and the final schedule will be adapted depending on the quantity and quality of the submissions.

Topics of interest (keywords)

  • Deformation sensing
  • Compact representation of deformation
  • Learning/identification/simulation of deformation
  • Multi-contact (environmental) planning for constraining deformation
  • Application-oriented research (agriculture, household, surgical, industrial)

Organizers:

  • Jihong Zhu, TU Delft/LIRMM, France/the Netherlands
  • Andrea Cherubini, LIRMM, France
  • David Navarro-Alarcon, The Hong Kong Polytechnic University, Hong Kong
  • Claire Dune, Laboratoire COSMER, Université de Toulon, France

List of Speakers:

  • Prof. Farshid Alambeigi, University of Texas at Austin (Confirmed)
  • Prof. Dmitry Berenson, University of Michigan (Confirmed)
  • Prof. Jeannette Bohg, Stanford University (Confirmed)
  • Dr. Frederic Colledani, CEA (Confirmed)
  • Prof. Fanny Ficuciello/Prof. Bruno Siciliano, Università degli Studi di Napoli Federico II (Confirmed)
  • Prof. Kensuke Harada, Osaka University (Confirmed)
  • Prof. Youcef Mezouar, Institut Pascal (Confirmed)
  • Prof. Jia Pan, University of Hong Kong (Confirmed)
  • Prof. Mathew Schwartz, New Jersey Institute of Technology (Confirmed)
  • Prof. Wenzhen Yuan, Carnegie Mellon University (Confirmed)
  • Prof. Carme Torras, CSIC-UPC (Tentative)
  • Prof. Akihiko Yamaguchi, Tohoku University (Tentative)

Dernière mise à jour le 31/05/2020