Institut des Systèmes Intelligents
et de Robotique

Partenariats

Sorbonne Universite

CNRS

INSERM

Tremplin CARNOT Interfaces

Labex SMART

Rechercher

ISIR's priority projects in 2021

Each year, on the occasion of the Single Call for Projects of the INS2I, Institut des Sciences de l'Information et de leurs Interactions du CNRS, ISIR researchers are invited to build research projects to define the annual programme of the laboratory.

 

This year, our Scientific Council has elaborated a list of 5 projects in line with the scientific priorities of the laboratory. All of these projects will be carried out thanks to the support of our supervisors and our own shared resources.

 

LIST OF PROJECTS

 

The APIER project, Interactive Child-Robot Learning.

One of the major challenges in child-robot interaction in an educational setting is to enable effective and beneficial interactive learning over time. The robot must adapt online to the different children and their progress. In turn, the child should progress in learning through interaction with the robot. The aim of this International Research Project is to consolidate a partnership with the Polytechnic University of Athens. In recent years we have done pioneering work in implementing the e-learning capabilities of humanoid robots during interaction with children who are typically developing or with autism spectrum disorders. The challenge now is to demonstrate that this provides significant educational benefit over time compared to pre-programmed robots.

Team members : Mehdi Khamassi, Mohamed Chetouani, David Cohen, Rémi Dromnelle, Petros Maragos, Paraskevas Oikonomou, Catherine Pelachaud, Félix Rutard, Olivier Sigaud, Costas Tzafestas.

 

The GATHERBOT project, a flexible manipulator for picking.

The aim of this project is to develop a flexible manipulator for picking fragile objects such as fruit. This demonstrator would address the issues of research, design, modelling and control, and machine learning. The design will be based on a deformable structure without articulation and with a transmission by cables. The modelling control will seek to find a good compromise between model fidelity and computational time efficiency. Finally, learning from real and/or simulated data is an approach that will be explored, as it is more robust for systems that are difficult to model and that present a large number of variables, such as a deformable object that interacts with the environment. 

Team members : Faïz Ben Amar, Jérôme Szewczyk. 

 

The GRETA project, the incremental platform for the avatar of the future.

The Greta platform allows to control the multimodal behaviour of an animated conversational agent in real time. The platform is modular, but with a linear structure which does not allow to take into account frequent phenomena during an interaction such as interruptions, socio-emotional reactions of the interlocutors, or the implementation of adaptation phenomena. It cannot take advantage of the incremental structure of the acoustic and visual signal processing modules, and the dialogue modules.  To this end, we aim to transform the modular structure of the Greta platform into an incremental architecture by modifying its main modules (planning of intentions, behaviours and their realisation).

Team members : Catherine Pelachaud, Catherine Achard.

 

The PANACHE project, Interactive Machine Learning: towards a human-machine partnership.

In a technological context where Machine Learning is integrated in an exponential number of application cases, the users and practitioners of this technology are diversifying. The PANACHE project aims to contribute to the development of the Interactive Machine Learning (IML) theme, putting the user at the heart of the development of learning algorithms and thus facilitating their democratisation to a wide range of users. The project proposes to : 

  • to support the development of a design and development platform dedicated to a panel of heterogeneous users
  • the creation of resources for the development of scientific knowledge in the field and for its pedagogy.

Team members : Baptiste Caramiaux.

 

The RobCell project, poly-articulated atomic force microscopy for cell diagnosis.

The RobCell project aims to extend ISIR's MicroRob platform, which is part of the Robotex national network, by acquiring an atomic force microscope (AFM) head for cell diagnosis. This support will provide leverage in the research theme of correlative microscopy. The AFM head will be coupled to a 6 ddl micro-robotic system developed at ISIR, operating in a scanning electron microscope (SEM). This combination will offer the possibility of scanning the AFM probe over a sample by controlling three translations and three rotations of the probe in independent directions. A 6 ddl poly-articulated AFM is a breakthrough in the state of the art. It will allow a new possibility for the reconstruction of the topography of intracellular structures of complex shapes, for which a standard 3 ddl translational AFM has limitations of accessibility. 

Team members : Mokrane Boudaoud, Catherine Achard, Sinan Haliyo, Pascal Morin, Stéphane Régnier.