[robotics-worldwide] [jobs] Postdoc and Funded PhDs for Robot Vision and Sim-to-Real Learning and Manipulation at IIT, Italy

Fei Chen Fei.Chen at iit.it
Mon Mar 25 10:44:51 PDT 2019


The Department of Advanced Robotics at Istituto Italiano di Tecnologia (IIT), an English-language research Institute placed in Genoa, Italy, is looking for highly motivated full-time fellows with a key research role in the area of robotics/computer sciences applied to robot vision and sim-to-real learning. The selected candidates will join the Active Perception and Robot Interactive Learning Laboratory (the APRIL lab) and have the opportunity to pursue research within the three year "Learn-Real" project and "VINUM" project.

l  "Learn-Real" is a Chist-Era European project that aims to improve the sim-to-real reproducibility in learning physical manipulation skills of robots with physics-based, photorealistic simulator considering realistic variations of the environment (e.g. light and weather conditions). They will work in direct collaboration with researchers having an established expertise in robotics, robot vision, robot learning with an international academic consortium (Idiap/EPFL from Switzerland and Ecole Centrale de Lyon from France), and on various robot hardware platforms existed in IIT.

l  "VINUM" as an end-user driven application, tackles the emerging challenge of the dramatic shortage in skilled labor and lower grape price using robotic solutions. The main goal of this project is to develop and test innovative robotic mobile manipulation technologies for grapevine winter pruning automation. Robot vision and learning technique and pruning skills will be implemented on a full torque controlled robot mobile manipulator for grapevine recognition, manipulation, and pruning. They will work in direct collaboration with researchers having an established expertise in viticulture from Università Cattolica del Sacro Cuore (Italy), and robotics researchers from Università degli Studi di Napoli Federico II (Italy).

The selected candidates will be responsible with the development of robot simulators with high-quality realistic rendering capability of the simulated scenes and accurate physics simulating ability for robot modelling and dynamics. A further task will be the development of user-friendly toolset to let users contribute to the skill acquisition process (e.g., robot manipulates deformable objects) within this simulated environment, allowing the user: 1) to provide demonstrations by considering the discrepancy between the user and robot embodiment; 2) to reconfigure the virtual environment so that the robot can learn faster. Some work package might be on transferring robotic technologies (vision, manipulation) learned from the simulator to real challenging applications, e.g., grape vine winter pruning case.

The formal application window for PhD candidates will open in June 2019.

What required:

l  Experience programming of robot using simulator (e.g. Gazebo, V-Rep, pyBullet) OR game engine (e.g. Unity 3D, Unreal Engine 4) as a plus.

l  Experience programming in C++, Python.

l  Working knowledge on robot vision OR deep learning OR robot learning by demonstration.

Application Procedure:
Interested applicants should submit their detailed (1) CV, (2) motivation letter (with available starting time and duration), (3) any helpful information or documents to Dr Fei Chen (fei.chen at iit.it<mailto:fei.chen at iit.it>) with "Application - Learn Real [*Role], Your Name" as the email header. Please apply by April 30th, 2019.
*Replace [Role] with Postdoc or PhD

The Fondazione Istituto Italiano di Tecnologia (Italian Institute of Technology, IIT) is a Foundation established jointly by the Italian Ministry of Education, University and Research and the Ministry of Economy and Finance to promote excellence in basic and applied research. IIT has a vast experience in managing and supervising research projects with a portfolio of more than 364 external funded research projects, 180 of which financed by EU funding programs as FP7 and H2020 - 25 are funded by European Research Council (ERC). The Department of Advanced Robotics (ADVR) which concentrates on an innovative, multidisciplinary approach to humanoid robot design and control, and the development of novel robotic components and technologies. This encompasses activities from both the hard and soft systems areas of robotics. ADVR has approximately 150 personnel working in Humanoids, Quadrupeds, Machine Learning, Haptics and Interaction, Industrial Automation and Medical Robotics.



More information about the robotics-worldwide mailing list