[robotics-worldwide] Marie Curie researcher at Italian Institute of Technology
Giorgio.Metta at iit.it
Thu Feb 14 14:53:19 PST 2013
Marie Curie Experienced Researcher
The Fondazione Istituto Italiano di Tecnologia - IIT (www.iit.it) - was founded with the objective of promoting the country's technological development and further education in science and technology. In this sense, IIT's scientific program is based on the combination of basic scientific research with the development of technical applications, a major inspirational principle. The research areas cover scientific topics of high innovative content, which represent the most advanced frontiers of modern technology, with wide application possibilities in various fields ranging from medicine to industry, from computer science to robotics, life sciences, and nanobiotechnology.
The iCub Facility at the IIT (an English language Institute) located in Genoa is seeking to appoint a Marie Curie Experienced Researcher in the framework of the FP7 Marie Curie ITN RobotDoc (http://robotdoc.org/). RobotDoc (Robotics for Development of Cognition) is a multi-national doctoral training network for the interdisciplinary training on developmental cognitive robotics. The RobotDoc network consists of an excellent balance of academic and industrial partners, and of European and international laboratory leaders in developmental cognitive robotics. The network activities will have a significant impact on the career perspectives of the Fellows through training opportunities on industrial and academic research projects and skills.
There are several possible themes numbered 1 to 4 as detailed later.
* PhD degree in Robotics or related fields;
* Experience in C++ programming is a plus;
* Experience in control theory and control related issues are a plus;
* Creativity and proactive attitude;
* Fluency in both spoken and written English;
* Immediate availability
The applicant should satisfy eligibility for experienced researchers within an ITN as defined by the EC. In short, they should still be within the first five years of their scientific career and must be nationals of a country other than that of the host organization, where they will carry out their project. Researchers must not have resided or carried out their main activity (work, studies, etc.) in Italy for more than 12 months, in the 3 years immediately prior to the date of selection.
This post will be offered on a fixed term contract to finish at the latest on September 2013. Salary will be according to the rules for Experienced Researchers set out in the FP7 People Work programme 2008, including living and mobility allowance (may vary according to the family situation of the researcher). The Experienced Researcher will also be eligible to receive a travel allowance, as specified in the Marie Curie Grant Agreement. Further information about Marie Curie projects can be found at the following link: http://cordis.europa.eu/fp7/mariecurieactions/home_en.html
Interested applicants should submit a cover letter and detailed Curriculum Vitae to Giorgio Metta (giorgio.metta at iit.it) before 28/02/2013. (Ref: ER Marie Curie).
In order to comply with the Italian law (art. 23 of Privacy Law of the Italian Legislative Decree n. 196/03), we have to kindly ask the candidates to give their consent to allow IIT to process his/her personal data according to the following conditions:
"We inform you that the information you provided will be used solely for the purpose of assessing your professional profile to meet the requirements of Istituto Italiano di Tecnologia. Your data will be processed by Istituto Italiano di Tecnologia, with headquarters in Genoa, Via Morego, 30, acting as the Data Holder, using computer and paper based means, observing the rules on protection of personal data, including those relating to the security of data. Please also note that, pursuant to art. 7 of Legislative Decree 196/2003, you may exercise your rights at any time as a party concerned by contacting the Data Manager."
The Istituto Italiano di Tecnologia is an Equal Opportunity Employer that actively seeks diversity in the workforce.
Event-driven vision: developing bio-inspired vision systems for the humanoid robot iCub
Smart vision sensors, instead of merely reproducing with high accuracy the visual scene, are capable of transducing the sensory signal in a way best suited to extract relevant information for behaviour. To build such types of sensors, a growing community is taking inspiration from biological perceptive systems that show performance so far unreachable by artificial sensory systems. Mapping the characteristics of biological sensory perception onto artificial systems allows them to be inherently capable of better performances and, at the same time, be compact, low power, low cost, robust to noise and capable of adapting to ever changing working conditions, making these products suitable for real world real time applications. The "neuromorphic" iCub is equipped with one of such smart sensors, the Dynamic Vision Sensor , that asynchronously sends information about scene changes. The use of bio-inspired sensors on such a sophisticated robotic platform allows for the development of dynamic tasks where the robot interacts with the environment. This short project aims at developing event-based algorithms based on the output of the DVS sensor, to guide the behavior of the robot in selected demonstrators.
The performance of the humanoid robot iCub, equipped with the newly developed vision system, will assess quality and advantages of the method.
The successful candidate holds a PhD degree in computer science, robotics or similar and a background in artificial vision.
Hierarchical Inverse Dynamics Position/Force Control of Floating-base Constrained Robots
Floating-base robots (e.g. humanoid robots, walking robots) are a well-known class of underactuated systems, so they are not feedback linearizable as standard industrial manipulator. For this reason, the control of floating-base robots is still nowadays a challenging and active research field. This short-term project will focus on the derivation and implementation of a control framework for floating-base robots. In particular, the test platform is going to be the highly redundant humanoid robot iCub. Following the latest ideas of analytical dynamics and inverse dynamics control, we want to devise a computationally efficient software for multi-task position-force control. In particular we are interested in a scenario in which the robot has to balance and reach with one of its hands (extending some of our previous work). The robot could also make contact with the environment, to extend its reachable workspace, so force control has to be addressed. Requirements: the candidate should have a strong background in analytical dynamics and C++ programming. A good knowledge of control theory is also expected.
Human detection for face-to-face Human Robot Interaction
Recent research in robotics is trying hard to push robots out of factories and research laboratories. Effective operation in everyday environment requires the ability to detect humans and interact with them. Most of the successful vision-based strategies for people/pedestrial localization are tuned for "at distance" scenario (see (Dollar, Wojek et al. 2012) for a review) and are not well suited for "face-to-face" Human Robot Interaction scenarios. Initial approaches for egocentric vision attempt mainly to infer the social environment where humans move (Fathi, Li et al. 2012; Park, Jain et al. 2012) and rely on face detection and orientation estimation. This information only allows performing simple inference about the social interaction in the scene and it loses almost all the information about actions which are mainly conveyed by limb motion. In more complex cases, where also the type of actions has to be inferred (e.g. pointing, grasping etc.), the egocentric vision algorithm has to also estimate the localisation and orientation of limbs. This problem is not yet efficiently solved in unconstrained scenarios with egocentric vision-like sensors. This project will be performed on the iCub robot (Metta, Sandini et al. 2008), with the aim to implement efficient tools for the detection of human body parts from the cameras in the head of the robot. We will develop new methodologies for detecting human body parts at close distance such has torso, arms and face (i.e. upper body). The input data will be the disparity and optical flow maps obtained from the stereo pair embedded in the robotic head as well the raw color images from the cameras. We will also consider using kinematic models to improve localization.
This project will be carried out within the iCub Facility in collaboration with the Department of Pattern Analysis and Computer Vision (PAVIS). The ideal candidate should have background in Computer Vision and/or Machine Learning. He should also have computer programming skills and be motivated to work on a robotic platform.
Reaching along with whole-body balancing through real-time nonlinear constrained optimization
Over the recent years, research on humanoid robotics has been giving an increasing attention to the problem of whole-body coordination. Indeed, as hardware components get constantly and quickly upgraded for bipedal platforms such as Asimo, HRP4, iCub, new theories and control frameworks are conceived to let humanoids behave safely during interaction in human environments. In this context, the purpose of the short-term project is to enhance the motor capabilities of the iCub robot, currently available for the reaching tasks, by incorporating a balancing system which will allow maintaining the upright posture on the legs in a scenario where the robot will perform complex bending movements with the upper-body limbs to attain far objects. In particular, a methodology that has been lately demonstrated to be effective in dealing with highly redundant structures makes use of state-of-the-art optimizers that can solve online for a given kinematic task, subject to a set of nonlinear constraints. The objective would be therefore to extend this approach including force/torque measurements and real-time control design to account for the robot dynamics and meet the balancing requirement.
Requirements: the candidate should have a strong background in control theory and software engineering. Knowledge of nonlinear optimization techniques would be beneficial.
Giorgio Metta <giorgio.metta at iit.it>
Istituto Italiano di Tecnologia
Via Morego, 30
16163 Genoa, Italy
Ph: +39 010 7178-1411
Fax: +39 010 7178-1205
More information about the robotics-worldwide