Today robots are largely used in factories but lack capabilities to sense and perceive the environment. They can interact with the environment only using accurate models and are not allowed to interact with humans. Programming robots is therefore expensive and time-consuming, and their use limited to very specific applications. The goal of our research is to develop humanoid robots that are progressively more autonomous and can effectively work in unstructured environments, operating in close interaction and cooperation with humans.
At this aim the Humanoid Sensing and Perception group studies algorithms and technologies that allow robots to sense the environment and react appropriately. Our strategy is to exploit the capability of robots to learn under human guidance or from the interaction with the environment by exploiting multiple sources of information (e.g. proprioception, vision, touch, and audition). Our activities focus on computer vision, tactile sensing and the development of the software technologies for seamless interaction of perception and action.
Our research tool is the iCub platform. We are one of the core groups of the iCub Facility department and we contribute to its development.
We have access to the equipment of the iCub Facility which includes among the other things: three fully-fledged iCub robots and related computing equipment, a small machine shop, mechanical and electronic design facilities.
The team is composed of computer scientists and engineers with competences ranging from computer vision, signal processing, machine learning to software engineering.
We work in close collaboration with other departments in IIT (see below) and have formal and informal international collaborations.
Robotics is a strongly interdisciplinary field to achieve our goals we work in close interaction with other groups at IIT.
- Giorgio Metta: iCub Facility
- Lorenzo Rosasco: Laboratory for Computational and Statistical Learning
- Alessio Del Bue: Visual Geometry and Modelling Lab
- Nikos Tsagarakis: software for whole-body control
- Lucia Beccai: human-like flexible tactile sensors