The IP SENSOPAC (www.sensopac.org) investigates sensing, cognitive interpretation, and handling of objects using cerebellar models. The experiences and development of an integrated hand‐arm system form a direct basis for this project.
PHRIENDS project
The STREP PHRIENDS (www.phriends.org) Within the EU‐Project PHRIENDS, DLR and UNIPI are cooperating on new dependable algorithms for supervision and planning and new control algorithms for handling safe human‐robot physical interaction and for fault‐tolerant behaviour of existing research robotic devices. The technology of variable stiffness actuators, pioneered in Phriends, will be used in THE for controlling the hand impedance in grasping.
VIACTORS Project
Within the EU‐Project VIACTORS, DLR and UNIPI are cooperating on the development and use of safe, energy‐efficient and highly dynamic variable impedance actuation (VIA) systems, which will permit the embodiment of natural characteristics found in biological systems, into the structures of a new generation of mechatronic arms, legs, and rehabilitation devices.
The DEXMART project tries to mix research on natural and artificial cognition to develop robotic systems endowed with dexterous and human‐aware dual‐arm/hand manipulation skills for objects. Within this project, task‐dependent decision functions are analyzed resulting to an activity repertoire, while learning of new actions is based on online clustering of manipulation activities by structural, kinematical or topological abstraction. The limitation of such an approach is again the discretization of the task space, which limits the generalization of grasp primitives, as well as the object structural dependency, the identification of which, depends on high‐level sensors and time‐consuming algorithms.
The HANDLE project aims at understanding how humans perform the manipulation of objects in order to replicate grasping and skilled in‐hand movements with an anthropomorphic artificial hand. The goal is to evolve robot grippers from current best practice towards more autonomous, natural and effective articulated hands. The main difference with our approach is in our relaying on the neuroscientific and geometrical basis of sensorimotor synergies, rather than on a more phenomenological learn‐by‐seeing approach;
The goal of the project is to equip a highly biomimetic robot hand‐arm system with the agility, robustness and versatility of the human motor system by understanding and mimicking the variable stiffness paradigms that are so effectively employed by the human CNS. A key component of the project is the anatomically accurate musculoskeletal modeling of the human arm and hand.
The overall objective is to develop an intelligent artificial hand that looks and feels like a real hand. Smart-Hand aims to integrate recent advances in nano‐bioscience, cognitive neuroscience and information technologies in order to develop such an intelligent artificial prosthetic hand with all basic features displayed by a real one. The approach of this project is much more mechatronic‐oriented, and complements well with our more theory‐driven approach.
The objective of the project is to introduce a discontinuity in the robot design, allowed by a strategic alliance between Neuroscience and Robotics, beyond present, mostly fragmented, collaborations. Beyond its political merits, Neurobotics has started the integrated development of some of the experimental platforms used in this project, which will be then able to build upon its results.
GRASP Project
The project aims at the emergence of cognitive grasping through introspection, emulation and surprise. GRASP implements a predict‐act‐perceive paradigm that originates from findings of human brain research and results of mental training in humans where the self‐knowledge is retrieved through different emulation principles. The knowledge of grasping in humans can be used to provide the initial model of the grasping process that then has to be grounded through introspection to the specific embodiment. The project differs from our goals in its focus on the high‐level cognitive aspects of grasping, whereas we are more concerned with the detailed analysis of sensorimotor synergies at the peripheral and cerebellar levels: the results might therefore complement well, if eventually the two models will show compatibility. The mechanisms to transfer the knowledge from all the above projects, as well as any new FP7 projects in related research areas, to this project will be agreed by the Consortium. The opportunity to organise joint meetings and workshops with the consortia of the related projects will be discussed, so as to exploit at best the existing synergies across the area of Cognitive Systems and Robotics
ROBOCUB Project
The IP RoboCub project is a research initiative dedicated to the realization of embodied cognitive systems and the creation of an advanced robotic platform for neuroscientific study. The two main goals of this project are; The creation of an open hardware/software humanoid robotic platform for research in embodied cognition and the advancing our neural understanding of cognitive systems by exploiting this platform in the study of the development of cognitive capabilities in humanoid robots. Most of the work in grasping however are either based on the analysis of grasping primitives at the sensor level and the development of a neural controller for grasping, or at the classification of objects using Support Vector Machines (SVM). However, these methods use extensive if very effective, generic learning paradigms, while the grasping planner is based on discrete commands that are object‐specific.
