The goal of this master project is to implement and validate basic planning and motion control system for human-like dual arm-hand grasping and manipulation for an anthropomorphic robot system.

The Robotic platform that will be used for the implementations is ARoS (figura ao lado), which was developed at the Mobile and Anthropomorphic Robotics Lab, Department of Industrial Electronics.

The work builds on the results of the European Project FP6 JAST, which has been considered an European ICT success, (see   http://staging.esn.eu/projects/ICT-Results/Success_stories.html).

The work is integrated in the EU funded project PF7 Marie Curie \"NETT - Neural Engineering Transformative Technologies\".

The control and sensory-motor coordination schemes for the two upper limbs of the robot will:

(1) take into account experimental knowledge about how the human-brain solves the control and coordination problem in a robust and flexible way; and

(2) will be based on a mathematically rigorous formalism, i.e. dynamical systems and non-linear optimization, for modelling the planning and execution of co-operative and co-coordinated control of the two arms movements necessary for manipulation.

Ultimately, we expect results may be used to advance towards a new generation of assistive robots that exhibit human-like morphology and human-like dual arm-hand manipulation

Objectivos/Main goals and tasks:

• (i) To generate collision-free dual arm-hand trajectories:
• (ii) To achieve spatial and/or temporal coordination of both arms/hands.

Metodologia/Methodology:

• Use of anthropomorphic robots
• Programming languages (C/C++, Matlab)