• Reference: DPI2015-68087-R
• Principal investigator: Jorge Pomares Baeza, Pablo Gil Vázquez
• Research Centers: University of Alicante
• Duration: January 1st, 2016 – December 31th, 2018
• Financial Entity: Ministerio de Economía y Competitividad
• Funding: 150.000 €

Nowadays, robots and the capabilities of new robotic manipulation systems, like robotic hands,  can be used in assistance applications to alleviate the needs arising from aging people or with the aim of improving the quality of life for people with mobility problems due to diseases, motor functional disability, etc. It is in this area where this project fits, having the purpose of serving for assistance in manipulation tasks to people who have some kind of motor disability, as welfare assistance in their daily tasks. For this, the design of a system composed of two robot arms with their respective robotic hand is done, to imitate the human manipulation mechanism.
In social robotic environments, like the one where this manipulation system is intended to be applied, the robot must be able to adapt to changing, dynamic and unstructured surroundings. It also has to be able to interact  with humans. For this reason, a sensory system is included in the robotic manipulation system, based on vision, force and tactile information which allows to integrate in the robot the perception capability needed to manipulate unknown objects that are not previously modeled.
Research has been focused in new vision systems, visual control and dual manipulation systems that provide greater flexibility compared to existing manipulation systems. In addition, the presence of humans is considered, with the dual aim of ensuring their security and adapting the control and manipulation strategies depending on the activity being carried out. Specifically, a structure acting as a torso has been developed. This structure supports two robotic arms, and their respective robot hands located at its ends. In this torso, a multisensory system is used, which is composed by RGBD and thermographic cameras, and force and tactile sensors. This system allows the robot to extract information from the environment where it is located.