Description

Within this sub-line we can differentiate projects between those that develop controllers and the ones that focus on designing and producing robotic structures with unique capabilities. In the former category, we would like to highlight one of the GII’s main lines of research in the last two decades, namely evolutionary robotics. Given that controlling autonomous systems is a highly complex task, the approach followed in the GII has been to automate their design using advanced computational techniques, in this case, evolutionary algorithms.

In this line of research we have developed a modular and incremental architecture to create drivers, based on the evolution of behaviours created with neural networks successfully tested on different real robots. Cognitive MDB architecture using evolutionary algorithms for obtaining behaviour models has also been developed. Moreover, evolutionary algorithms have been applied in the design process of modular robots, leading to the EDHMoR (Evolutionary Designer of Heterogeneous Modular Robots) system. This system has been successfully applied on real modules manufactured in the GII.

On the other hand, first steps have been taken in a line concerned with multi-robot systems. We work with the ASIC (Asynchronous Situated Co-evolution) algorithm based on natural evolution to optimise the control of robot equipment on a permanent basis. This system has been put into practice with simulated and e-puck robots.

The second sub-line, designing and producing robots within the GII, responded to the need to solve real problems faced by many companies. The GII has already developed a series of robotic structures for performing tasks of radically different nature: a robotic system for blasting metal hull surfaces, an underwater robotic system for cleaning vessel hulls, a robotic system for the torque tightening of bolts in the ferrule-ferrule and reducing-axis flanged connections of wind turbines and a segment of automatic tightening by tensioning in flanged connections.

The heterogeneity of different lines of research carried out in the GII means that the team has an ever increasing number of resources that can be used to develop new robotic solutions for real applications.

Related publications