Abstract
This column will try to describe the characteristics of current cyberpsychology research in Europe. In particular, CyberEurope aims at describing the leading research groups and projects running on the other side of the Ocean.
For many applications, future robots should be able to learn how to solve multiple tasks in unstructured environments in an autonomous way. This project aims to build artificial intelligence architectures that should allow robots to self-generate goals and autonomously learn the sensorimotor skills needed to accomplish them.
Imagine your friend asks you to tidy up her room, which is full of furniture and objects. Now imagine that your friend shows you a photo of how the room should look after it has been tidied. This task would be boring for you, but you would nevertheless be able to carry it out with ease. Indeed, when you were a child, you played with all sort of objects, and driven by curiosity, you learned many flexible sensorimotor skills to manipulate them at will.
Differently from you, current robots are ill-suited for these types of challenges. Indeed, although new architecture and algorithms are being developed to control robots, getting them to accomplish tasks in unstructured environments still requires much programming or supervised training. Moreover, while current robots can solve specific tasks, in comparison to biological agents, they have very limited autonomy and flexibility.
To solve these issues, this project aims to develop a new paradigm to build open-ended learning robots called “Goal-based Open-ended Autonomous Learning” (GOAL). This paradigm rests upon two key insights. The first of these is that to exhibit an autonomous open-ended learning process, robots should be able to self-generate goals and hence tasks to practice. Second, new learning algorithms can leverage self-generated goals to accelerate skill learning dramatically.
The project follows a previous European project called IM-CLeVeR (“Intrinsically Motivated Cumulative Learning Versatile Robots”), which investigated how intrinsic motivations (IMs) can support autonomous learning in biological and artificial agents. IMs, related to novelty, surprise, and competence acquisition, are maximally apparent in children at play: driven by IMs, children explore and interact with objects in the environment, thus getting to know how they work and acquiring a wide set of sensorimotor skills that enable them to manipulate them.
The central idea of GOAL-Robots is that IMs can fully support open-ended learning only if a further critical ingredient is considered: goals. A goal is an internal representation of a state (or set of states or trajectories of states) of the world that can be internally activated by the agent in the absence of its current perception: when a goal representation is internally activated, it can contribute to move the agent's attention, behavior, and learning resources toward accomplishing the goal. The project idea is that endowing robots with the capacity to self-generate goals and use them to learn the skills for their accomplishment is the key ingredient to have truly autonomous open-ended learning robots.
The new paradigm will allow robots to acquire a large repertoire of flexible skills in conditions unforeseeable at the point of design with little human intervention, and then to exploit these skills to solve new user-defined tasks efficiently with no/little additional learning. The project will advance our understanding of the fundamental principles of open-ended learning in autonomous robotics by producing robots that can autonomously accumulate complex skills and knowledge in a truly open-ended way. This innovation will be essential in the design of future service robots, addressing pressing societal needs.
The project will develop the GOAL paradigm by pursuing three main objectives: (a) advance our understanding of how goals are formed and underpin skill learning in children; (b) develop innovative computational architectures and algorithms supporting the self-generation of useful goals based on user/task independent mechanisms such as intrinsic motivations, and the use of such goals to build large repertoires of skills efficiently and autonomously; and (c) demonstrate the potential of GOAL with a series of increasingly challenging demonstrators in which robots will autonomously develop complex skills and use them to solve difficult challenges in real-life scenarios.
The interdisciplinary project consortium is formed by leading international roboticists, computational modelers, and developmental psychologists working with complementary approaches. This will allow the project to advance our understanding greatly of the fundamental principles of open-ended learning and to produce a breakthrough in the field of autonomous robotics by producing for the first time robots that can autonomously accumulate complex skills and knowledge in a truly open-ended way.
Sources: Cordis, European Commission and European Union