PILLAR-Robots - Purposeful Intrinsically motivated Lifelong Learning Autonomous Robots

Pillar-Robots focuses on developing a new generation of autonomous robots, able to work in unknown scenarios to determine their own goals and strategies to fulfil users’/designers’ desires. The objectives of the project are manifold:

  • from a theoretical perspective, to define and operationalise the concept of Purpose, to provide a path towards the design of more autonomous and multifunctional robots, able to operate in domains that were not known at design-time while performing tasks compatible with the desires of the designers;
  • from a computational perspective, developing algorithms and mechanisms within an appropriate cognitive architecture for lifelong, open-ended learning to allow the implementation of the new concepts introduced in the project within mainstream robotic platforms; from a practical perspective, to have a direct impact on the three fields of application contemplated in the project, i.e. agri-food processing, edutainment and industry/retail;
  • from a social perspective, to assess the ethical, economical and legal impact of adopting purposeful lifelong open-ended learning robots in real environments.

To achieve its objectives PILLAR-Robots, starting from a theoretical analysis, will first operationalise the concept of purpose and integrate it with intrinsic motivations and open-ended learning. Leveraging on previous results, the project will then focus on elaborating a new robotic motivational system supporting autonomous learning processes: new machine learning algorithms and mechanisms will be developed to foster autonomous goal setting, skill acquisition and representation processes devoted to the generalization and re-use of the acquired knowledge in different tasks and experimental scenarios. All these mechanisms will be integrated in a cognitive robotic architecture that will then be implemented and tested in real robotic platform performing in the fields of agri-food processing, edutainment and industry/retail, open up the automatization of areas that were previously out of reach for these systems. Finally, the project will address the dissemination of its outcomes among stakeholders, the integration with other relevant technologies, and the evaluation of possible impacts through socio-economic, ethical and legal studies around the main concepts of the project.

PILLAR-Robots aims at developing a new generation of robots endowed with a higher level of autonomy, that can determine their own goals and establish their own strategies, creatively building on the experience acquired during their lifetime to fulfil the desires of their designers/users in real-life. The project will operationalize the concept of Purpose, drawn from the cognitive sciences, to increase the autonomy and domain independence of robots during autonomous learning and, simultaneously, to lead them to acquire knowledge and skills that are relevant for operating in target real applications. In particular, the project will develop algorithms for the acquisition of purpose, ways to bias the perceptual, motivational and decision systems of the robots’ cognitive architectures towards purposes, and strategies for learning representations, skills and models that allow the execution of purpose-related deliberative and reactive decision processes. PILLAR-Robots will implement and validate demonstrators of purposeful lifelong open-ended autonomy within three different application fields: Agri-food, Edutainment, and unstructured Industrial/retail.

 

For more information:

 

This project has received funding from the EU Horizon 2020 research and innovation programme under grant agreement No. 101070381, project 'PILLAR-Robots - Purposeful Intrinsically motivated Lifelong Learning Autonomous Robots'