Current Challenges

  • Improve productivity and supply chain efficiency in hazardous conditions.
  • Prioritise health and safety by minimising long-duration tasks using robotics.
  • Foster human-robot collaboration in everyday working environments.
  • Develop multipurpose robots to address labor shortages.
  • Enhance robotics data collection and analysis for improved decision-making.
  • Implement human-like senses in robots for intelligent decision-making.
  • Increase investment in AI and assess the return on investment.
  • Promote scalability of AI-driven robotics for business integration.
  • Align with the 2030 Climate Targets for energy efficiency and green job creation.

Objectives

  • Goal

    Develop a novel, pneumatic-based reconfigurable manipulator with advanced soft end-effectors capable of operating in environments with high risk of dust or water ingression while carrying out tasks involving both high forces/torques and delicate, complex manipulation

  • Goal

    Develop a Hierarchical Imitation Learning module, grounded on acquired knowledge alongside task planning algorithms with reactive planning capabilities including human-robot interaction

  • Goal

    Develop an Ontological framework for Knowledge Representation to enable robust and fault-tolerant collaboration and autonomous task completion through reasoning based on domain-specific fact understanding

  • Goal

    Develop a cognitive algorithm module encompassing Active Perception, Semantic Mapping and Localisation capabilities to fuse and orchestrate perception modalities in a dynamic context-aware that will enable identification changes in the environment and autonomous operation for longer periods while maintaining trustworthiness and dependability

  • Goal

    Establish an edge-native, resource-optimised and automated computing infrastructure able to support dynamic computing at the Cloud-Edge continuum realising an ecosystem where ML-models for new tasks and/or applications are efficiently updated deployed on-board the robot or in a distributed manner.

Expected Results

  • Two reconfigurable pneumatic-based robotic manipulators mounted on a mobile robotic platform
  • Novel soft end-effectors with variable stiffness that will allow for a diverse set of manipulation tasks
  • A Robotic perception module comprising 3D vision algorithms
  • A Localisation and Mapping module that allows the robot to accurately identify its position within the environment
  • A Semantic Mapping module powered by scene understanding algorithms
  • A Knowledge Representation framework that will capture important information regarding objects’ properties and relationships
  • A Hierarchical Imitation Learning framework for acquiring robotic skills to accomplish complex tasks directly from human demonstrators
  • A Human-Interaction conditioned Path and Task planning module enabling reactive robot control
  • An edge-AI framework for deploying Machine Learning models and computer vision algorithms at the edge in a streamlined fashion
  • Two demonstrators in solar and hydroponic farms