WP 1: Robot learning, autonomy, and mobility
Workpackage leader: Robert Babuška
- Autonomy, shared autonomy, mobility and persistence are important preconditions for future robotic applications. WP1 encompasses R&D efforts in this direction.
- Robotic tasks and their simulation provide data that can be used by machine learning algorithms to continuously improve the performance of robotic and other industrial production systems. Robert Babuška will bring his expertise in machine learning for robotic tasks. At the same time, R4I will provide him and the Czech industry real practical scenarios.
- Tasks 1D and 1E connect bidirectionally R4I to the extensive ČVUT background knowledge in mobile robotics.
- Task 1A: Machine learning methods for collaborative robots. Leader: Robert
- Task 1B: Multispectral and hybrid sensing for search and exploration in unknown areas. Leader Luděk Žalud.
- Task 1C: Model of robot environment, life-long autonomy. Leader Libor
- Task 1D: Robot mobility in unstructured environments. Leader: Libor Přeučil.
- Task 1E: Perception and calibration for robot navigation. Leader: Tomáš Pajdla.
WP 2: Perception, grasping and manipulation in industrial environments
Workpackage leader: Václav Hlaváč
- Exploring the perception-action cycle for tasks motivated either by the robot-human cooperation or by industrial manufacturing.
- Development of scientific methods, which are prerequisites for future autonomous industrial robots in the framework of Industry 4.0.
- Using the competitive advantage with respect to the state-of-the-art, which ČVUT has in its experience and in a testbed for dual-arm manipulation with flexible objects (including vision, tactile and force/torque feedback).
- Task 2A: Robot perceptual subsystems.. Leader: Václav Hlaváč.
- Task 2B: Dexterous dual-arm manipulation with soft objects. Václav Hlaváč.
- Task 2C: Human-robot collaboration in a shared environment. Leader Michael Valášek.
- Task 2D: Exploring friction and gravitation in sensory rich robotic structures. Leader: Michael Valášek.
WP 3: Networked control systems
Workpackage leader: Michael Šebek
- Networked systems become ubiquitous both on a lower cyberphysical level and on a higher knowledge representation and planning level. WP 3 deals with selected areas, in which both the CVUT team and Robert Babuška have experience, which are relevant to Industry 4.0.
- Closed robotic kinematic mechanisms can be viewed as networked systems too. Tasks 3D, 4D explore them in two areas vital for the future autonomous robotics relevant to the factory shop floor.
- Task 3A: System of systems. Leader: Michael Šebek.
- Task 3B: Human interactions with complex networks. Leader Michael Šebek.
- Task 3C: Multi-agent robotics. Leader Michael Šebek.
- Task 3D: Kinematic structures for controlled mechanical impedance. Leader: Michael Valášek.
- Task 3E: Mechatronic structures for grasping and manipulation. Leader: Michael Valášek.