Current projects

ROBOPROX: Robotics and Advanced Industrial Production

The growing complexity of industrial production, combined with quality and time-to-market requirements, creates new challenges for engineering practices. This project focuses on research and development in robotics and advanced manufacturing. It aims to advance methods in control and optimization for systems, materials, and manufacturing, as well as robotics and computational methods for production. Our team specializes in human-machine collaboration – we investigate methods to enable more intuitive task specification by combining machine learning with multiple modalities (gestures, eye gaze, language) and prior knowledge in, e.g., CAD models and algorithms for improved task planning and scheduling under uncertainty. ROBOPROX is a partnership between Czech Technical University in Prague, Brno University of Technology, The University of West Bohemia in Pilsen, and Technical University of Ostrava. It is funded by the Ministry of Education, Youth and Sports of the Czech Republic, 2023-2028. Visit the project website for more information.

CoreSense: A Hybrid Cognitive Architecture for Deep Understanding

This project addresses research problems related to autonomous robots in open environments. Current robots do not fully understand their environments, complex missions, and the unexpected events that affect their performance. The lack of system-level awareness severely limits their performance, dependability, and trustworthiness. CoreSense aims to develop a profound core architecture, scientific theory, and technology for deep understanding and awareness of autonomous systems and cognitive robots to improve their performance, flexibility, reliability, and user experience. Our team contributes a generic methodology for physics-aware, data-driven modeling to generate concise, physically plausible, and explainable dynamic models of robots or their environments. We also support knowledge representation and reasoning by leveraging foundational ontologies such as SUMO and using state-of-the-art automated theorem provers to complement the models with formalization, automated reasoning, and query-answering capabilities. Finally, we contribute a dedicated risk awareness module to predict safety and reliability risks associated with cognitive robot operations in open environments. CoreSense is funded by the Horizon Europe, 2022 – 2026. Visit the project website for more information.

FlexCRAFT: Cognitive Robots for Flexible Agro Food Technology

The aim of the FlexCRAFT program is to develop new robotics solutions for agriculture and food processing. In these application domains robots must be able to handle a substantial variation in shape, size and hardness of products, which is still a major challenge for the current robotics technology. We focus on developing generic robot skills for operations like automated harvesting of tomatoes, processing and packaging of meat and the packaging of bags of crisps and boxes of biscuits.The consortium includes the following universities and companies: Wageningen Universiteit & Research, Eindhoven University of Technology, Delft University of Technology, University of Twente, University of Amsterdam, ABB, AgriFoodTech Platform, Aris BV, BluePrint Automation, Cellar Land, Cerescon, Demcon, Festo, GMV, Houdijk Holland, Marel Stork Poultry Processing, Maxon Motor, Priva, Protonic Holland, Rijk Zwaan, and 3DUniversum. FlexCRAFT is funded by the Netherlands Organization for Scientific Research (NWO), 2019 – 2025. Visit the project website for more information.

Completed projects

OpenDR: Open Deep Learning Toolkit for Robotics
The aim of OpenDR is to develop a modular, open and non-proprietary deep learning toolkit for robotics. We will provide a set of software functions, packages and utilities to help roboticists develop and test robotic applications that incorporate deep learning. OpenDR will enable linking robotics applications to software libraries such as tensorflow and the ROS operating environment. We focus on the AI and cognition core technology in order to give robotic systems the ability to interact with people and environments by means of deep-learning methods for active perception, cognition and decisions making. OpenDR will enlarge the range of robotics applications making use of deep learning, which will be demonstrated in the applications areas of healthcare, agri-food and agile production. The project is funded by the EU Horizon 2020 program, call H2020-ICT-2018-2020 (Information and Communication Technologies), 2019 – 2023. Visit the project website for more information.
 
R4I: Robotics for Industry 4.0
This project aimed at accelerating fundamental research in robotics and cybernetics, which is a necessary condition for the success of the Industry 4.0 initiative. The main focus was on industrial robotic manipulators, compliant mechanisms, mobile robotics, networked control systems, machine perception and learning. These topics are the pillars of modern robotics and have a strong growth potential. The project strived to connect scientific research to industrial needs and ran from June 2017 to June 2023 as a partnership of the Czech Institute of Informatics, Robotics and Cybernetics (CIIRC) at Czech Technical University in Prague with the Central European Institute of Technology, Brno University of Technology (CEITEC BUT) and the Faculty of Applied Science (FAV), University of West Bohemia in Pilsen. It was co-funded by the EU, under the registration number CZ.02.1.01/0.0/0.0/15_003/0000470, within the call Excellent Research Teams within the Operational Programme Research, Development, and Education. Visit the project website for more information.
DL Force: Deep Learning for Robust Robot Control
While robots can flawlessly execute a set of commands to achieve a task, these commands are mostly encoded by hand. There is a need for effective learning methods that can deal with the uncertainty in the robot’s environment, in particular when only broad goals are specified, and the learning algorithm has to learn motor commands to achieve these goals. This typically involves reinforcement learning (RL). However, current RL for robotics tasks relies on ad hoc function approximators and is typically not robust to changes in the task, environment, or robot uncertainty (compliant robot actuators, or wear and tear). The aim of this project is to integrate two emerging notions in order to make reinforcement learning for robot control more robust and efficient: dynamic feedback control policies for robust control combined with deep neural networks to learn low-dimensional parameterizations of such control policies. This approach promises a generic and robust approach to reinforcement learning for robotic control. The project was funded by Natural Artificial Intelligence program of the Netherlands Organization for Scientific Research (NWO), 2016 – 2019.
 
Cadusy: Control and data-driven modelling using symbolic methods
The goal of this project is to fully automate the process of system identification and control synthesis. We employ evolutionary symbolic methods to produce compact, interpretable models and control laws with minimal required human supervision. The symbolic nature of these methods enables the representation and manipulation of models containing continuous dynamical elements alongside with logic expressions, typical of cyber-physical systems. We automatically synthesize control laws in order to satisfy complex modeling and control objectives dictated by economical, performance and safety specifications. This project is a cooperation between Delft University of Technology and Eindhoven University of Technology and it involves industrial partners ASML, Flanders Make, Evolved Analytics and National Instruments. The project was funded by the Netherlands Organization for Scientific Research, domain of Applied and Engineering Sciences (TTW), 2015 – 2019.
 
Close Menu