BRUVIT: A new ultrasonic-assisted ball burnishing tool to improve the finish in turning operations
March 19, 2024
SASACI: A smart assistance system for architecture and the construction industry
April 19, 202404/04/2024
The LENA project emerged with the aim of creating continuous learning algorithms in the field of artificial intelligence (AI) that are robust and flexible. This project is led by the Mobile Robotics and Artificial Intelligence Group (RAIG) from the Institute of Robotics and Industrial Informatics (IRI, CSIC-UPC) at the Universitat Politècnica de Catalunya - BarcelonaTech (UPC). LENA is focused on developing innovative technologies that overcome the current limitations in AI algorithms that, once trained, tend to show a certain degree of rigidity in the face of new, unexpected situations.
Currently, most AI algorithms are optimised through data-based training processes. Sets of data are selected to try to attain the widest spectrum of possible situations. However, once the training is completed, the derived networks tend to remain static. LENA seeks to mitigate this restriction in the case of autonomous robot navigation, given that unexpected situations are likely to be faced.
As part of LENA, methods will be developed for incremental or online learning and the transfer of learning. In addition, there will be a focus on interactions between humans and robots, to explore how learning can be optimised through two-way communication that imitates human interactions. To ensure efficacy, all the techniques that are developed will be exhaustively tested with real data from robot navigation.
These new developments will enable algorithms to adapt to changes in the availability and nature of data over time. They will become valuable tools for applications with constantly evolving data. In addition, they will facilitate a substantial improvement in the capacity of robots to understand and interact with the physical world, expanding their knowledge through new experiences. The accuracy of robots’ predictions and decisions will improve due to better understanding of the context in which they process the information. The capacity to learn from data generated by humans and the adjustment of the behaviour of AI systems based on this learning will promote the generation of innovative ideas and more natural and effective interaction with human users.
The LENA project not only represents a significant advance in the field of robotics and AI but will also contribute substantially to improving the interaction between humans and robots in dynamic, changing environments.
Budget and funding
The total budget for the LENA project is €101,250, and the expected duration is from 1 September 2023 until 31 August 2026. It was funded through the State Plan of Scientific and Technical Research and Innovation 2021-2023 (Spanish State Research Agency).
Project PID2022-142039NA-I00 funded by MCIN/ AEI /10.13039/501100011033 and by "ERDF A way of making Europe".
Technology
You want to know more?
Related Projects
- The UPCxels project, led by the Intelligent Data sciEnce and Artificial Intelligence Research Group (IDEAI-UPC) at the Universitat Politècnica de Catalunya - BarcelonaTech (UPC), aims to create an open, cross-sectoral platform for a shared data space across multiple fields, with a special focus on the agri-food sector.
- A multidisciplinary research team led by the Intelligent Data sciEnce and Artificial Intelligence Research Group (IDEAI-UPC) at the Universitat Politècnica de Catalunya - BarcelonaTech (UPC) is participating in the Agrixels project, which aims to create an agricultural information standard—essentially, a data ecosystem—designed for AI-based services in the agri-food sector.
- The Food Safety and Control Research Center (CRESCA) at the Universitat Politècnica de Catalunya – BarcelonaTech (UPC), in collaboration with the Aquamaris Foundation, has developed a system that allows irrigation using seawater without the need for filtration, desalination, or chemical treatment, both in agriculture and gardening. The new system enables the adaptation of ornamental garden plants and leafy vegetable crops, eliminating the need for fresh water entirely.
- The Environmental Engineering (ENMA) research group leads the WhATTer project with the aim of developing an innovative and sustainable solution for the treatment of textile wastewater through an electrochemical/alkaline electrolysis system that allows for simultaneous hydrogen production. The project is carried out in collaboration with the Textile Technology (TECTEX) and Polymeric Materials and Textile Chemistry (POLQUITEX) research groups at the facilities of the Institute of Textile Research and Industrial Cooperation of Terrassa (INTEXTER) at the Universitat Politècnica de Catalunya - BarcelonaTech (UPC).
