Line “Application of Big Data to predictive maintenance“
The Chair is working on the development of visual language models for the detection of defects in steel plates, an aspect of great relevance for the maintenance of military vehicles. Advanced Zero Shot Learning techniques are being used, which consists of applying a model to a task for which it has not been trained. More specifically, we are using a visual model of the CLIP-based language that is able to segment objects on an image using another reference image (with the type of defect to be detected) or even a reference text (descriptive of the defect).
Line “Application of AI and data analysis to radar signal processing”.
The activities in this line focus on assessing the state of the art of drone classification against birds and on analyzing the capabilities and developing classification algorithms based on machine learning and target tracking radar data (trajectories).
The analysis of capabilities and development of classification algorithms has focused on machine learning techniques using features of the measured trajectories.
Line “Combat Cloud“
This line of research began with a study of the state of the art. Due to the fact that the concept of Combat Cloud is still somewhat diffuse, the analysis has been deepened by performing a taxonomy of the concept, identifying the different technologies that converge in it, in coordination with the study of the state of the art. The different elements that constitute this infrastructure are analyzed, highlighting their importance and role in the design and effective implementation of a Combat Cloud.
Web of the Chair at the UPM:
