Industrial

El proyecto PREDILECT tiene como objeto el desarrollo de modelos no lineales de predicción de series temporales, basados en distintos esquemas de aprendizaje-máquina e inteligencia artificial (ML-AI) para la predicción en distintas escalas temporales (horaria, diaria, semanal, mensual, anual), de los consumos de electricidad y gas para los clientes de la empresa Aldro, de tal forma que dicha empresa pueda estimar de manera más precisa que en la actualidad el volumen de electricidad/gas que debe comprar.

This project aims to develop efficient communication schemes over a non-coherent channel for the multi-user MIMO regime. Non-coherent MIMO communication is gaining renewed interest due to the advent of massive MIMO system. When the transmitter and/or receiver is equipped with a large antenna array, the overhead incurred by the channel estimation may significantly limit the effective throughput, and make it more difficult to acquire channel state information (CSI) for channels that vary rapidly.

Desarrollo de un nuevo software para el análisis automático de tubos de generadores de vapor inspeccionados con la técnica de las corrientes inducidas.

This Project will provide technical and scientific support for the design and development of a distance measurement industrial system and for its integration in the SAYME wireless sensor networks system. Several sensor technologies will be considered and their compliance with the low cost, low power and small form factor specifications of the SAYME system will be evaluated. Moreover, signal processing algorithms will be developed to process sensor signals in order to provide reliable and accurate measurements.

The goal of this project is to develop a wireless distributed system for monitoring and predictive maintenance of tower cranes. The system will increase the security at work in this field since it is able to predict potential failures and to assess the risk of fall of the crane. The system consists on a set of wireless sensors (to measure load, wind speed, etc ) transmitting data to an acquisition and signal processing unit located close to the crane. The sensors and the unit communicate to each other using the ZigBee standard.

In this project the GTAS group developed a wireless system for information delivery among the different coachs of a train. Basically, the system is composed of an industrial PC, working as a video server, and several devices equipped with wireless 802.11 cards. The system permits the transmission of video, images or other information to passengers, as well as the transmission of other information among coachs useful, for instance, for synchronization between carriages. The system was successfully installed in several trains of the spanish FEVE railroad company.

This is one of a set of collaboration projects with Nuclenor to improve and update the integrated intrumentation system of the nuclear power plant of Santa María de Garoña (Spain). This project includes hardware and software developments.

In this project we developed a digital modem for communications through low power electric lines, specifically in the CENELEC A band (9-95kHz), with application to remote control and billing. During the project the particular characteristics of this transmission channel were studied (low SNR, impulsive noise, impedance variations, ...). Also the influence of these impairments in the parameters of the designed modem (shaping filter, oversampling ratio, synchronism, ...).

This was a european project within the BRITE-EURAM framework with the final goal of monitoring large turbo-generator sets in hydroelectric plants. The project leader was Iberdrola (SP) and, besides the GTAS group, the rest of partners were ANSALDO (IT), IRCOM (FR), LETI (FR) and the groups of mechanical engineering and photonics of the University of Cantabria. The work of our group was focused on the development of a flexible and powerful acquisition and signal processing unit that we called DASPU (Digital Acquisition Signal Processing Unit).