Research Lines/Projects


Summary: the group has 4 ongoing financed projects and 2 concluded projects in trading, energy and situational awareness applications and in basic multiagent systems research.

Energy

“Sustainable energy systems of the future will need more than efficient, clean, low-cost, renewable energy sources; they will also need efficient price signals that motivate sustainable energy consumption as well as a better real-time alignment of energy demand and supply.” ~W. Ketter, J. Collins, C.A. Block, Smart Grid Economics, working paper, 2010.

Ongoing

  • H2020-CONACyT big-data project  with applications to Smart Mobility, Pollution Management, Security, Energy and Health.
  • SmartSDK will deliver a smart city platform for emerging countries based on FIWARE technology and combining previous successful EU R&D experiences and support the creation of an ecosystem around the platform. The validated and integrated result will be transferred back to the European context and made available as instantiable platform in FIWARE Lab.
  • The reference architecture for smart city platform will be created as a software infrastructure built on top of basic Generic Enablers of FIWARE (Context Broker, IDAS, BigData) as software components that use the functionalities of the Generic Enablers in a transparent way for the final user by enriching in this way the FIWARE Platform. In this context, the smart city platform will be developed to be an intermediate implementation model between GE and the developers interface. In this “second level” the basic GE will be linked with mechanisms to represent the context information using maps, routes, etc.

more info

The leading trend to generate cleaner energy and reduce CO2 emissions is to  disintegrate energy generation. This trend is coined as distributed energy, to which  microgrids are cornerstone.
We take as a microgrid the interconnection of small, modular generation (micro-turbines, fuel cells, PV, etc.), together with storage devices (flywheels, energy capacitors and batteries) and controllable loads at low voltage distribution systems (see figure on the right). Normally, microgrids operates connected to and synchronous with the traditional centralized grid (macrogrid), but can disconnect and function autonomously as physical and/or economic conditions dictate.

microGrid

This project investigates how micro-generators should coordinate and collaborate to function as a bigger, more reliable and stable source of generation. Furthermore, we also investigate how should these microgrids adapt in a coordinated fashion to respond to load variabilities. more info

Past

forecast
El objetivo de este proyecto es desarrollar nuevas técnicas que permitan pronosticar la producción eléctrica en parques eólicos con una anticipación de más de 5 horas y desarrollar nuevas técnicas de Inteligencia Artificial distribuida con la finalidad de optimizar la compra-venta de energía.

Situational Awareness

Situational awareness is the ability to  monitor, make sense and predict what is happening in an environment. We’re interested in monitoring environments through multiagent systems, for example to maximize the number of events detected while monitoring an environment. read more…

Ongoing

  • H2020-CONACyT big-data project  with applications to Smart Mobility, Pollution Management, Security, Energy and Health.
  • SmartSDK will deliver a smart city platform for emerging countries based on FIWARE technology and combining previous successful EU R&D experiences and support the creation of an ecosystem around the platform. The validated and integrated result will be transferred back to the European context and made available as instantiable platform in FIWARE Lab.
  • The reference architecture for smart city platform will be created as a software infrastructure built on top of basic Generic Enablers of FIWARE (Context Broker, IDAS, BigData) as software components that use the functionalities of the Generic Enablers in a transparent way for the final user by enriching in this way the FIWARE Platform. In this context, the smart city platform will be developed to be an intermediate implementation model between GE and the developers interface. In this “second level” the basic GE will be linked with mechanisms to represent the context information using maps, routes, etc.

more info

agriculturaUAV

This is a continuation of Precision agriculture through collaborative UAVs. The objective here is to augment previous achieved results with more information to automate the decision making process. The technical and scientific challenges will be to coordinate  two sensor networks: a ground WSN that obtains soil and plant data through proximity sensors and a mobile WSN that obtains multispectral and IR imagery through UAVs.

more info

Past

agriculturaUAV
This is an industrial project whose objective is to compute UAV trajectories given human input about the objectives. UAVs gather information in the form of multispectral images and the information is then aggregated in a data center for it’s analysis. The analysis include sectioning the terrain into zones and classifying each zone into it’s most probable type.

more info

Multiagent Systems

We conduct basic and applied research in multiagent systems.

Ongoing

  • H2020-CONACyT big-data project  with applications to Smart Mobility, Pollution Management, Security, Energy and Health.
  • SmartSDK will deliver a smart city platform for emerging countries based on FIWARE technology and combining previous successful EU R&D experiences and support the creation of an ecosystem around the platform. The validated and integrated result will be transferred back to the European context and made available as instantiable platform in FIWARE Lab.
  • The reference architecture for smart city platform will be created as a software infrastructure built on top of basic Generic Enablers of FIWARE (Context Broker, IDAS, BigData) as software components that use the functionalities of the Generic Enablers in a transparent way for the final user by enriching in this way the FIWARE Platform. In this context, the smart city platform will be developed to be an intermediate implementation model between GE and the developers interface. In this “second level” the basic GE will be linked with mechanisms to represent the context information using maps, routes, etc.

more info

microGrid
We take as a microgrid the interconnection of small, modular generation (micro-turbines, fuel cells, PV, etc.), together with storage devices (flywheels, energy capacitors and batteries) and controllable loads at low voltage distribution systems (see figure on the right). Normally, microgrids operates connected to and synchronous with the traditional centralized grid (macrogrid), but can disconnect and function autonomously as physical and/or economic conditions dictate.
This project investigates how micro-generators should coordinate and collaborate to function as a bigger, more reliable and stable source of generation. Furthermore, we also investigate how should these microgrids adapt in a coordinated fashion to respond to load variabilities. more info

ewa_walle
This project searches for ways to design intelligent “sophisticated” agents that can modify an existing MAS without the need to reprogram it. These agents reason about the rational responses of its opponents (teammates) in search for good equilibrium points.

More info

Past

agriculturaUAV
This is an industrial project whose objective is to compute UAVs trajectories given human input about the objectives. UAVs gather information in the form of multispectral images and the information is then aggregated in a data center for it’s analysis. The analysis include sectioning the terrain into zones and classifying each zone into it’s most probable type.

more info

 

Trading agents

We’re interested in designing autonomous agents capable of interacting with an electronic commerce environment in order to optimize its objective function. read more…

Ongoing

We’re exploring machine learning and game theory techniques to abstract knowledge and design autonomous decision making strategies in financial markets. Some of the ideas we’re exploring include transfer learning, reinforcement learning, and apprenticeship learning. We use PowerTac as our simulation environment to test our ideas.

microGrid
We take as a microgrid the interconnection of small, modular generation (micro-turbines, fuel cells, PV, etc.), together with storage devices (flywheels, energy capacitors and batteries) and controllable loads at low voltage distribution systems (see figure on the right). Normally, microgrids operates connected to and synchronous with the traditional centralized grid (macrogrid), but can disconnect and function autonomously as physical and/or economic conditions dictate.
This project investigates how micro-generators should coordinate and collaborate to function as a bigger, more reliable and stable source of generation. Furthermore, we also investigate how should these microgrids adapt in a coordinated fashion to respond to load variabilities.