José Ramón Vázquez-Canteli
I’m a PhD student at the University of Texas at Austin. I have a multidisciplinary background which includes electrical engineering, energy management and machine learning. I did my master thesis at École Polytechnique Fédérale de Lausanne (EPFL), Switzerland, and I graduated from the Universidad Pontificia Comillas in 2016, with a M.S. in Industrial Engineering (Electrical Engineering focus).
My current topic of research is the use of Deep Learning to make smart buildings learn from diverse factors such as occupants’ behavior, weather conditions, energy prices, and from each other. My goal is to develop and simulate an AI-algorithm that allows buildings to learn and coordinate among each other, so they avoid consuming energy simultaneously (demand response). Other projects I am working on include:
- Developing new ways of forecasting energy demand by encoding energy profiles into images, and using computer vision (tiled convolutional neural networks in TensorFlow) to make the predictions.
- Modelling human temperature preferences in the built environment using ecobee thermostat data and implementing recurrent neural networks in Keras.
- Building energy simulation at the urban scale for the city of Austin, and development of population growth scenarios and integration of renewable energy technologies.
Vázquez-Canteli, J., Ulyanin, S., Kämpf J., and Nagy, Z., “Adaptive Multi-Agent Control of HVAC Systems for Residential Demand Response Using Batch Reinforcement Learning”, (under review)
Vázquez-Canteli, J., Ulyanin, S., Kämpf J., and Nagy, Z., “CityLearn: Fusing Deep Reinforcement Learning with Urban Energy Modeling for Adaptive Building Energy Control in Smart Cities”, (under review)
Vázquez-Canteli, J., and Nagy, Z., “Reinforcement Learning for Demand Response: A Review of algorithms and modeling techniques”, Submitted on March 2018 (under review)
Nagy, Z., Park, J.Y., and Vázquez-Canteli, J., “Reinforcement learning for intelligent environments: A Tutorial”, in Press
Nagy, Z., Vázquez-Canteli, J., and Park, J.Y., “Using Bluetooth Based Occupancy Estimation for HVAC Set-Back to Reduce Energy Consumption in Buildings”, ASHRAE Conference, Houston, May 2018
Vázquez-Canteli J., Kämpf J., and Nagy, Z., “Balancing comfort and energy consumption of a heat pump using batch reinforcement learning with fitted Q-iteration”, CISBAT, Lausanne, 2017 (best paper award)
Nagy, Z., Park, J.Y., and Vázquez-Canteli, J., “Reinforcement Learning for smart buildings and cities” Passive and Low Energy Architecture (PLEA), Edinburgh, 2017
Vázquez-Canteli J., Kämpf J. Energy simulation at the urban scale: a focus on Geneva and climate change scenarios, Sustainable City 2016; WIT Transactions on Ecology and The Environment, Vol 204.
Awards & Recognitions
2018 Kolodzey travel grant, The University of Texas at Austin
2017 BEST PAPER AWARD – CISBAT 2017 International conference, École Polytechnique Fédérale de Lausanne (EPFL), Switzerland
2017 Award of British Petroleum Chair on Energy and Sustainable Development. One of the twelve best projects at Universidad Pontificia Comillas: “Massive 3D Models and Physical Data for building simulation at the urban scale: a focus on Geneva and climate change scenarios” [link, in Spanish]
2010 Excellence Scholarship of the Community of Madrid