Roberto Castello
Scientist
roberto.castello@epfl.ch +41 21 693 45 47
EPFL ENAC IIC LESO-PB
LE 2 204 (Bâtiment LE)
Station 18
CH-1015 Lausanne
Web site: Web site: https://leso.epfl.ch
Fields of expertise
Applied Machine Learning and Big Data analytics for the built environment and renewable energy, Urban and building physics simulation, Statistical methods for particle physics, Geographic Information System
Biography
Roberto Castello is a senior scientist and group leader at the EPFL Laboratory of Solar Energy and Building Physics. Physicist by training, he has extensive experience in collecting, classifying and interpreting large datasets using advanced data mining techniques and statistical methods. He received his MSc (2007) in Particle Physics and PhD (2010) in Physics and Astrophysics from the University of Torino. He worked as a postdoctoral researcher at the Belgian National Research Fund (2011-2014) and at the CERN Experimental Physics Department (2015-2017) as a research fellow and data scientist. He is primary author of more than 20 peer-reviewed publications and he presented at major international conferences in the high energy physics domain.
In 2018 he joined the Solar Energy and Building Physics Laboratory (LESO-PB) to work on data mining and Machine Learning techniques for the built environment and renewable energy. His main research interests are: spatio-temporal modeling of renewable energy potential, energy consumption forecasting techniques, anomaly detection, and computer vision techniques for automated classification in the built environment.
He leads the group of Urban Data Mining, Intelligence and Simulation at LESO-PB and he is a member of the NRP75 Big Data project (HyEnergy) of the Swiss National Science Foundation. He is a member of the Swiss Competence Centre for Energy Research (SCCER) and deputy leader of the working group on Leveraging Ubiquitous Energy Data. He has served as a scientific committee member, workshop organizer and speaker at international conferences (ICAE 2020, Applied Machine Learning Days 2019 and 2020, CISBAT 2019 and 2021 and SDS2020).
Since 2017 he is member of the Geneva 2030 Ecosystem network, promoting the United Nations agenda towards the realization of the Sustainable Development Goals (SDGs).
Publications
Infoscience publications
A machine learning-assisted building electricity consumption profiling for anomaly detection
2020-12-14. International Conference on Applied Energy (ICAE 2020), Virtual Conference, December 1-10, 2020.Big data mining for the estimation of hourly rooftop photovoltaic potential and its uncertainty
Applied Energy. 2020-01-28. DOI : 10.1016/j.apenergy.2019.114404.Spatio-Temporal Relationship between Land Cover and Land Surface Temperature in Urban Areas: A Case Study in Geneva and Paris
ISPRS International Journal of Geo-Information. 2020. DOI : 10.3390/ijgi9100593.A fast machine learning model for large-scale estimation of annual solar irradiation on rooftops
2020. SHC 2019/SWC 2019. ISES Solar World Congress, Santiago, Chile, November 3-7, 2019. DOI : 10.18086/swc.2019.45.12.Wind profile prediction in an urban canyon: a machine learning approach
2019-11-20. CISBAT 2019 | Climate Resilient Cities – Energy Efficiency & Renewables in the Digital Era, Lausanne, Switzerland, September 4-6, 2019. DOI : 10.1088/1742-6596/1343/1/012047.Geospatial analysis and optimization of the incoming and stored CO2 emissions within the EPFL campus
2019-11-20. CISBAT 2019 | Climate Resilient Cities – Energy Efficiency & Renewables in the Digital Era, Lausanne, Switzerland, 4-6 September 2019. DOI : 10.1088/1742-6596/1343/1/012118.Deep learning in the built environment: automatic detection of rooftop solar panels using Convolutional Neural Networks
2019-11-20. CISBAT 2019 | Climate Resilient Cities – Energy Efficiency & Renewables in the Digital Era, Lausanne, Switzerland, 4–6 September 2019. DOI : 10.1088/1742-6596/1343/1/012034.Topology classification with deep learning to improve real-time event selection at the LHC
Computing and Software for Big Science. 2019-08-31. DOI : 10.1007/s41781-019-0028-1.Spatio-temporal modelling and uncertainty estimation of hourly global solar irradiance using Extreme Learning Machines
2019-03-15. 10 th International Conference on Applied Energy ( ICAE2018), Hong Kong , China, August 22- 25, 2018. p. 6378-6383. DOI : 10.1016/j.egypro.2019.01.219.From a Vegetation Index to a Sustainable Development Goal Indicator: Forest Trend Monitoring Using Three Decades of Earth Observations across Switzerland
ISPRS International Journal of Geo-Information. 2018-11-24. DOI : 10.3390/ijgi7120455.Other publications
Selected publications
- R. Castello and CMS collaborators, Search for new physics in events with two low momentum opposite-sign leptons and missing transverse energy at 13 TeV, Physics Letters B, 2018
- R. Castello and CMS collaborators, Search for narrow resonances in dilepton mass spectra in proton-proton collisions at 13 TeV and combination with 8 TeV data, Physics Letters B, 2017
- R. Castello and CMS collaborators, Search for neutral resonances decaying into a Z boson and a pair of b jets or tau leptons, Physics Letters B, 2016
- R. Castello and CMS collaborators, Alignment of the CMS tracker with LHC and cosmic ray data, Journal of Instrumentation, 2014
- R. Castello and CMS collaborators, Measurement of the production cross sections for a Z boson and one or more b jets in pp collisions at 7 TeV, Journal of High Energy Physics, 2014
- R. Castello and CMS collaborators, Observation of a new boson at a mass of 125 GeV with the CMS experiment at the LHC, Physics Letters B, 2012 - Science Breakthrough of the Year 2012
- R. Castello and CMS collaborators, Performance of CMS muon reconstruction in pp collision events, Journal of Instrumentation, 2012
- R. Castello and CMS collaborators, Alignment of the CMS Silicon Tracker during Commissioning with Cosmic Rays, Journal of Instrumentation, 2010 (PhD thesis work)
Teaching & PhD
Courses
- AR-241 Building technology III - Module on Building Physics (Teaching assistant)
- AR-242 Building technology IV - Module on Building Physics (Teaching assistant)
- AR-496 Behind/Beyond Future Cities (Invited lecturer)
Supervised Master students
- Mr Lorenzo Donadio (EPFL and Eindhoven University of Technology), Temporal hierarchical forecasting of energy consumption in office buildings
- Mr Romain Sibuet (HES-SO Geneva), Optimization of a hybrid solar-geothermal energy system potential at building scale
- Ms Xu Ge (Technical University of Denmark - DTU), Spatio-temporal relationship between land cover and Land Surface Temperature in urban areas
- Ms Dan Chai (EPFL), Automatic detection of rooftop solar installations in the built environment and its validation on the EPFL campus
- Mr Giovanni Mori (University of Bolzano), Geospatial analysis and optimization of the incoming and stored CO2 emissions within the EPFL campus