Roberto Castello
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Roberto Castello has a total of 43 co-authors. Here are the top co-authors by number of shared publications:
Jean-Louis Scartezzini
Solar Energy and Building Physics Laboratory, EPFL, Lausanne, Switzerland
7 shared publicationsNahid Mohajeri
UCL Institute for Environmental Design and Engineering, University College London, London, United Kingdom
4 shared publications- Alina Walch
EPFL
2 shared publications - Dasaraden Mauree
WSP Ingénieurs Conseils SA
2 shared publications - Jean-Louis Scartezzini
École Polytechnique Fédérale de Lausanne
2 shared publications Adrian Guerra
Solar Energy and Building Physics Laboratory, Ecole Polytechnique Fédérale de Lausanne, CH-1015, Lausanne, Switzerland
1 shared publication
Latest Publications
- Open Access
- 12 October 2024
- Published by Elsevier in Applied Energy
- Vol. 377
Real-time estimates of Swiss electricity savings using streamed smart meter data
- 12 October 2024
- Published by Elsevier in Applied Energy
- Vol. 377
Open AccessReal-time estimates of Swiss electricity savings using streamed smart meter data
- 22 April 2022
- Published by Elsevier in Applied Energy
- Vol. 316
Large-scale evaluation of the suitability of buildings for photovoltaic integration: Case study in Greater Geneva
- 22 April 2022
- Published by Elsevier in Applied Energy
- Vol. 316
Large-scale evaluation of the suitability of buildings for photovoltaic integration: Case study in Greater Geneva
- Open Access
- 12 April 2022
- Published by Elsevier in Sustainable Cities and Society
- Vol. 82, 103896
Impact of the COVID-19 pandemic on the energy performance of residential neighborhoods and their occupancy behavior
- 12 April 2022
- Published by Elsevier in Sustainable Cities and Society
- Vol. 82, 103896
Open AccessImpact of the COVID-19 pandemic on the energy performance of residential neighborhoods and their occupancy behavior
- Open Access
- 1 November 2021
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 2042 (1) , 012027
Dimensionality reduction and clustering of time series for anomaly detection in a supermarket heating system
A timely identification of an anomalous functioning of the energy system of an industrial building would increase the efficiency and the resilience of the energy infrastructure, beside reducing the economic wastage. This work has been inspired by the need of identifying, for a series of supermarket buildings in Switzerland, the failures happening in their heating systems across the years in an unsupervised and easy-to-visualize fashion for the building managers. The lack of any a-priori label differentiating between typical and anomalous behaviors calls for the usage of unsupervised machine learning methods to extract the relevant features to describe the system operations, to reduce the dimension of the feature space, and to cluster together similar patterns of operations. The method is validated on a standard supermarket building, where it successfully discriminates winter and summer operations from periods of refurbishment or malfunctioning of the heating system.
- 1 November 2021
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 2042 (1) , 012027
Open AccessDimensionality reduction and clustering of time series for anomaly detection in a supermarket heating system
A timely identification of an anomalous functioning of the energy system of an industrial building would increase the efficiency and the resilience of the energy infrastructure, beside reducing the economic wastage. This work has been inspired by the need of identifying, for a series of supermarket buildings in Switzerland, the failures happening in their heating systems across the years in an unsupervised and easy-to-visualize fashion for the building managers. The lack of any a-priori label differentiating between typical and anomalous behaviors calls for the usage of unsupervised machine learning methods to extract the relevant features to describe the system operations, to reduce the dimension of the feature space, and to cluster together similar patterns of operations. The method is validated on a standard supermarket building, where it successfully discriminates winter and summer operations from periods of refurbishment or malfunctioning of the heating system.
- Open Access
- 10 October 2020
- Published by MDPI AG in ISPRS International Journal of Geo-Information
- Vol. 9 (10) , 593
Spatio-Temporal Relationship between Land Cover and Land Surface Temperature in Urban Areas: A Case Study in Geneva and Paris
Currently, more than half of the world’s population lives in cities, which leads to major changes in land use and land surface temperature (LST). The associated urban heat island (UHI) effects have multiple impacts on energy consumption and human health. A better understanding of how different land covers affect LST is necessary for mitigating adverse impacts, and supporting urban planning and public health management. This study explores a distance-based, a grid-based and a point-based analysis to investigate the influence of impervious surfaces, green area and waterbodies on LST, from large (distance and grid based analysis with 400 m grids) to smaller (point based analysis with 30 m grids) scale in the two mid-latitude cities of Paris and Geneva. The results at large scale confirm that the highest LST was observed in the city centers. A significantly positive correlation was observed between LST and impervious surface density. An anticorrelation between LST and green area density was observed in Paris. The spatial lag model was used to explore the spatial correlation among LST, NDBI, NDVI and MNDWI on a smaller scale. Inverse correlations between LST and NDVI and MNDWI, respectively, were observed. We conclude that waterbodies display the greatest mitigation on LST and UHI effects both on the large and smaller scale. Green areas play an important role in cooling effects on the smaller scale. An increase of evenly distributed green area and waterbodies in urban areas is suggested to lower LST and mitigate UHI effects.
- 10 October 2020
- Published by MDPI AG in ISPRS International Journal of Geo-Information
- Vol. 9 (10) , 593
Open AccessSpatio-Temporal Relationship between Land Cover and Land Surface Temperature in Urban Areas: A Case Study in Geneva and Paris
Currently, more than half of the world’s population lives in cities, which leads to major changes in land use and land surface temperature (LST). The associated urban heat island (UHI) effects have multiple impacts on energy consumption and human health. A better understanding of how different land covers affect LST is necessary for mitigating adverse impacts, and supporting urban planning and public health management. This study explores a distance-based, a grid-based and a point-based analysis to investigate the influence of impervious surfaces, green area and waterbodies on LST, from large (distance and grid based analysis with 400 m grids) to smaller (point based analysis with 30 m grids) scale in the two mid-latitude cities of Paris and Geneva. The results at large scale confirm that the highest LST was observed in the city centers. A significantly positive correlation was observed between LST and impervious surface density. An anticorrelation between LST and green area density was observed in Paris. The spatial lag model was used to explore the spatial correlation among LST, NDBI, NDVI and MNDWI on a smaller scale. Inverse correlations between LST and NDVI and MNDWI, respectively, were observed. We conclude that waterbodies display the greatest mitigation on LST and UHI effects both on the large and smaller scale. Green areas play an important role in cooling effects on the smaller scale. An increase of evenly distributed green area and waterbodies in urban areas is suggested to lower LST and mitigate UHI effects.
- Open Access
- 1 March 2020
- Published by Elsevier in Applied Energy
- Vol. 262
Big data mining for the estimation of hourly rooftop photovoltaic potential and its uncertainty
- 1 March 2020
- Published by Elsevier in Applied Energy
- Vol. 262
Open AccessBig data mining for the estimation of hourly rooftop photovoltaic potential and its uncertainty
- Open Access
- 1 November 2019
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 1343 (1) , 012034
Deep learning in the built environment: automatic detection of rooftop solar panels using Convolutional Neural Networks
Mapping the location and size of solar installations in urban areas can be a valuable input for policymakers and for investing in distributed energy infrastructures. Machine Learning techniques, combined with satellite and aerial imagery, allow to overcome the limitations of surveys and sparse databases in providing this mapping at large scale. In this paper we apply a supervised method based on convolutional neural networks to delineate rooftop solar panels and to detect their sizes by means of pixel-wise image segmentation. As input to the algorithm, we rely on high resolution aerial photos provided by the Swiss Federal Office of Topography. We explore different data augmentation and we vary network parameters in order to maximize model performance. Preliminary results show that we are able to automatically detect in test images the area of a set of solar panels at pixel level with an accuracy of about 0.94 and an Intersection over Union index of up to 0.64. The scalability of the trained model allows to predict the existing solar panels deployment at the Swiss national scale. The correlation with local environmental and socio-economic variables would allow to extract predictive models to foster future adoption of solar technology in urban areas.
- 1 November 2019
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 1343 (1) , 012034
Open AccessDeep learning in the built environment: automatic detection of rooftop solar panels using Convolutional Neural Networks
Mapping the location and size of solar installations in urban areas can be a valuable input for policymakers and for investing in distributed energy infrastructures. Machine Learning techniques, combined with satellite and aerial imagery, allow to overcome the limitations of surveys and sparse databases in providing this mapping at large scale. In this paper we apply a supervised method based on convolutional neural networks to delineate rooftop solar panels and to detect their sizes by means of pixel-wise image segmentation. As input to the algorithm, we rely on high resolution aerial photos provided by the Swiss Federal Office of Topography. We explore different data augmentation and we vary network parameters in order to maximize model performance. Preliminary results show that we are able to automatically detect in test images the area of a set of solar panels at pixel level with an accuracy of about 0.94 and an Intersection over Union index of up to 0.64. The scalability of the trained model allows to predict the existing solar panels deployment at the Swiss national scale. The correlation with local environmental and socio-economic variables would allow to extract predictive models to foster future adoption of solar technology in urban areas.
- Open Access
- 1 November 2019
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 1343 (1) , 012047
Wind profile prediction in an urban canyon: a machine learning approach
Resolving the wind profile in an urban canyon environment means dealing with the turbulent nature of the stream and the presence of non-negligible flux exchanges with the atmosphere inside the canopy, making any deterministic model solution computationally very intensive. In this paper, a learning-from-data method is explored, which is able to predict the wind speed in an urban canyon at different heights, given a minimal set of input features. The experimental location is provided by a street canyon located at the Swiss Federal Institute of Technology campus in Lausanne, equipped with several measuring stations to record data at high temporal resolution. Different machine learning approaches are compared in order to predict the wind speed in two directions and at different heights inside the urban canyon: an optimized Ridge Regression outperforms the Random Forest algorithm. We find particularly high accuracy in predicting the wind speed in the highest part of the canyon. None of the proposed algorithms however is able to model in an accurate way the variation of the wind speed close to the ground.
- 1 November 2019
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 1343 (1) , 012047
Open AccessWind profile prediction in an urban canyon: a machine learning approach
Resolving the wind profile in an urban canyon environment means dealing with the turbulent nature of the stream and the presence of non-negligible flux exchanges with the atmosphere inside the canopy, making any deterministic model solution computationally very intensive. In this paper, a learning-from-data method is explored, which is able to predict the wind speed in an urban canyon at different heights, given a minimal set of input features. The experimental location is provided by a street canyon located at the Swiss Federal Institute of Technology campus in Lausanne, equipped with several measuring stations to record data at high temporal resolution. Different machine learning approaches are compared in order to predict the wind speed in two directions and at different heights inside the urban canyon: an optimized Ridge Regression outperforms the Random Forest algorithm. We find particularly high accuracy in predicting the wind speed in the highest part of the canyon. None of the proposed algorithms however is able to model in an accurate way the variation of the wind speed close to the ground.
- Open Access
- 1 November 2019
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 1343 (1) , 012118
Geospatial analysis and optimization of the incoming and stored CO2 emissions within the EPFL campus
All the projected scenarios agree on the immediate necessity of reducing greenhouse gas emissions, especially carbon dioxide (CO2). This research proposes an analysis on the EPFL campus of all the potential reductions of CO2emissions: energy refurbishment, the request of heat and cooling of the buildings, emissions deriving from commuting and business travel as well as the embedded emissions of the buildings through their Life-Cycle Assessment. Subsequently, compensation strategies have been evaluated: Vertical Greenery Systems, CO2uptake from trees of the campus through an automated computing technique and Direct Air Capture technology. Together, all these contributions allow to significantly reduce the annual emissions of the campus, by up to 60 %, getting below the critical threshold of 1 ton of CO2/person/year.
- 1 November 2019
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 1343 (1) , 012118
Open AccessGeospatial analysis and optimization of the incoming and stored CO2 emissions within the EPFL campus
All the projected scenarios agree on the immediate necessity of reducing greenhouse gas emissions, especially carbon dioxide (CO2). This research proposes an analysis on the EPFL campus of all the potential reductions of CO2emissions: energy refurbishment, the request of heat and cooling of the buildings, emissions deriving from commuting and business travel as well as the embedded emissions of the buildings through their Life-Cycle Assessment. Subsequently, compensation strategies have been evaluated: Vertical Greenery Systems, CO2uptake from trees of the campus through an automated computing technique and Direct Air Capture technology. Together, all these contributions allow to significantly reduce the annual emissions of the campus, by up to 60 %, getting below the critical threshold of 1 ton of CO2/person/year.
- 31 August 2019
- Published by Springer Nature in Computing and Software for Big Science
- Vol. 3 (1) , 12
Topology Classification with Deep Learning to Improve Real-Time Event Selection at the LHC
- 31 August 2019
- Published by Springer Nature in Computing and Software for Big Science
- Vol. 3 (1) , 12
Topology Classification with Deep Learning to Improve Real-Time Event Selection at the LHC