An Integrated Optimization Model Using Machine Learning for Daylighting, Thermal Comfort, and Energy Efficiency in Historic Townhouses L&E, Vol.34, No.2, 2026
Articles authors:
Nurefşan Sönmez, Arzu Cilasun Kunduraci, Cemre Ugurlu, Yonca Erkan
Nurefsan Sonmez, Joint Ph. D. Candidate, holds a master's degree in interior architecture and environmental design from Yasar University, Turkey. She is actively pursuing her Ph. D. studies in Architecture at the Yasar University since 2021, and in Heritage Studies at the University of Antwerp since 2024, with a primary research focus on the areas of daylighting optimization and energy-efficient building design in historic buildings. Her doctoral studies are financially supported by The Scientific and Technological Research Council of Turkey (TÜBİTAK)
Arzu Cılasun Kunduracı, Ph. D, holds a Doctor of Philosophy in Architecture from Izmir Institute of Technology. At present, she is an Associate Professor in Department of Architecture in Yasar University, Turkey, and is also a postdoctoral researcher at Penn State University. At Penn State, she is focusing on light pollution research using machine learning. Her expertise includes architectural lighting design, light pollution, building physics and universal design, and she is publishing extensively in these areas. She also uses computational and generative models to explore innovative uses of light in architecture
Cemre Ugurlu, Ph. D., holds a Doctor of Philosophy in Design Informatics from TU Delft, the Netherlands. She is currently an Assistant Professor in the Department of Interior Architecture and Environmental Design at Yasar University, Turkey, and participates as a researcher in a university-supported BAP project. Her research focuses on computational and AI-driven design approaches with a human-centered perspective, addressing performance criteria such as sustainability, human factors, indoor environmental quality, and resilience in high-occupancy environments including educational and healthcare facilities
Yonca Erkan, Ph. D., holds a Doctor of Philosophy in Architecture from Istanbul Technical University. At present, she has been working as a Professor of Built Heritage at the University of Antwerp since 2022. She was the UNESCO Chair holder, Management and Promotion of World Heritage Sites: 'New Media and Community Involvement' (2015–2024). Coordinated the EU Horizon 2020 MSCA – RISE Project "Sustainable Management of Industrial Heritage as a Resource for Urban Development" (2021–2025). She has research interests in world heritage, industrial heritage, and historic urban areas
Abstract:
Historic buildings pose challenges for daylighting improvements due to preservation constraints, however, improving their performance is essential for climate goals. This study seeks to improve daylighting performance, thermal comfort, and energy efficiency in historic buildings while maintaining their historic value through various design strategies, including glass upgrades, shading elements, and light tube integration, using a multi-objective optimization model.To analyse the performance of the model, two case studies from distinct climates, Mediterranean and Temperate Oceanic, were evaluated through Grasshopper plug-ins, Ladybug and Honeybee, and optimized via Octopus plug-in. A machine learning tool, Dodo plug-in was integrated to speed up the evaluation process and estimate simulation results while exploring a larger number of design alternatives.The optimization results reveal that daylighting strategies improved UDI up to 83.6 % and reduced ASE to 0 %, and improved TCP up to 6.73%, while energy use decreased by up to 8.81 % with shading and light tubes proving most effective in warm climates and roof windows in colder ones. The study confirms that performance-based design can enhance environmental efficiency without compromising buildings' historic values.
References:
1. ICOMOS. The Venice Charter-International charter for the conservation and restoration of monuments and sites, 1964. 2. ICOMOS. ICOMOS Charter – Principles for the analysis, conservation and structural restoration of architectural heritage, 2003. 3. European Commission. Communication from the commission to the European Parliament, the council, the European economic and social committee and the committee of the regions, 2020. 4. Troi, A., Bastian, Z., Eds. Energy efficiency solutions for historic buildings / A handbook, Berlin, München, Boston: Birkhäuser, 2015. DOI:10.1515/9783038216506 5. Bodart, M., A. De Herde, A. Global energy savings in offices buildings by the use of daylighting // Energy and Building, Vol. 34, # 5, pp. 421–429, 2002. DOI:10.1016/S0378-7788(01)00117-7 6. Bakmohammadi, P., Noorzai, E. Optimization of the design of the primary school classrooms in terms of energy and daylight performance considering occupants' thermal and visual comfort // Energy Reports, Vol. 6, pp. 1590–1607, 2020. DOI: 10.1016/J.EGYR.2020.06.008 7. Brzezicki, M. An evaluation of useful daylight illuminance in an office room with a light shelf and translucent ceiling at 51◦ n // Buildings, Vol. 11, # 11, 2021. DOI: 10.3390/buildings11110494 8. Lo Verso, V. R., Fregonara, E., Caffaro, F., Morisano, C., Peiretti, M.G. Daylighting as the driving force of the design process: From the results of a survey to the implementation into an advanced daylighting project // Journal of Daylighting, Vol. 1, # 1, pp. 36–55, 2014. DOI: 10.15627/jd.2014.5 9. Nasteski, V. An overview of the supervised machine learning methods // HORIZONS.B, Vol. 4, pp. 51–62, 2017. DOI: 10.20544/horizons.b.04.1.17/p05
Keywords
Number of images - 5
Tables and charts - 2
