Session: 02-02: Advances in Green Energy Modeling and Innovative Technologies

Paper Number: 142224

142224 - Advancing Building-Integrated Photovoltaics (Bipv) Integration in Architecture: A Machine Learning-Based Revit Plugin for Sustainable Design 

Abstract: 

The building sector significantly contributes to global energy consumption and greenhouse gas (GHG) emissions, making it a pivotal area for climate change mitigation efforts. Addressing energy consumption and GHG emissions within this building sector is significant for achieving global climate change objectives. Building-integrated photovoltaics (BIPV) is one of the practical means to advance sustainable architectural design by generating on-site clean energy and thus reducing its carbon footprints. However, the current challenges in BIPV integration include a lack of comprehensive libraries and information, making it difficult for architects to collect data and apply it effectively to their designs. Despite its advancement, the challenge lies in effectively selecting the most suitable BIPV solutions for diverse building elements while maintaining the integrity and aesthetic of the original architectural design as much as possible. This study introduces a machine learning-based approach, incorporating a newly developed plugin module within the Building Information Modeling (BIM) framework to streamline the selection of BIPV systems. By leveraging a comprehensive dataset of BIPV and BIM information, this method utilizes BIM data to estimate the building's approximate energy consumption. It then calculates energy usage across all BIPV library options and determines energy coverage percentages and costs. This process significantly enhances decision-making for architects. Additionally, the algorithm incorporates climate data and a BIPV library to enhance the precision of its recommendations. By analyzing the relationship between building geometries within the modeling environment and their respective BIPV performance metrics, a machine learning model is trained to identify patterns and predict optimal BIPV configurations. This strategy facilitates a detailed analysis of environmental factors, guaranteeing that the BIPV recommendations not only complement the architectural design but also enhance energy efficiency and sustainability to the fullest extent. The integration of these datasets and the application of machine learning techniques facilitate a deeper understanding of the complex interplay between architectural design, solar potential, and climate conditions, leading to more informed and context-specific BIPV recommendations. This method not only enhances the precision of BIPV system selection but also aligns with unique design and environmental prerequisites. The algorithm's effectiveness in improving the accuracy and efficiency of selecting BIPV systems is well-documented. Its integration with BIM tools significantly streamlines the architectural design process, facilitating easier adoption and more precise implementation of BIPV solutions, thereby offering a more user-friendly experience for architects. This algorithm enables them to incorporate sustainable BIPV elements into their designs effectively. Overall, the development of this machine learning algorithm, supported by an extensive BIPV and BIM dataset, marks a significant advancement in sustainable architectural design. This approach offers a sophisticated, data-driven method for selecting BIPV systems, providing more sustainable construction practices. Moreover, by improving the accessibility and precision of BIPV system integration, this research significantly contributes to the broader adoption of renewable energy technologies in the building sector, aligning with global efforts to combat climate change.

Presenting Author: Dawon Lee Junglim Architecture

Presenting Author Biography: Dawon Lee
Computational Designer
Junglim Architecture

Biography:
Dawon Lee is a Computational Designer at Junglim Architecture, focusing on sustainable architectural solutions. With a Bachelor's degree in Architecture from Kyunghee University, Dawon has dedicated four years to integrating sustainability into architectural design, emphasizing machine learning applications in Building Information Modeling (BIM).

A key aspect of Dawon's expertise is his use of computational design to enhance sustainable architectural practices. He has notably contributed to the research, “Parametric Design Study of a Proposed Photobioreactor-Integrated Vertical Louver System for Energy-Efficient Buildings,” published in the Journal of Green Building. This work underscores his dedication to sustainable design, utilizing genetic algorithms for design optimization.

Dawon is excited to present his insights on "Advancing Building-Integrated Photovoltaics (BIPV) Integration in Architecture: A Machine Learning-Based Revit Plugin for Sustainable Design" at the ES 2024 : 18th International Conference on Energy Sustainability. His presentation will focus on innovative, energy-efficient building solutions that underscore his dedication to sustainable design.

Connect with Dawon :
Email: dawon2100@gmail.com
LinkedIn: www.linkedin.com/in/dawon-lee-7395371a9

Authors:

Dawon Lee Junglim Architecture
Hyojung Kim Junglim Architecture
Hyunah Kim Junglim Architecture
Kyungwoo Lee Junglim Architecture
Sungwoo An Junglim Architecture
Dongsu Kim Department of Architectural Engineering, Hanbat National University