Residential Construction Carbon Reduction Using Multilayer Perceptron and Renewable Energy Integration
多層パーセプトロンと再生可能エネルギー統合を用いた住宅建設の炭素削減 (AI 翻訳)
Dr. Ayalapogu Ratna Raju, T. Akila, Dr.K Umapathy, S. Sujatha, Dr. R. Jagadeesh Kannan, Sam Zenith
🤖 gxceed AI 要約
日本語
住宅建設のライフサイクル全体の炭素排出削減を予測するため、多層パーセプトロン（MLP）を用いたデータ中心の枠組みを提案。1,640件の住宅プロジェクトデータを分析し、排出削減レベルを高・中・低に分類。MLPは99.76%の分類精度を達成し、再生可能エネルギー統合と材料再利用が主要な削減要因であることを感度分析で示した。低炭素意思決定を支援するMLPモデルの有効性を実証。
English
This study proposes a data-centric framework using a Multilayer Perceptron (MLP) to predict carbon emission reduction potential across the lifecycle of residential buildings. Analyzing 1,640 projects, the model classifies reduction levels with 99.76% accuracy, identifying renewable energy integration and material reuse as key drivers. The findings validate MLP-based modeling for low-carbon decision-making in residential construction.
Unofficial AI-generated summary based on the public title and abstract. Not an official translation.
📝 gxceed 編集解説 — Why this matters
日本のGX文脈において
日本の住宅建設業界では、脱炭素化に向けた建材のライフサイクル評価や再生可能エネルギー導入が重要課題。本手法は、SSBJ開示やサプライチェーン排出量算定に応用可能で、住宅事業者の低炭素投資判断を支援する。
In the global GX context
This research aligns with global efforts to decarbonize the building sector, a key focus of TCFD and ISSB disclosures. The MLP-based framework offers a scalable approach for assessing supply chain emissions and renewable energy integration, relevant for companies reporting under CSRD and SEC climate rules.
👥 読者別の含意
🔬研究者:Provides a novel application of MLP for lifecycle carbon prediction in residential construction, with high accuracy and sensitivity analysis.
🏢実務担当者:Offers a data-driven tool for housing developers to prioritize carbon reduction measures, such as renewable energy and material reuse.
🏛政策担当者:Demonstrates the potential of AI in supporting building decarbonization policies and could inform regulatory frameworks for emissions reporting.
📄 Abstract（原文）
Residential construction substantially contributes to global carbon emissions via material manufacture, intricate supply networks, and building operations. Precise forecasting of emission reduction potential across lifetime phases is crucial for sustainable house building. This research introduces a datacentric framework for analyzing lifetime and supply chain carbon emissions in residential building via a Multilayer Perceptron (MLP) neural network. A publicly accessible Kaggle dataset consisting of $\mathbf{1, 6 4 0}$ residential building projects was used, including embodied and operational emissions, supply chain hotspots, and sustainability metrics. Projects were categorized into Low, Medium, and High emission reduction tiers according to their realized carbon mitigation results. The proposed MLP model was trained using lifetime emission measurements, material efficiency indicators, renewable energy proportions, and policy intervention factors. Experimental findings indicate that the MLP achieved an overall classification accuracy of 99.76%, with precision, recall, and F1-score metrics of $99.64 \%, 99.57 {\%}$, and 99.60%, respectively. In comparison to conventional machine learning models, the MLP demonstrated enhanced efficacy in identifying nonlinear correlations among supply chain and lifespan characteristics. Sensitivity analysis indicated that the integration of renewable energy and material reuse rates were the primary contributors to the categorization of substantial emission reductions. The findings validate the efficacy of MLP-based modelling in facilitating low-carbon decision-making for residential building projects.
🔗 Provenance — このレコードを発見したソース
- semanticscholar https://doi.org/10.1109/icctwc68241.2026.11583604first seen 2026-07-13 06:31:17 · last seen 2026-08-02 06:11:53
🔔 こうした論文の新着を逃したくない方は キーワードアラート に登録（無料・3キーワードまで）。
gxceed は公開メタデータに基づく研究支援データセットです。要約・翻訳・解説は AI 支援で生成されています。 最終的な解釈・検証は利用者が原典資料に基づいて行うことを前提とします。