Leveraging Generative AI to Learn Impact of Climate Change on Buildings on Urban Areas

Abstract

Climate change, global warming, and pollution are increasing daily. Population growth demands more buildings, so understanding the environmental impact on buildings and vice versa is essential. We must know what the environmental effects on buildings are and how buildings contribute to environmental changes. Most research focuses on building monitoring using Internet of Things (IoT), such as energy consumption, data collection, etc., but overlooks the outdoor environmental impacts on buildings and vice versa. Some studies examine the impact of global warming on residential buildings in specific areas, but they often lack comprehensive reports explaining the results. This work aims to find and understand the environmental effects on buildings, indoor environments, and residents. It also seeks to generate reports of these impacts, providing recommendations to mitigate and minimize them by using Generative Artificial Intelligence. We fine-tuned Large Language Models (LLMs) such as Generative Pre-trained Transformer 2 (GPT-2) and Large Language Model Meta AI 2 (LLAMA2-7b), specifically using NousResearch LLAMA2-7b-hf version from Hugging Face to generate the reports based on the outdoor environment values i.e. temperature, humidity, and air index. The generated reports define the environmental impacts on buildings and recommendations to minimize them. These reports can be used for effective urbanization planning and sustainable development. By following the recommendations or best practices we can maintain our indoor environment with reduced contributions to global warming. Future work involves monitoring buildings’ indoor environments, energy consumption, and greenhouse gas (GHG) emissions to reduce GHG emissions further and address global warming.