Graphs4CO2

The construction industry is responsible for a significant share of global carbon dioxide (CO₂) emissions. To reduce its climate impact, we must both use materials with a low carbon footprint and increase the reuse of building materials. However, one of the biggest challenges is that access to relevant and reliable data is limited and often difficult to apply in practice.

The Graphs4CO₂ project explores the development of an AI-driven solution to streamline the management of CO₂ data and circular materials in building design. By using Graph Machine Learning (GML), we create a smart data integration system that makes it easier for architects, engineers, and decision-makers to select sustainable materials and strategies.

The Challenge: CO₂ Data in the Construction Industry

Current databases on the climate impact and circularity of building materials are often difficult to combine, compare and integrate within BIM environments. They use different formats, measurement units, and only cover certain types of material properties. Additionally, data on reused materials is often scattered, incomplete, and difficult to align with architectural project requirements, making it complicated to choose recycled options over newly manufactured ones.

There are specialized companies that offer services to calculate the carbon footprint of buildings and maximize the use of recycled materials. However, these services are often exclusive and expensive, making it difficult for small and medium-sized enterprises (SMEs) to access the best information. This creates an inequality in the market and hinders the widespread adoption of sustainable building strategies.

The Solution: AI and Graph Machine Learning

By utilizing GML, we can develop a more intelligent and flexible solution. Instead of relying on isolated databases, Building Information Modeling (BIM) models are transformed into a graph – a network of interconnected information. This allows AI to understand and analyze the relationships between materials, building components, and climate impact in a more advanced way than traditional methods.

The proposed method, which we aim to explore in this project, consists of four steps:

1. Data Integration and Harmonization – Co lects and standardizes data from various sources, ensuring it is accurate, consistent, and useful.
2. BIM Integration – Links data to digital building models (BIM), making it easily accessible to architects and engineers in their workflows.
3. Smart Recommendations – AI suggests sustainable materials based on project needs, similar to how Spotify recommends music.
4. Interactive AI – A chatbot enables users to search for the best materials, compare alternatives, and receive instant responses.

Impacts
Through Graphs4CO₂, we can accelerate the transition to a more circular and sustainable construction industry.
Architects and construction companies gain better tools to reduce CO₂ emissions while increasing the reuse and recycling of materials.

The project enhances operational efficiency by automating the collection, standardization, and analysis of CO₂ data, improving decision-making and reducing the time required for material selection, CO₂ calculations, and resource optimization.

At the same time, it strengthens workforce empowerment by making sustainability data more accessible and user-friendly, reducing the need for technical expertise, minimizing manual labor, and freeing up time for more strategic and creative work.

For small and medium-sized enterprises (SMEs), this means fairer access to high-quality data through an open and interoperable platform, enabling them to implement sustainable solutions without significant upfront investments, thereby strengthening their competitiveness.

Conclusion

This project has the potential to revolutionize how the construction industry handles CO₂ data and circularity, both in Sweden and globally. By making the best information available to more stakeholders, we can create a fairer, more efficient, and more sustainable construction sector. With the help of AI and graph machine learning, we can transform complex and fragmented data into smart, integrated solutions that enable a more climate-friendly and resource-efficient construction sector.