Goal of the Research Project
With the DiziPro research project – a digital platform for circular value creation in the maintenance of existing production buildings – we at thinkBIC aim to fundamentally improve the revitalisation and sustainable development of existing industrial buildings through digital technologies. Many production buildings contain valuable structural and material resources, the potential of which often cannot be fully realised due to incomplete documentation, heterogeneous data sets and complex decision-making processes. To address this, DiziPro is developing an open, AI-supported platform that captures and analyses existing building data from various sources and consolidates it into a digital twin. Using AI methods, semantic knowledge models and simulation-based assessment procedures, transparent decision-making frameworks are created to evaluate refurbishment, demolition or repurposing scenarios from both an economic and environmental perspective. In this way, the project makes an important contribution to the circular economy in the construction sector and supports companies and public sector organisations in developing existing buildings in a resource-efficient and future-proof manner.

Project Participants
DiziPro is being implemented jointly by partners from the fields of research, software development and industry. The consortium comprises the Fraunhofer Institute for Machine Tools and Forming Technology (IWU) as a consortium partner, Spinnwerk GmbH & Co. KG, Skill Software GmbH, Building Information Cloud GLWG GmbH and Munich University of Applied Sciences. The diverse expertise of the project partners enables the integration of artificial intelligence, Building Information Modelling (BIM), semantic data models and practical use cases. As thinkBIC, we are contributing our many years of expertise in the fields of BIM management, digital planning processes and AI-supported data analysis, and are supporting the development of an end-to-end, practice-oriented solution for the sustainable maintenance of production buildings. The project is funded by the Federal Ministry of Research, Technology and Space (BMFTR) and will run from February 2026 to January 2029.

Our Role in the Project
As part of DiziPro, we undertake key tasks at the interface between BIM, data management and artificial intelligence. One focus is on BIM and stakeholder management, where we define BIM use cases for the regeneration of existing buildings and develop standardised templates for Client Information Requirements (CIR) and Project Information Requirements (PIR). In addition, we produce practical guidelines to support the involvement of all relevant stakeholders throughout a building’s life cycle.

Another key focus is the AI-supported analysis of existing documentation. Building on our planBIC solution, we develop and train Large Language Models (LLMs) that can automatically recognise, structure and analyse historical building records, inspection reports and as-built plans. This makes valuable information from often unstructured documents digitally accessible and usable for further analysis.

In addition, we are responsible for extracting relevant information from BIM models and providing quality-assured IFC data for the platform’s semantic knowledge model. On this basis, we work with our project partners to develop AI-powered decision-support systems capable of assessing building stock in terms of, for example, sustainability, potential for reuse or DGNB-relevant criteria. Finally, we oversee the integration of the developed solutions into the overall platform, ensure BIM compliance and validate the results in real-world application scenarios from the perspective of different BIM roles and stakeholders.

Added Value for the Industry/Clients
DiziPro addresses key challenges in the sustainable maintenance and revitalisation of industrial buildings. By consolidating a wide variety of data sources into a digital twin, DiziPro provides, for the first time, comprehensive and reliable decision-making foundations for managing existing buildings. Companies gain transparency regarding the condition of their properties, available material resources, levels of hazardous substances, and the potential for refurbishment, conversion or demolition.

The automated digitisation and analysis of historical documents significantly reduces the effort involved in data collection and inventorying, and lays the foundation for more efficient planning and decision-making processes. At the same time, the platform enables a comprehensive comparison of various courses of action from both economic and environmental perspectives. Local authorities, industrial companies and owners of existing properties, in particular, benefit from improved tools for the regeneration of brownfield sites and the sustainable development of existing locations. In this way, DiziPro not only supports the digitalisation of the construction and property sectors, but also promotes climate protection, resource efficiency and the circular economy.

Conclusion
From thinkBIC’s perspective, DiziPro demonstrates how the combination of Building Information Modelling, artificial intelligence and semantic knowledge models opens up new prospects for the sustainable use of existing buildings. The platform developed lays the foundation for recording existing buildings on a data-driven basis, assessing them transparently and developing them further in a resource-efficient manner. Rather than a purely linear approach involving demolition and new construction, the focus shifts to the circular use of existing building stock. This creates new opportunities for the economic and environmental optimisation of building stock. The approach developed in the project can be extended to other types of buildings in the future and makes an important contribution to climate neutrality and the sustainable transformation of the construction and property sectors.