NaiS

Sustainable, intelligent renovation measures

Renovation measures offer tremendous potential when it comes to increasing sustainability in the construction industry. In 2019, the construction and operation of buildings were responsible for 38 percent of global CO₂ emissions. Of this, 10 percent was attributable to production and 28 percent to operation. As around 75 percent of the buildings were constructed before 1978 – and therefore before the first Thermal Insulation Ordinance came into effect – the need to renovate existing buildings is acute. However, digital plans and information on the renovation status, the building materials and the proportion of grey emissions generated during the production of these materials are frequently lacking. The digital tools currently available on the market, for example for evaluating buildings, do not allow users to perform intelligent life cycle analyses. This makes it impossible to fully assess emissions and optimise renovation measures. Furthermore, existing analysis options do not allow forecasts to be made, for example on how CO2-intensive the energy mix is or on energy costs. These challenges have resulted in a significant amount of manual effort for experts, specialists and reviewers.

The NaiS research project has a solution approach that is focused on residential and office buildings. The first step is to digitise information that is available as PDF or image files, for example, and is not machine-readable. For instance, floor plans or energy performance certificates are prepared using machine learning models and converted into the ‘Industry Foundation Classes’ (IFC) standard. The information is then checked for completeness and enriched with additional digital data. The next step is to use a platform with open-source standards and interfaces: This will make it possible to enrich the existing digitalised buildings with sustainability-related information from external data platforms. There are also intelligent analysis tools for sustainable renovation: Once everything has been digitalised, the information is used to assess the existing buildings and the need for renovation or refurbishment. This makes it possible to compare alternatives and prepare recommendations, e.g. in the form of a renovation roadmap. The results can also be used as a basis for a digital building resource passport and for the new energy performance certificate.

The research team is committed to taking a human-centred approach with its developments: For example, a collaboration partnership between humans and machines (MMK) will be used to intuitively supplement any missing information, including any data on the suitability and type of materials. Suggestions made by the AI are then evaluated with the help of human expertise. This feedback forms the basis of a learning and intelligent system for sustainable renovation measures.

The project team has specific partners and application scenarios in mind for transferring the research project to the market: The German Sustainable Building Council – DGNB e. V. will expand its current product portfolio to include the newly developed open-source interfaces – as a neutral provider and central point of contact for objective KPIs that will be used to assess sustainability in the construction industry. The project partners CAALA GmbH and Concular GmbH will also integrate the new digital tools into existing software solutions. And the pilot projects at Ed. Züblin AG are expected to have a major external impact in the construction industry. Therefore, this project may also indirectly contribute to strengthening the competitiveness of German construction companies and Germany as a centre for AI technology.

Consortium:

• Karlsruhe Institute of Technology: Institute for Technology and Management in Construction (TMB; consortium leader); Karlsruhe Service Research and Innovation Hub (KSRI)
• German Sustainable Building Council (DGNB) e. V.
• Karlsruhe Institute of Technology (major research task): Institute for Automation and Applied Computer Science (IAI)
• Ed. Züblin AG
• CAALA GmbH, Concular GmbH, Hof University of Applied Sciences (subcontracted)