Objective

Around 75 percent of buildings in Germany were built before 1978, before the first heat insulation regulations came into force. The enormous challenge of improving energy efficiency begins long before the actual renovation: housing companies and property managers need easy-to-understand action plans - and they need to implement them quickly and efficiently. The NaiS project aims to develop a digital platform for the sustainable renovation of buildings. The focus is on energy renovation measures for sustainable building operation. On the basis of NaiS, digital twins will be created in order to analyze the economic, ecological, socio-cultural and technical aspects of renovation measures. In addition, the platform will allow an objective comparison of implementation alternatives for refurbishment measures. Through efficient digitization methods, objective evaluation benchmarks and scalable renovation measures, NaiS contributes to the achievement of German and European climate and environmental goals.

In the project, the Karlsruhe Service Research Institute (KSRI) will use AI technologies to digitize and enrich information from existing buildings that is difficult to capture. The focus is on the use of computer vision approaches for the classification, localization and segmentation of objects, e.g. in point clouds (e.g. windows in facades), in photos (e.g. wall texture) or in building plans (e.g. walls, staircases, windows). The researchers will further explore how Large Language Models (LLM) can be used to navigate Germany’s complex public construction laws, building codes, and related documents. On the research side, novel approaches will be developed in the areas of Human-AI Collaboration (i.e., “How can humans and AI work together effectively?“) and Data-Centric AI (i.e., “How do I train AI models with limited data?“).

Funding

The joint project NaiS is funded by the Bundesministerium für Wirtschaft und Klimaschutz (BMWK) within the program “Green Tech”.

Duration

01.05.2023 – 30.04.2026

Website

https://nais.tech

Partners

Involved DSI Researchers

• Dr.-Ing. Michael Vössing (Project Lead)
• Patrick Bigge
• Patrick Hemmer
• Sebastian Schäfer

News