BEYOND: AI-based optimization in early design phases of buildings using crowd simulation in intelligent building models
Important design decisions are made in early phases of the design process. So far, however, the data basis for such decisions is limited in this phase, which often leads to wrong decisions and higher costs. In order to improve this decision basis, the BEYOND research project has started: it aims to use new BIM-based solutions in combination with artificial intelligence (AI) and crowd simulation. The current project BEYOND is funded by the Federal Ministry of Transport and Digital Infrastructure (BMVI) within the framework of the mFUND with a total of 905,170 Euros.
The construction industry is currently experiencing a boost towards digital transformation. This is particularly noticeable in complex buildings, such as infrastructural buildings. The planning and construction period of such projects is very long and the implementation oftentimes is carried out during operations. This often leads to delays and cost increases. This can be counteracted with the help of new, digital applications.
This is why the project consortium of BEYOND has formed. BEYOND will investigate how the planning of large construction projects in the infrastructure sector can be carried out more cost-effectively, faster, with higher quality and lower error rates by combining Building Information Modeling (BIM), Artificial Intelligence (AI) and crowd simulation. To this end, methods and technologies are being developed to support architects and engineers in decision-making in the early phases of planning or redesigning train stations. By analyzing the movement of people during normal operation as well as in case of evacuation, planning reliability of such buildings and the safety in stations can be significantly increased. The feasibility has already been confirmed in the predecessor project AHEAD. This will now be continued in BEYOND with an extended consortium.
accu:rate GmbH, which is a leading expert in the field of crowd simulation, was able to connect suitable project partners for this: the Chair of Computational Modeling and Simulation at the TU Munich, headed by Prof. Borrmann, who plays a leading role in BIM research; DB Netz AG with its major project 2. Stammstrecke München, who are submitting the reference project and providing their input from the point of view of the client; and SSF Ingenieure AG, specialized in BIM, who as expert planners are contributing their practical view to the project and testing the developed methods for applicability.
Based on an exemplary station model, data structures, algorithms and technical procedures for the fast check of different variants of the BIM models are developed in close cooperation with the project partners. This interdisciplinary work and close integration with practical applications is ought to ensure the later use of the models. Therefore, regular meetings with national and international BIM experts will take place during the project in order to map both national and international requirements. A standardized data interface is to guarantee a manufacturer-independent usability from the technical side.
With BEYOND, variants of building designs can then be examined more quickly for safety aspects, decision-making can be supported more sustainably and thus better solutions can be found.