Reconstruction of large fires in buildings using computational methods can assist the police investigation. In practice, investigators currently collect the data required for the elucidation of a fire by notes and video recordings at the scene of the fire. Because there is no standardized acquisition or data storage system, these notes are often very subjective and barely let in a neutral and objective assessment of the fire event later on.
A new approach is needed, which can systematically generate and evaluate the necessary scenarios for reconstruction of a fire event in order to assist the experts in their judgments. In addition this research project shall examine a systematical, automated and standardized affiliation of the data. Furthermore this will ease the digital data processing and analysis. This approach is intended to improve the quality of investigative reports and to facilitate comparison of similar cases.
In addition to this research an algorithm can be developed, which limits the possible scenarios based on the data collected. Besides due to a model-based approach, the relevant data shall be integrated into the building model to unify and simplify the creation of possible scenarios.
This process is to be implemented by the approaches of “Building Information Modeling (BIM)”. The model-based approach will also provide the opportunity, based on the definition of use of space in the building model, to standardize the creation of fire scenarios that can be implemented in the context of BIM.