FEIT Research Project Database

Automatic generalisation of BIM for land administration

Project Leader: Davood Shojaei
Staff: Behnam Atazadeh
Primary Contact: Davood Shojaei (shojaeid@unimelb.edu.au)
Keywords: artificial intelligence; asset management; database systems; land administration; machine learning
Disciplines: Infrastructure Engineering
Research Centre: Centre for Spatial Data Infrastructures and Land Administration

BIM models include a large number of physical data elements and some of them, such as furniture data elements, are not important in the context of land administration. Therefore, 3D building models in IFC format should be generalised to eliminate unnecessary physical information.

From land administration perspective, there is information redundancy in architectural BIM models. Not all of the physical building elements are used for modelling spatial structure and boundaries of legal interests. Thus, another direction for future research could include developing generalisation approaches to automatically extract required building elements from BIM models and eliminate unnecessary elements.

Considering that A is legal space and B is a building element, the containment topological relationship can be used as one potential solution to eliminate unessential building elements for the purpose of urban land administration:

A Contains B >> Eliminates B from BIM