Development of infrastructure information models according to open standards and automatic parametrization from geomatic data

Resumo

The complexity, extension and deterioration through time and events of transport infrastructure calls for cost-effective technologies for their monitoring. The digitalization of these assets into information models is a fitting approach for monitoring, planning, and analysing and predicting behaviour. Nonetheless, the obtention of the infrastructure information models needs to be as automatized as possible, while also following existing standards in order to be interoperable with the different technologies and software tools. In this context, this thesis proposes a set of different methodologies for the automatic generation of infrastructure information models from segmented point cloud data following the IFC (Industry Foundation Classes) schema. It its heavily focused on both the alignment-based positioning system underlying these models, and the geometric definition of the elements that form the asset. Nonetheless, it also pays attention to the available semantics and further processing that might be of interest, such as structural analysis. In essence, the proposed methodologies build on top of each other to increase the details and amount of information that the model represents, from simple curves to complex geometries and relationships. Ideally, the obtained models should be introduced in a BIM (Building Information Model) workflow where they can be utilized for the aforementioned applications, such as cost-analysis. They should also be constantly updated through time to always reflect the current state of the asset and serve as a unique source of truth for the interested parties. The methodologies and algorithms presented in this dissertation have been tested in different real scenarios of road and railway environments (including truss bridge assets). This, in turn, resulted in four publications in high impact, peer reviewed, international journals indexed on the Journal Citation Report (JCR) and one conference paper. Through these publications, the presented work advanced the state of the art and contributed to the knowledge in its respective field.