The city or the objects that are part of it (such as buildings) now have real digital twins that are increasingly precise in their representation, whether geometric and/or semantic. These data play an important role in understanding the city and its evolution.

Despite the large number of objects represented (for example, a few gigabytes of data to represent a building in 3D and the objects that make up its environment, such as neighbouring buildings, infrastructures, terrain, vegetation, etc.), it is now possible to obtain a 3D visualisation in a simple browser, thus offering playful representations that allow you to move around a scene as if you were in a video game.

Beyond performance, this representation, missing semantic information, is inappropriate for technical management applications of the city and its heritage. The semantics to which we refer must complete the knowledge and make it possible to document the city, all the buildings on a site or the objects that makes them up. It is also necessary to be able to move between representations at different scales, from the territory to the interior of a building, or even to the finest level of the technical plan (CAD) of the equipment.

The objective of this thesis is to offer users the possibility to query the urban maintenance dataset in order to visualise the equipment and its life cycle, from the city scale to the building scale and its interior. Users must also be able to interact with this equipment, by displaying information available from different data sources, but also by carrying out classical maintenance management operations : asking for an intervention, report a disorder ...

Solutions to solve scientific challenges arising when visualizing urban and technical objects enriched with existing semantics (2D+Time, 3D, 3D+Time, 3D+semantics, etc.), using  multiple views defined by the user according to their needs, should be defined. As all the information comes from different production sources, it will also be necessary to propose a method for integrating the various equipment representations, as well as information concerning their maintenance. Finally, in order to respond to maintenance business queries, it will be necessary to offer an approach for identifying the representations of the equipment concerned and their associated maintenance data. Interaction with users throughout this research will be important in order to make the proposed approaches relevant in an industrial context.