Reality Data is a high-fidelity digital representation of an infrastructure or site at a specific moment. It provides digital context to design projects, continuous monitoring for construction projects, and multiple snapshots of an asset’s as-built conditions during its operation. Reality Data is also a practical medium from which insights are extracted throughout an asset’s lifecycle.
In Digital Twins, Reality Data are seldom used alone. They are typically visualized in conjunction with various other geolocated data, including BIM models, GIS layers, readings from IoT sensors, inspection artifacts, and more. The virtue of a Digital Twin is achieved from the synergy of all the information it contains.
Unlike the Earth, the computer monitors and mobile devices we use to experience Digital Twins are (mostly) flat. The Digital Twin software performs mathematical transformations, called reprojection, to spatially align datasets in the application’s viewport. Coordinate Reference Systems (CRS) define the parameters used in those transformations that accurately calculate the relation between the position of pixels on a screen with their location in the real world.
There are many CRSs in existence. Each has its own purpose, strengths, and drawbacks, and those nuances are admittedly difficult to master. It is therefore recommended to consult a domain expert when selecting a CRS for Reality Data.
This situation happens when a Reality Data delivery contains no information about its CRS. If such Reality Data are published as-is to a Digital Twin, they will be positioned in the wrong location, far away from where they are expected to display.
Like how a GPS navigation system needs a starting point and a destination to calculate a route, iTwin software needs a source CRS (defined in the Reality Data) and a destination CRS (defined in the iModel) to reproject and display Reality Data in the right spatial location.
Image 1 - illustration of data published in a wrong location
This situation happens when Reality Data deliverables are processed in a Local Coordinate System. Such Reality Data are registered against landmarks that are positioned on the site and unrelated to the CRS that are known by the Digital Twin software.
Spatial alignment of such Reality Data can only be achieved when all data in the Digital Twin are created in the same Local Coordinate System.
Image 2 - illustration of reality data registered against landmarks
This situation happens when the Coordinate Reference System of a Reality Data delivery is incorrect. In most cases, the incorrect CRS is a variation of the right one where one parameter is inaccurate. When this happens, the display of the Reality Data will typically be shifted horizontally or vertically.
A horizontal shift is typically a symptom of an inaccurate definition of units in the Reality Data. To confirm, measure the location of a known point on the Reality Data, divide its X or Y coordinate with the corresponding coordinate of the known point, and compare the result with the conversion factors in the table below.
A vertical shift is typically a symptom of a missing or inaccurate definition of the Vertical Coordinate System (VCS) component of the Reality Data’s CRS. A VCS is a reference from which elevations are measured to accurately represent topography.
All VCS can be classified into two categories: Ellipsoid and Geoid. Data collected from satellite navigation systems, such as GPS, is measured against an ellipsoidal VCS. However, the most common types of VCS used to visualize 3D Digital Twins are gravity-based surfaces representing mean sea level (geoids).
The distance between an ellipsoid surface and the topography is called ellipsoid height, the distance between a geoid surface and the topography is called orthometric height, and the difference between an ellipsoid and a geoid surface is called geoid separation.
Vertical shifts are typically equal to the geoid separation, and they happen when:
Image 3 - Vertical Coordinate Systems
The following Wiki articles describe the best practices for achieving perfect geospatial alignment for Reality Data.