Determining the geographical location of assets with a Workbench
There is a desire at Utilities to store the asset data of home connections geographically within the existing GIS system. For example, for the purpose of trace functionality within the Esri ArcGIS platform, to determine which addresses are connected to a certain asset or area in the event of a malfunction.
These asset data therefore often relate to the home connection and the geographic relationship with the distribution network. This usually involves a number of data points relating to the objects — such as connecting pipes or (home) connecting cables, but also the service faucets or the connection sleeves.
In some cases, the position of the various objects is not recorded with x, y coordinates, but recorded in a notation form — often in combination with the connection sketch (these too can be converted to vector data using FME, but that is beyond the scope of this blog post).
This notation form can, for example, be included in location dimensions. These dimensions can refer to where the pipe or cable enters the building in relation to the façade, viewed from the side of the building or the intermediate façade in combination with the distance to the distribution network.
These objects can be provided with x, y coordinates using FME through, for example, vectoral calculations based on the dataset and the geometry of the building.
CHOSEN PATH OF SOLUTION
The distribution network is searched for (which meets a number of conditions such as diameter, material and network accuracy), as well as the point where the projected connection crosses the distribution network. This is the theoretical point of, for example, the service valve or the connection sleeve. The connection line is drawn in a straight line from the starting point.
The location of the connection can be determined theoretically from a point in the home (address or building), but the exact location is not. This could be better approximated by adjusting the XY value with the help of the connection location dimensions.
Above, a schematic representation of the location dimensions.
The algorithm for calculating the home connection assets based on FME Workbench is developed by Tensing. By choosing FME, the data-enrichment process is system-independent and can be added to multiple GIS systems.
The algorithm uses data from different sources, namely:
Notation of the home connection data (multiple notations are possible);
The geographical and asset data of the distribution network from the GIS system to be enriched (for linking the home connections to the distribution network);
BAG data from the Land Registry (property and address points for the calculation of the location dimensions from the correct facade and to determine which property belonged to the connection address);
The GBKN for any deviating building characteristics (in some cases, the BAG supplies the contour of the entire block of houses, the intermediate walls are in the GBKN);
The ACN file for any addition to address points (it is possible that the BAG data cannot be completely matched with the connection addresses);
National road database (NWB) for determining the facade (in this, it is assumed that the corresponding street name of the connection address and the NWB determine the location of the facade).
FUNCTIONING OF THE ALGORITHMS
Subsequently, the following items were executed with FME Workbench algorithms:
Above, one of the FME Workspaces for calculating the point of entry of the connection into the building and the perpendicular angle.
It is determined which BAG building belongs to the corresponding address of the connection;
Subsequently, the facade of the building was determined using the National Road Database based on street names. "Direction of rotation" of the geometry has been converted so that it rotates to the left (normal vector> 0);
So, head and tail change as the normal vector <0. And the building geometries are generalized and then converted to line geometries;
All convex corners of the correct façade line (<180 degrees) are reviewed whther the dimensioned length here corresponds to the length of the facade;
Subsequently, for each of these calculated angles, the calculated location is point of entry into the building (with a shift x, y over the facade);
Next, it was determined where the distribution network is closest to the calculated location (material and diameter are taken into account);
Per facade is using the normal vector where the connection enters the building perpendicularly.
To determine the location of the connections, a custom transformer has been developed within FME Workbench called the VECTOR_SETTER.
Above, a small extract of the VECTOR_SETTER workspace
FME Datainspector intermediate result
The data sets used should be up-to-date to achieve the desired results, and BAG compliance of the data sets is a prerequisite for achieving good results.
With FME it is possible to achieve these things, but if the source data are not consistent, the expected result may turn out differently.