Available for an ArcGIS organization licensed with the IPS extension.
Assessing the quality of an ArcGIS IPS deployment is important to the user experience in your facility. The quality assessment workflow provides insights into how IPS is performing in different areas of your environment. You can run this assessment for several positioning files. The results allow you to compare the performance of each of the positioning files you have created.
After quality assessment, you can go back on-site, perform new survey recordings for that area, and generate a new positioning file. With the new positioning file in place, you can perform a second assessment and compare the result with the previous one to verify improved quality.
IPS quality recording
The first step in the quality assessment workflow is to collect quality recordings. This procedure is similar to the collection of survey recordings, but a positioning file must already be in place. Follow the instructions to collect quality recordings for iOS or Android devices.
As you perform a quality recording with the ArcGIS IPS Setup app, the active positioning file is used by the IPS Engine to compute your real-time position, and this information is stored with the recording data.
Quality measures are generated along the trajectory of the quality recordings. It is therefore recommended that you perform as many quality recordings as necessary to properly cover the area.
Create an IPS Quality Dataset
The IPS Quality Dataset contains the following feature classes necessary to store a quality assessment:
- Computed Positions—Indoor positions computed by the IPS Engine while performing a quality recording.
- Reference Positions—Actual positions of the surveyor while performing a quality recording.
You can generate an IPS Quality Dataset using theCreate IPS Quality Dataset tool in ArcGIS Pro.
Complete the following steps to create an IPS Quality Dataset in ArcGIS Pro:
- Open a map in ArcGIS Pro.
- On the Analysis tab, click Tools
.The Geoprocessing pane appears.
- In the Geoprocessing pane, click the Toolboxes tab.
- Expand Indoor Positioning Tools and expand Quality Assessment.
- Click the Create IPS Quality Dataset tool.
The Create IPS Quality tool appears.
- Browse and choose the workspace where you want to create the quality dataset.
Note:
This tool honors the Allow geoprocessing tools to overwrite existing datasets option. If overwrite is enabled and the output quality dataset already exists in the target workspace, it is overwritten.
- Optionally, specify a coordinate system for the Spatial Reference parameter.
Note:
If no coordinate system is specified, the quality dataset will be created in WGS84.
- In the Dataset Name parameter, choose a name for the quality dataset.
- Click Run.
A feature dataset with the chosen coordinate system and name is created in the target workspace that contains Reference Positions and Computed Positions feature classes.
You are now ready to compute the positioning quality from the collected quality recordings.
Compute IPS positioning quality
You can compute the positioning quality for a set of quality recordings by running the Compute Positioning Quality tool.
- Open the map in ArcGIS Pro.
- On the Analysis tab, click Tools .
The Geoprocessing pane appears.
- In the Geoprocessing pane, click the Toolboxes tab.
- Expand Indoor Positioning Tools and expand Quality Assessment.
- Click Compute Positioning Quality tool.
The Compute Positioning Quality tool appears.
- For Target IPS Quality Dataset, select the feature dataset generated previously.
- For IPS Recordings Features, select a feature layer containing the quality recordings.
- Optionally, provide values for Sight Blocking Details Features.
If not provided, the line-of-sight between reference and computed positions will not be computed.
- Click Run.
This tool extracts from each recording the real-time positions computed by the IPS Engine as you were performing the recording and the corresponding actual positions of the surveyor at that same time. These are stored in the Computed Positions and Reference Positions feature classes, respectively. Each computed position is associated with a unique reference position by recording ID and time.
In every positioning file, each point in Reference Positions has associated information about the quality of the positioning performance at the specific location. This information can be used to compare a reference position with the corresponding computed position and the obstacles present in the environment.
The positioning performance is estimated according to the following metrics:
- Level Match—A boolean value denoting if the reference and computed positions are on the same level.
- LOS to Computed—A boolean value denoting if the reference position is visible from the corresponding computed position. In case of a floor mismatch, this is set to FALSE, otherwise it is the result of a line-of-sight calculation, where given Sight Blocking Details Features are considered as obstacles. If no sight obstruction details are given, the line-of-sight is not computed, and this field is set to NULL.
- Distance to Computed—The distance between reference and computed positions. In case of a floor mismatch, this is set to NULL, otherwise it represents the Euclidean distance between a computed position and its corresponding reference position.
- Level of Accuracy—A simple classification of the positioning accuracy based on the Distance to Computed field. In case of a floor mismatch, this evaluates to Invalid, otherwise it takes one of the following values: Low, Medium, or High.
Note:
The computed values are encoded in the DOM_IPS_QA_ACCURACY_LEVEL domain that is created together with the quality dataset by the Create IPS Quality Dataset tool.
The possible combinations of the Level Match, Line of Sight (LOS) to Computed, and Level of Accuracy fields provide a qualitative assessment of your deployment. Note that not all combinations of values are possible:
| Level Match | LOS to Computed | Level of Accuracy |
|---|---|---|
TRUE | TRUE | High |
TRUE | TRUE | Medium |
TRUE | TRUE | Low |
TRUE | FALSE | High |
TRUE | FALSE | Medium |
TRUE | FALSE | Low |
FALSE | FALSE | Invalid |
As mentioned, the accuracy level is a classification of the positioning accuracy based on the Distance to Computed field and the following threshold values.
| Distance to Computed | Level of Accuracy |
|---|---|
<5 m | High |
5 m - 10 m | Medium |
>10 m | Low |
The reference positions and computed positions features can be used to assess your IPS deployment.