| Product(s): | WaterSight |
| Version(s): | 10.00. |
| Area: | Documentation |
WaterSight New Updates (January 2026)
A new release has been published in January 2026, bringing the following additional functionalities:
Alerting based on patterns generated by AI
In the administration >> Alerts it is now possible to define simple and combined trigger rules using as reference a pattern that is generated by AI. It is valid only when defining triggers for zones.
This new option will appear under the "Type" settings of the trigger with the name "Pattern AI", see example below. The Pattern (P50) refers to the traditional pattern, calculated using statistical analysis.
This new AI trigger type follows the exact same logic as the typical pattern (P50) trigger. Pattern based alarms are evaluated using the pattern as a reference and alerts are only generated if i) time series values are above or below percentile 95th or 5th of the pattern (depending if high or low type is defined) for the minimum duration defined and ii) at least one of outliers has a difference relative to the percentile 50th (absolute or relative/percentage value) higher than the threshold defined. More complete information about alert triggers can be seen here - Alert Administration.
More information about patterns and how AI is used can be seen here - Patterns and demand forecast calculation.
Multi-edit for Zones
Multi-edit for zones is now available (inside Admin >> Zones). The user can now select multiple zones and make several changes, all at once.
Map Improvements
Movable time-series graphs
The time-series graphs that are displayed when clicking on an element in the map (sensor, zone, large customer, model element) can now be dragged by the user.
Zoom-in and Zoom out improvements
Zoom-in and zoom-out functions were improved. Now only when user selects an element using the side panel, the auto-zoom in function will act. If selecting from the map, there is no change of the zoom extents.
Operational Response simulations: improving the way affected customers are identified
The identification of affected customers during operational simulations (such as pipe breaks, fires, pump shutdowns or valve operations) was improved:
- Time-steps related to unbalanced time steps are now ignored when calculating affecting customers
- Low pressure customers rule was updated. Customers with low pressure are now calculated according with the following:
-
- All customers that in the baseline scenario (with no event) were above the minimum pressure threshold (defined in the administration, under the service expectations tab) and during the event scenario end up with pressure below the minimum pressure threshold, during at least one time step of the all simulation period with no pressure,
- All customers that were already below the minimum pressure in the baseline scenario (no event) but had a pressure decrease greater than 30% with the event, during at least one time step of the all simulation period (before the threshold was 15%).
More information about operational events can be seen here - Operational response.
Model reupload: improving messaging when it fails
In case a model reupload fails, a message will be displayed (in Admin >> Numerical Models) asking the user to press the rebuild button, which should solve the uploading issue. More information about the rebuild function here - Error uploading shapefiles or hydraulic model.
Reduce Sensor Delay
The time difference between the last time series instant observed in the data source (database or restAPI) and the time it takes to get displayed in WaterSight was significantly reduced, which better supports near-real time workflows of our users.
See Also
Patterns and demand forecast calculation
Error uploading shapefiles or hydraulic model
OpenFlows WaterSight TechNotes and FAQ's
WaterSight Learning Resources Guide