Skip to content

Feature Description

The Model-Based Signals allow you to mark a model to provide model-based signals independently of a specific machine.
This allows users to see Model-Based Signals in the DataPortal and understand the efficiency of each machine-model even without owning a machine of that model.
The metrics are also available via the REST API Timeseries, and can be used to compare a machine's consumption with the model fleet's averages or benchmark own machines with other machines from the same model.
The signals defined by the OEM for a model are managed via the PDC Manual the same as machine signals.
Contact your Proemion representative for more information.
The Model-Based Signals feature is an included DataPortal feature.

Use Case

A sample use case is the analysis of "Fuel Consumption":

The Model-Based Signals measure the performance (i.e.Key Performance Indicator - KPI) of the fuel consumption over time.
An analysis will be performed and a KPI created which focuses on the:

  • Lowest (25%) - 25th percentile of fuel consumption.
    25% of the machines of this model consume less fuel.
  • Average - Average fuel consumption of all machines of this model.
  • Highest (75%) - 75th percentile of fuel consumption.
    75% of the machines of this model consume less fuel.

This data will be pushed back to the DataPlatform and visualized in the DataPortal via the Capacity widget, see Widget Catalog.
To set up this process, define the signals as described in Configuration.
For more information, contact your Proemion representative.

Activation

In the DataPortal, proceed as follows:

  1. Activate the Model Visibility feature via the Feature Switch.
  2. You can adapt the model's visibility, see Model-Based Signals Visibility.
  3. You can additionally change the usage of the model across organizations, see Models Management in the DataPortal Manual.

Configuration

The Model-Based Signals script1 automatically runs once a day and processes data from the previous day.

For the analysis of "Fuel Consumption", proceed as follows:

  1. Define the signals and the aggregation function as described in the following table within the PDC, see PDC Manual → "ValueDefinition".

    Parameter Values Additional Information
    Signal keys for operating hours and total fuel consumption2 value.common.machine.hours.operation.total Operating Hours and Fuel Engine Total
    value.common.engine.fuel.used.total
    boundTo="" model The signal keys must be marked with boundTo="model" in the PDC to indicate the signals are provided by the model, see PDC Manual → "ValueDefinition".

  2. Inform your Proemion representative about the signals to analyze, the aggregation functions and other information as described in the following table.

    Parameter Values Additional Information
    Signal keys for operating hours and total fuel consumption2 value.common.machine.hours.operation.total Operating Hours and Fuel Engine Total
    value.common.engine.fuel.used.total
    Aggregation Delta Both signals
    Ignore machines name : ["Test", "Demo"]
    cu.name : ["missing"]
    model.name : ["missing", "Default"]    		 Can be used to exclude machines from being visible for other users.
    Bucket size 3600000 ms 1 Hour
    Selected Models [] List of models - Empty means all models

  3. The defined signals will be populated via the REST API by using the script.

  4. The following metrics for each result of the analysis can be then obtained via the REST API, see Measurement of Model-Based Signals.

    Name Signal
    Fuel Consumption lowest 25% value.custom.model.fuel_consumption_25
    Fuel Consumption average value.custom.model.fuel_consumption_average
    Fuel Consumption highest 75% value.custom.model.fuel_consumption_75

  1. The script can be created together with Proemion based on your input.
    Contact your Proemion representative. ↩

  2. It is recommended to use the standard keys noted there, see also Proemion Standard Metrics. ↩↩