Skip to content

KPI Calculation Service

Idea

The KPI Calculation Service computes Key Performance Indicators (KPIs) for an asset. It uses data source such as sensors, control events and calendar entries.

The service enables a user to get a unified view on an asset and to outline the following aspects:

  • Industrial procurement: reliability, availability and maintainability
  • Conditions monitoring: operating characteristics
  • Alarm management and inferred events
  • Risk assessments
  • Diagnostics applications and Root-cause analysis
  • Condition-based maintenance

Access

For accessing this service you need to have the respective roles listed in Analytics Services roles and scopes.

Basics

KPI calculation

The KPI Calculation Service offers a set of computational procedures based on ISO 3977-9:1999. It works for both offline batch analysis and the online analysis of new data.

The calculation uses the following inputs:

  • Time period
  • Sensor data as time series {TimeStamp, SensorValue} and a threshold
  • Events/states as time series {TimeStamp, EventID/State}
  • Maintenance calendar data {StartTimeStamp, EndTimeStamp}
  • Initial and default unit states

Depending on the availability of the input data sources a specific computational procedure is applied.

The calculated output varies based on the input, but always consists of a set of KPI time and factor values.

KPI states and time values

A unit spends time in various KPI states.

For a considered time period (Period Hours) the following hierarchy shows the resulting time values of a KPI calculation:

  • Available Hours
    • Service Hours
      Unit is in-service.
    • Reserve Shutdown Hours
      Unit is available, but non-generating and not in service (default).
  • Unavailable Hours
    • Forced Outage Hours
      Unit is unavailable due to a failure.
    • Planned Outage Hours
      Unit is unavailable and in a planned outage by calendar.
  • Unknown Hours (NoData)
    Not all required data is available.

KPI factors

The KPI calculation also produces values for the following KPI factors:

  • availabilityFactor
    Probability that the unit will be usable (available without any outages).
  • unavailabilityFactor
    Probability that the unit will be unusable (not available due to outages).
  • reliabilityFactor
    Probability that the unit will not be in a forced outage condition.
  • serviceFactor
    Probability that the unit will be in an operating condition.
  • forcedOutageFactor
    Probability that the unit will be in a forced outage condition.

Features

The KPI calculation Service exposes its API for realizing the following tasks:

  • Calculate a set of KPIs within a time period
  • Get the all KPI states within a time period

Example Scenario

The manager of a brewery wants to calculate KPIs for parts of the production line.

The manager collects time series data of a relevant sensor of the production line using the IoT Time Series Service and feeds the KPI Calculation Service API with this time series and the desired time period.

Any questions left?

Ask the community


Except where otherwise noted, content on this site is licensed under the MindSphere Development License Agreement.