Proper Functioning Index

The Proper Functioning Index (PFI) refers to a percentage of time a machine has been running properly every day.

The index calculation takes into account the number of FAILURES and ANOMALIES having an impact on the machine health (event topic is HEALTH).

Failures and anomalies are weighted differently. A FAILURE indicates that the machine is blocked (i.e. it cannot function), whereas an ANOMALY is not blocking but can cause a breakdown if it is not resolved in time.
Minor anomalies can be discarded according to the threshold specified in the insight metric (by default at least 5 per day).

Operation managers track this index to assess the reliability of their machines and to identify any potential issues that may affect the machine health and thus the productivity and costs.

In order to reduce downtime, the PFI is crucial to identify on which machines the maintenance is needed. In the Smart Maintenance use case, this insight can be very helpful to prioritize the machine’s service.

Within Servitly you can define Insight Metrics based in the Proper Functioning Index algorithm described below.

You can define Insight Metrics based in the PFI built-in algorithm described below.

This feature is available for the following modules:
VALUE-ADDED DIGITAL SERVICES, SMART AFTER SALES & ADVANCED SERVICES

Inputs

Here is the list of inputs required by this Algorithm.

	INPUTS
Events	The FAILURE and ANOMALY events occurred on the machine during the last 24 hours and having HEALTH as topic. This input does not need to be configured. During the computation, the system automatically provides all events (FAILURE and ANOMALY) that have an impact on the health status of the machine. For a correct computation, make sure that you have marked with the topic HEALTH all events that impact the machine's health status. Type: ACTIVE_EVENTS \| Mandatory
Minor Anomalies Threshold	The minimum number of minor anomalies to be considered in the algorithm computation. Type: INTEGER \| Mandatory \| Default: 5

INPUTS

Events

The FAILURE and ANOMALY events occurred on the machine during the last 24 hours and having HEALTH as topic. This input does not need to be configured. During the computation, the system automatically provides all events (FAILURE and ANOMALY) that have an impact on the health status of the machine. For a correct computation, make sure that you have marked with the topic HEALTH all events that impact the machine's health status.

Type: ACTIVE_EVENTS | Mandatory

Minor Anomalies Threshold

The minimum number of minor anomalies to be considered in the algorithm computation.

Type: INTEGER | Mandatory | Default: 5

Outputs

Here is the list of outputs provided by this Algorithm.

	OUTPUTS
Proper Functioning Index	Provides the daily percentage time the machine has an active problem. PFI is calculated by summing the time spent by the machine with an active FAILURE or ANOMALY. Minor anomalies are considered only if they occurred more than a threshold (default 5). A high value indicates that the product is reliably. Conversely, a low value indicates that the product is prone to fault and problems. Type: FLOAT \| Insight Properties: AVG LAST_30_DAYS, AVG LIFETIME

OUTPUTS

Proper Functioning Index

Provides the daily percentage time the machine has an active problem.

PFI is calculated by summing the time spent by the machine with an active FAILURE or ANOMALY. Minor anomalies are considered only if they occurred more than a threshold (default 5).

A high value indicates that the product is reliably. Conversely, a low value indicates that the product is prone to fault and problems.

Type: FLOAT | Insight Properties: AVG LAST_30_DAYS, AVG LIFETIME

Insight Metric Configuration

To use this Algorithm in an Insight Metric, you need to:

Go to the Insight / Insight Metrics page.
Select the Thing Definitions tab.
Select the Thing Definition where to create the insight metric.
Click the Add Metric button.
Select the Proper Functioning Index algorithm.
Select the output for which you want to calculate the metric.

Displaying Insight Values

You can use any widget supporting metric data loading, for instance:

Value: to display a single value, also aggregated on a period.
Benchmark: to compare your machine with similar ones.
Time Series Chart: to display the trend on a period.
Bar Chart: to display aggregated data through statistics.