06 Downtime Reporting

Downtime reporting in Grafana with the ENLYZE API.

What You'll Learn

• Query the downtimes API endpoint

• Handle ongoing downtimes (no end time)

• Calculate downtime duration with transformations

• Visualise downtimes with pie charts and bar gauges

• Create a state timeline for time-based analysis

• Build summary statistics for downtime reports

Prerequisites

• Completed Tutorials 4.1-4.3

• Understanding of Infinity queries and transformations

• Familiarity with dashboard variables

---

Overview

Downtime analysis helps identify:

• How much time is lost to stoppages

• What categories of downtime occur (planned, unplanned, not scheduled)

• Which reasons are most common

• When downtimes occur (patterns over time)

Downtime Categories

Category	Description	Color (Convention)
PLANNED	Scheduled maintenance, changeovers	Blue
UNPLANNED	Manual operator-reported stoppages	Orange
NOT_SCHEDULED	No production scheduled	Gray
(none)	Downtime without assigned reason	Yellow

---

Downtimes API Endpoint

The downtimes endpoint returns machine stoppages:

URL:

downtimes?machine=${machine}&start=${__from:date:iso}&end=${__to:date:iso}

Query Parameters:

Parameter	Description
machine	Filter by machine UUID(s)
start	Filter downtimes starting from (ISO 8601)
end	Filter downtimes up to (ISO 8601)

Response Structure

{
  "data": [
    {
      "uuid": "abc123",
      "machine": "141e0927-62b3-4e76-8398-ad82d20f397f",
      "start": "2024-01-15T08:30:00.000Z",
      "end": "2024-01-15T09:15:00.000Z",
      "reason": {
        "uuid": "reason-uuid",
        "category": "UNPLANNED",
        "name": "Material jam"
      }
    }
  ],
  "next_cursor": null
}

Important: If end is null, the downtime is still ongoing.

---

Basic Downtime Query

Step 1: Create the Query

Infinity Configuration:

Setting	Value
Type	JSON
Parser	Backend
Source	URL
URL	downtimes?machine=${machine}&start=${__from:date:iso}&end=${__to:date:iso}
Root selector	$.data

Step 2: Extract Fields with Columns

Add columns to flatten the nested structure:

Selector	Title	Type
uuid	uuid	String
machine	machine	String
start	start	Timestamp
end	end	Timestamp
reason.category	category	String
reason.name	reason	String

[SCREENSHOT: Basic downtime query configuration]

---

Handling Ongoing Downtimes

Downtimes without an end time are still in progress. Handle them with JSONata:

JSONata to Fill Missing End Times

$.data.{
  "uuid": uuid,
  "machine": machine,
  "start": start,
  "end": end ? end : $now(),
  "category": reason.category,
  "reason": reason.name
}

This replaces null end times with the current time for calculations.

Alternative: Use Query Time Range End

(
  $endMillis := $number($__to[0]);
  $.data.{
    "uuid": uuid,
    "machine": machine,
    "start": start,
    "end": end ? end : $fromMillis($endMillis),
    "category": reason.category,
    "reason": reason.name
  }
)

This uses the dashboard's time range end instead of current time.

---

Calculating Downtime Duration

Grafana transformations can calculate duration from start/end times.

Step 1: Convert Field Types

Add transformation -> Convert field type:

Field	Type
start	Time
end	Time

Step 2: Calculate Duration

Add transformation -> Add field from calculation:

Setting	Value
Mode	Binary operation
Operation	end - start
Alias	duration_ms

This gives duration in milliseconds.

Step 3: Convert to Seconds

Add another calculation:

Setting	Value
Mode	Binary operation
Operation	duration_ms / 1000
Alias	duration

Now duration is in seconds, which works with the dtdhms unit (days, hours, minutes, seconds).

[SCREENSHOT: Duration calculation transformations]

---

Downtime Pie Chart by Category

Visualise the distribution of downtime categories.

Query Configuration

Use the basic downtime query with JSONata:

$.data.{
  "machine": machine,
  "start": start,
  "end": end ? end : $now(),
  "category": reason.category ? reason.category : "UNKNOWN"
}

Transformations

1. Convert field type -> start, end to Time

2. Add field from calculation -> duration_ms = end - start

3. Add field from calculation -> duration = duration_ms / 1000

4. Group by:

   • Group by: category

   • Aggregate: duration -> Sum

Panel Settings

Setting	Value
Visualization	Pie Chart
Reduce	Fields -> duration (sum)
Legend	Table, show percent

Value Mappings

Map category values to readable names:

Value	Display
PLANNED	Planned Downtime
UNPLANNED	Unplanned (Manual)
NOT_SCHEDULED	Not Scheduled
UNKNOWN	No Reason

Color Overrides

Add field overrides for each category:

Category	Color
Planned Downtime	dark-blue
Unplanned (Manual)	orange
Not Scheduled	gray
No Reason	yellow

[SCREENSHOT: Downtime pie chart by category]

---

Downtime Bar Gauge

Show duration by category as horizontal bars.

Panel Settings

Setting	Value
Visualization	Bar Gauge
Orientation	Horizontal
Display mode	Basic
Show unfilled area	Off
Unit	Duration (d hh:mm:ss)

Use the same query and transformations as the pie chart.

[SCREENSHOT: Downtime bar gauge]

---

State Timeline

Show when downtimes occurred over time.

Query Configuration

$.data.{
  "start": start,
  "end": end ? end : $now(),
  "state": reason.category ? reason.category : "UNKNOWN"
}

Panel Settings

Setting	Value
Visualization	State timeline
Merge values	On
Show value	Auto

Value Mappings with Colors

Value	Display	Color
PLANNED	Planned	dark-blue
UNPLANNED	Unplanned	orange
NOT_SCHEDULED	Not Scheduled	gray
UNKNOWN	No Reason	yellow

The state timeline shows colored bands for each downtime period.

[SCREENSHOT: State timeline panel]

---

Downtime Details Table

Show individual downtime events with full details.

Query Configuration

Use the basic query with all fields:

$.data.{
  "start": start,
  "end": end,
  "duration_display": end ? "Completed" : "Ongoing",
  "category": reason.category,
  "reason": reason.name,
  "machine": machine
}

Join with Machine Names

Add a second query for machine names:

Query B:

URL: machines
Root selector: $.data.{"machine": uuid, "machine_name": name}

Transformations:

1. Merge

2. Organize fields -> Reorder and hide machine UUID

Column Configuration

Column	Type	Special
start	Timestamp	dateTimeAsIso
end	Timestamp	dateTimeAsIso
duration_display	String	-
category	String	Value mappings
reason	String	-
machine_name	String	-

[SCREENSHOT: Downtime details table]

---

Summary Statistics

Add stat panels for quick KPIs.

Total Downtime Duration

Setting	Value
Visualization	Stat
Reduce	Sum
Field	duration
Unit	Duration (d hh:mm:ss)

Downtime Count

Add transformation -> Group by:

• Group by: (none)

• Aggregate: uuid -> Count

Setting	Value
Visualization	Stat
Field	uuid (count)
Title	Downtime Events

Unplanned Downtime Only

Add transformation -> Filter by value:

• Field: category

• Condition: equals "UNPLANNED"

Then calculate sum of duration.

[SCREENSHOT: Summary stat panels]

---

Multi-Machine Comparison

Compare downtime across machines.

Query with Machine Names

Join downtimes with machines query:

Query A (Downtimes):

URL: downtimes?start=${__from:date:iso}&end=${__to:date:iso}
Root selector: $.data

Query B (Machines):

URL: machines
Root selector: $.data.{"machine": uuid, "machine_name": name}

Transformations

1. Merge (join on machine field)

2. Calculate duration

3. Group by: machine_name, Aggregate: duration -> Sum

4. Sort by: duration (descending)

Visualization

Use a Bar Chart with:

• X axis: machine_name

• Value: duration (sum)

• Unit: Duration (d hh:mm:ss)

[SCREENSHOT: Multi-machine downtime comparison]

---

Downtime Pareto Analysis

Identify the top reasons for downtime.

Transformations

1. Calculate duration

2. Filter by value -> category = "UNPLANNED" (focus on actionable downtimes)

3. Group by: reason, Aggregate: duration -> Sum

4. Sort by: duration (descending)

Visualization

Use a Bar Chart or Bar Gauge:

• Sorted by duration (descending)

• Top 10 reasons

This creates a Pareto-style view of the biggest downtime contributors.

[SCREENSHOT: Downtime Pareto analysis]

---

Complete Dashboard Layout

Row 1: Overview

• Pie chart by category (6 units)

• Total duration stat (6 units)

• Downtime count stat (6 units)

• Unplanned duration stat (6 units)

Row 2: Timeline

• State timeline (24 units)

Row 3: Analysis

• Duration by category bar gauge (12 units)

• Top reasons bar chart (12 units)

Row 4: Details

• Downtime details table (24 units)

---

Practical Exercise

Build a downtime reporting dashboard with:

1. Variables:

   • Site dropdown (filter machines)

   • Machine dropdown (multi-select)

2. Overview Section:

   • Pie chart by category

   • Three stat panels (total, count, unplanned)

3. Timeline:

   • State timeline showing all downtimes

4. Analysis:

   • Bar gauge by category

   • Top 10 reasons Pareto chart

5. Details:

   • Full downtime table with machine names

---

Summary

You've learned to build downtime reports:

• Query the downtimes endpoint with time filters

• Handle ongoing downtimes (null end times)

• Calculate duration using transformations

• Visualise with pie charts, bar gauges, and state timelines

• Create summary statistics

• Perform multi-machine comparison and Pareto analysis

---

What's Next?

Part 4 Complete! Continue with:

• Part 5: Shopfloor & Mobile – Build dashboards for big screens and mobile devices

---

Quick Reference

Downtimes Query URL

downtimes?machine=${machine}&start=${__from:date:iso}&end=${__to:date:iso}

JSONata for Handling Null End Times

$.data.{
  "start": start,
  "end": end ? end : $now(),
  "category": reason.category
}

Duration Calculation Transformations

1. Convert field type: start, end -> Time

2. Add field: duration_ms = end - start

3. Add field: duration = duration_ms / 1000

Category Value Mappings

API Value	Display	Color
PLANNED	Planned Downtime	dark-blue
UNPLANNED	Unplanned (Manual)	orange
NOT_SCHEDULED	Not Scheduled	#b0b7af (gray)
(empty)	No Reason	yellow

Duration Unit

Use dtdhms (Duration: days, hours, minutes, seconds) for downtime durations.

Group By for Aggregation

Purpose	Group By	Aggregate
Total by category	category	duration -> Sum
Total by reason	reason	duration -> Sum
Count by category	category	uuid -> Count