Custom Air Quality Reports let you choose which parameters to report on and set your own alert levels for each one. Use them when a built-in standard doesn't match your requirements, such as a company CO₂ policy, a lab's temperature range, or a client's contract limits.
The report shows how much time each device spent in each alert level and how readings were spread across the range. It also lists every exceedance event with its start, end, duration, and peak.
For the settings every report shares (name, schedule, working hours, Noesis summary), see Creating & Scheduling Reports.
Set up a Custom Air Quality Report
In Reports, click + New Report.
Fill in the Info card.
In Settings, set Report type to Air Quality and Standard to Custom.
Select your Devices.
Add your parameters and alert levels (see below).
Choose a Data Aggregation interval.
Set the Content and Time Range & Schedule cards, then click Create.
Parameters & Alert Levels
Each parameter you add gets its own block with its own alert levels.
Click Add parameter and choose a parameter, such as CO₂, ppm or Temperature, °F.
In Parameters, choose which devices' readings to include for that parameter.
Click + Add alert level to add an alert level for this parameter:
Operator – greater than (>), greater than or equal to (≥), less than (<), or less than or equal to (≤)
Value – the threshold, in the parameter's units
Name – the label used in the report. If left empty, it defaults to Alert Level 1 (or Alert Level 2 for the second level).
To add a second alert level click + Add alert level 2.
To delete an alert level, click ⊖ next to it.
To remove the whole parameter, click Remove.
Example: CO₂ with two levels
Operator | Value | Name | Meaning |
> | 800 ppm | Alert Level 1 | Readings from 800 to 1000 ppm |
> | 1000 ppm | Alert Level 2 | Readings above 1000 ppm |
Example: temperature with a comfort range
Operator | Value | Name | Meaning |
> | 74 °F | Alert Level 1 | Too warm |
< | 67 °F | Alert Level 2 | Too cool |
Readings that don't fall into any alert level count as Passed. In the temperature example, that means anything from 67 to 74 °F.
Data Aggregation
Aggregation sets the interval that readings are averaged over before the report is built, such as 5 minutes. Every chart, statistic, and exceedance event uses these averaged values.
Shorter intervals catch brief spikes. They suit short periods and close investigation.
Longer intervals smooth out brief spikes. They keep reports for long periods easy to read.
What's in a Custom Air Quality Report
Header
The report name and description, plus an Air Quality · Custom label. It also shows the account name, reporting period, working hours (if set), and time zone.
Parameters Summary
Each parameter has a card with one row per device and an All devices row that combines them. Each row shows:
Stats – Min, Avg, and Max, plus the P50, P95, and P98 percentiles (explained below)
Data coverage – how much of the expected data for the period the device actually reported. Devices with no data show No data available for this device.
Time at each alert level – the percentage of time readings spent in each alert level
A histogram of how readings were spread across the range (explained below)
The legend at the top of each card shows the range for Passed and for each alert level, such as Alert Level 1 5-10 µg/m³.
Reading the histograms
Each histogram shows how much of the reporting period a device spent at each value.
Each bar is a small range of values, such as 50 ppm for CO₂ or 1 °F for temperature. The label above the chart shows the range, e.g. Share of time per 50 ppm bin.
Bar height is the share of time readings fell in that range.
Bar color matches the legend: blue for Passed, or the color of the alert level the range falls in.
Dashed lines mark your alert level thresholds.
The histogram shows more than whether readings passed. It shows how close they came to a threshold and whether they stayed steady or varied widely. For example, a narrow cluster of bars far below the first dashed line means a device stayed well within limits the whole time. Bars piled up just past a line mean the readings were consistently a little over the limit, which may only need a small adjustment.
Understanding P50, P95, and P98
Percentiles show what was typical without being thrown off by a single spike.
P50 (median) – half of the readings were at or below this value.
P95 – 95% of the readings were at or below this value.
P98 – 98% of the readings were at or below this value.
If Max is much higher than P98, the highest readings were brief spikes rather than a lasting problem.
Analysis per Device
Each device has its own section with its data coverage and reporting period. Each parameter shows:
A line graph of readings over time, with dashed lines at each alert level
Stats for that device (Min, Avg, Max, P50, P95, P98)
Exceedance Events – a table of each time readings crossed an alert level, with Level, Start, End, Duration, and Peak. A Total row sums the duration and shows the highest peak. If nothing crossed a level, the table shows No exceedances.
Summary by Noesis
If Include Summary by Noesis is on, the report ends with an AI-written summary and key findings. Noesis points out which parameters went over your alert levels, when the peaks happened, and which devices had no data.
Tips
Name alert levels in your team's terms, such as Warning and Action. The names appear in the legends, histograms, and exceedance tables.
Turn on Working hours to report only on occupied hours. Overnight readings won't dilute the results.
Check data coverage first. Low coverage means the report is based on partial data. See Troubleshooting Devices.