AirGradient Arduino Library

The corrections object allows you to apply local correction algorithms to PM2.5 (pm02), Temperature (atmp), and Humidity (rhum) values. This affects the local server response, the physical display, and open metrics.

PM 2.5 (pm02)

Available correctionAlgorithm values:

• "none": No correction (default).
• "epa_2021": Uses EPA 2021 correction factors.
• "slr_PMS5003_20240104", "slr_PMS5003_20231218", "slr_PMS5003_20231030": Specific SLR corrections for sensor batches. Requires slr object with intercept, scalingFactor, and useEpa2021.

Temperature (atmp) & Humidity (rhum)

Available correctionAlgorithm values:

• "none": No correction (default).
• "ag_pms5003t_2024": Standard AirGradient correction (for outdoor monitors).
• "custom": Uses manually set intercept and scalingFactor in the slr object.

Note: If configurationControl is set to local, corrections must be set manually.

# Example: Apply SLR correction for PMS5003 batch 20231030
curl --location -X PUT 'http://airgradient_84fce612eff4.local/config' \
--header 'Content-Type: application/json' \
--data '{"corrections":{"pm02":{"correctionAlgorithm":"slr_PMS5003_20231030","slr":{"intercept":0,"scalingFactor":0.02838,"useEpa2021":true}}}}'

# Example: Apply custom temperature correction
curl --location -X PUT 'http://airgradient_84fce612eff4.local/config' \
--header 'Content-Type: application/json' \
--data '{"corrections":{"atmp":{"correctionAlgorithm":"custom","slr":{"intercept":0.2,"scalingFactor":1.1}}}}'

The Sensirion Gas Index Algorithm converts raw sensor signals (SRAW_VOC and SRAW_NOX) from SGP40/41 sensors into robust VOC and NOx Index outputs.

Key Concepts

• Normalization: The algorithm uses statistical gain-offset normalization that adapts constantly using an exponentially decaying function to handle changing environments and minimize sensor-to-sensor variation.
• Dual Instances: To get both indices, you must instantiate the software twice: once as a VOC Algorithm (using SRAW_VOC) and once as a NOx Algorithm (using SRAW_NOX). Note that the SGP40 sensor only provides SRAW_VOC.
• Sampling Interval: Raw signals must be fed to the algorithm at a constant sampling interval that matches the sensor's readout interval. The default interval is 1 second. If you change the sensor sampling interval, you must also update the sampling interval definition in the algorithm's .h file.
• Recursive Calculation: The algorithm calculates indices recursively using a single raw tick value per time step, maintaining internal states like estimated mean, variance, and uptime.

AirGradient ONE and Open Air monitors (firmware version 3.1.1+) support Over-the-Air (OTA) updates. The device automatically checks for updates on startup and at regular intervals.

Update Mechanism

1. Version Identification: During compilation of an official release, the git tag (GIT_VERSION) is embedded into the binary.
2. Update Request: The device sends a GET request to the following URL, passing its current version as a query parameter: http://hw.airgradient.com/sensors/{deviceId}/generic/os/firmware.bin?current_firmware={GIT_VERSION}
3. Server Response Scenarios:
   • 200 OK: An update is available. The server returns the binary data, and the device performs the update.
   • 304 Not Modified: The device is already running the latest firmware.
   • 400 Bad Request: The firmware version is unknown to the server. This typically happens when running local development builds where GIT_VERSION defaults to snapshot. In this case, no update is performed.