DustSensor

Project Dust Sensor
Experiments with a dust sensor
Status	Completed
Contact	bertrik
Last Update	2018-06-13

Introduction

PM10 in the morning of 2018-06-03

PM10 at around 2018-06-03 20:50 UTC

To get an idea of the levels of fine dust around my house, I ordered some dust sensor modules to play around with. In particular, I ordered these ones:

• The Plantower PMS 7003 (AliExpress link);
• The SDS011 (AliExpress link).

The PMS7003 is being advertised as an advanced generation of dust sensor (7th generation), while still reasonably priced (E15,-). They use a laser to perform the measurement and contain a small fan to move the air around.

In parallel with the fine dust level measurement, my device also measures basic meteo data using a BME280 module: temperature, relative humidity and air pressure.

I recommend to also get the PMS7003 cable converter board. The pitch of the connector on the sensor module is slightly non-standard: 2 x 5 pins with a spacing of 0.05 inch (instead of the common 0.1 inch spacing).

The data produced by this sensor is sent as JSON to a MQTT server. From there it is picked up by a protocol converter: a custom written Java application to forward it to the RIVM samenmeten website and forward it to luftdaten.info

Future work

Next steps:

• create animated dust maps from open data from luftdaten.info
• build more luftdaten.info sensors and put them up in Gouda
• build a luftdaten.info sensor and put it up at RevSpace
• investigate a LoRa version of the software + TTN forwarder so we can put up a sensor at Meteo Gouda (kinderboerderij).

More distant future:

• investigate whether it is possible to power the dust sensor off a solar panel (backup battery during the night), this would completely eliminate any wiring to the sensor
• for RIVM samenmeten: use a scheme where the unique id is derived from the ESP id, this should make it very easy to add more sensors with minimal administrative overhead
• for RIVM samenmeten: automatically determine geo position from wifi APs and the google/mozilla wifi location APIs

Results

Below is a graph of the dust levels around new years eve 2017/2018, separated by particle size

You can clearly see a peak just after midnight.

See also the map of these two experiments:

• Samenmeten dataportaal
• Vuurwerk-experiment 2017/2018

Hardware

PMS7003 data sheets can be found:

• here (english) and
• here (chinese)

The PMS7003 takes 5V to run and communicates using 3.3V levels. I connect it using a NodeMCU.

The module gives an estimate of the total mass of the particles (microgram/m3) in 3 categories: PM1.0, PM2.5 and PM10, both for "standard particle" (CF-1) and "standard atmosphere". It also makes an estimate of the raw number of particles per size category, total 6 categories: 0.3-0.5-1.0-2.5-5.0-10 micrometer. I don't know how it actually works on the inside and is able to make a distinction between particles of different size.

Connections

NodeMCU	PMS7003/BME280	Remark
D3	PMS7003-RST	Pulled-up to 3.3V on NodeMCU side, may not be actually needed
D4	PMS7003-SET	Pulled-up to 3.3V on NodeMCU side, may not be actually needed
D5	BME280-SCL	I2C-SCL
D6	BME280-SDA	I2C-SDA,
D7	PMS7003-TX	NodeMCU receive, PMS7003 transmit
D8	PMS7003-RX	NodeMCU transmit, PMS7003 receive
GND	PMS7003-GND / BME280-GND	GND ground reference
VU	PMS7003-VCC	USB voltage (5V)
3.3V	BME280-VCC

Special thanks to Crashjuh for helping with the cable, putting dupont connectors on them, making it a lot easier to connect the module to an ESP8266.

Software

The software archive can be found at github, it consists of the following parts:

• esp arduino software for reading the SDS011 and sending it as an MQTT stream;
• arduino software for reading the PMS7003 and sending it as an MQTT stream;
• a Java based bridge application that takes the MQTT data produced by the arduino and forwards it to other backends, like the RIVM influx database, luftdaten.info API.

Dust sensor

The dust sensor software is written for the Arduino environment.

Typing 'make' builds and runs unit tests that verify parsing of measurement data and construction of command data. The sub-directory 'pms7003_esp' contains the .ino file to be opened in the Arduino IDE.

Libraries used:

• SoftwareSerial (built-in) for serial communication with the sensor
• WiFiClient (built-in) for WiFi connectivity
• WiFiManager (tzapu) to present a captive portal and allow selection of an AP to connect to the internet
• PubSubClient (Nick O'Leary) to handle publishing of data over MQTT
• BME280I2C (Tyler Glenn) and wire for interacting with a BME280 for basic meteo data (temperature, humidity, pressure)

The Arduino software has platformio.org configuration files, so if you're using that, you can just build it with

 pio run

PMS7003

Reading measurements from the pms7003 works, sending commands to the module does not. I don't know yet whether this is a hardware or software problem.

Typing 'make' builds and runs unit tests that verify parsing of measurement data and construction of command data. The sub-directory 'pms7003_esp' contains the .ino file to be opened in the Arduino IDE.

SDS011

Typing 'make' builds and runs unit tests that verify parsing of measurement data and construction of command data. The sub-directory 'sds011_esp' contains the .ino file to be opened in the Arduino IDE.

Infrastructure

The idea is that the dust sensor send its data to an MQTT server. From there on, there is an application written in Java that picks up the data from MQTT and redistributes it to other backends, like samenmeten.rivm.nl and luftdaten.info.

The topic that the sensor publishes on is

 bertrik/dust/<sensorid>

where sensorid is the id of the ESP8266, in hexadecimal format, e.g. AB12CD

References

Grafana page of my sensor data.

Page on aqicn about the PMS5003/7003

Overview of dust sensor types and comparisons between them:

• list of sensors from "samen meten aan luchtkwaliteit".
• another overview of dust sensors.

Citizen science projects for measuring airborne dust:

• Dutch RIVM official air quality map
• Dutch RIVM citizen science air quality map
• Dutch RIVM fireworks smog monitoring during new year's eve
• Belgian project "ik adem" - fijnstofmetingen
• German project luftdaten - measure air data by yourself

Dust measurement blog:

• scapeler

Visualizing dust data

See also my Stofradar page.

Grafana

To create nice graphs in Grafana, I used the following stack of tools/applications:

• a dust sensor, as described here, that publishes measurement data towards an MQTT server
• an MQTT server to accept the data and forward it to subscribers
• the 'Telegraf' importer that listens on the MQTT stream and converts the data to influx database
• an influx database, to store the measurement data
• grafana, to grab the data from the database and display it

Sensor

The sensor produces JSON, grouping the information from one message from the module together. Example

  bertrik/pms7003/json {"alive":1138,"pms7003":{"pm1_0":2,"pm2_5":3,"pm10":4},"bme280":{"t":21.1,"rh":50,"p":1012.4}}

Telegraf

Things I did:

• downloaded and installed the Telegraf .deb from here.
• generated a default configuration using:

  telegraf --input-filter mqtt_consumer --output-filter influxdb config >telegraf.conf

• edited the configuration to set mqtt and influxdb settings
  • influxdb output plugin
    • urls = ["http://172.29.0.1:8086"]
  • mqtt_consumer plugin
    • topics=[ "bertrik/pms7003/json" ]
    • data_format="json"
    • data_type="integer"
• test-run using 'telegraf <TODO>'
• copied the final telegraf.conf to /etc/telegraf and restarted the telegraf service

  systemctl restart telegraf

Grafana

Add a data source, pointing to the influx DB. Provide credentials and verify by pressing the 'test connection' button.

Add a dashboard, add a row to the dashboard, add a graph panel to the row. Under 'metrics', add a query with the following properties:

• FROM mqtt_consumer WHERE topic = bertrik/pms7003/json
• SELECT field(amb_pm10)
• GROUP BY
• ALIAS BY PM10
• repeat for other particle sizes (amb_pm2_5 and amb_pm_1_0)

Regelgeving

From: https://www.rivm.nl/Documenten_en_publicaties/Algemeen_Actueel/Uitgaven/Milieu_Leefomgeving/Dossier_Fijn_stof/Maart_2013/Regelgeving.pdf

 Regelgeving voor fijn stof (PM10)
 De regelgeving voor fijn stof (PM10) kent twee doelstellingen:
 - Een grenswaarde voor het jaargemiddelde: 40 µg/m3 als jaargemiddelde mag niet worden
   overschreden. Aan deze grenswaarde moet sinds 2005 worden voldaan. Behoudens
   ‘derogatie’ zijn geen uitzonderingen mogelijk (->Derogatie). Deze grenswaarde beoogt
   vooral bescherming te bieden tegen de langetermijneffecten van fijn stof.
 - Een grenswaarde voor het daggemiddelde: 50 µg/m3 als daggemiddelde mag op niet meer
   dan 35 dagen per jaar worden overschreden. 5 Aan deze grenswaarde moet sinds 2005
   worden voldaan. Behoudens ‘derogatie’ zijn geen uitzonderingen mogelijk (->Derogatie).
   Deze grenswaarde is vooral bedoeld om bescherming te bieden tegen de
   kortetermijneffecten van fijn stof.

Luftdaten.info

I'm forwarding my data also to the luftdaten.info website so it appears on the lufdaten.info map. I use the same kind of mechanism as used for RIVM: write a Java program to capture the MQTT stream and convert it to their API.

Luftdaten.info have a pretty good description of how to build your own sensor for their network. They're keeping it simple, just wire an sds011 dust sensor with some du-pont cable to a nodemcu and put it in a plastic enclosure anyone can build using some drain water piping.

It appears hack42 is also considering to join the luftdaten.info network.

protocol

Most of the Luftdaten.info upload protocols are described on the luftdaten wiki.

Some examples of how to use it, can be found here: here and this code from ttn-ulm-muecke.

• data is sent as a HTTP POST to https://api.luftdaten.info/v1/push-sensor-data/ (NOTE: the trailing slash!)
• the HTTP POST uses headers "X-Pin" and "X-Sensor" (not case-sensitive it appears)
  • header "X-Pin" indicates the type of data
    • 1 = SDS011 or PMS7003
    • 3 = BMP180
    • 5 = PPD42NS
    • 7 = DHT22
    • 11 = BME280
  • header "X-Sensor" indicates the unique sensor id, for example "esp8266-9137604"
  • header "Content-Type" with value "application/json" (NOTE: API describes this header, but it's not explicitly sent by the implementations above)
• the body of the POST is JSON with the following fields:
  • "software_version": string containing the software version of the sender, for example "python-dusty 0.0.1"
  • "sensordatavalues": array of structures containing measurement data, looking like this:
    • "value_type" (when X-PIN=1): string describing the measurement item type, can be "P1" (PM10 value) or "P2" (PM2.5 value) with dust value in ug/m3. It appears there is also an unofficial "P0" field which encodes for the PM1.0 value.
    • "value_type" (when X-PIN=7): string describing the measurement item type, can be "temperature", "humidity"
    • "value_type" (when X-PIN=11): string describing the measurement item type, can be "temperature", "pressure", "humidity"
    • "value": field containing the measurement value (NOTE: in string quotes, not as numeric values!)

Other meta-information like latitude/longitude of the sensor is sent by an e-mail registration process. This is linked to your measurements by means of the unique sensor id.