In COMPAIR we use a variety of sensors to collect hyperlocal data on air quality and traffic. Some are pre-assembled devices, requiring only self-installation by participating citizens, others are partly or fully DIY (Do-It-Yourself), and as such are more demanding in terms of efforts and skills required to assemble/operate them. In this article we'll introduce sensors used in our citizen science pilots, focusing on key characteristics of a device and its basic functional requirements.
Pre-assembled sensors
SODAQ AIR
The SODAQ AIR is a portable air quality monitor that provides insights into the air you breathe. It monitors particulate matter (PM1, PM2.5, PM10), temperature, and humidity of the air. Besides the air quality, AIR also monitors the device’s movement and location, sending the data to the cloud though LTE-M or NB-IoT. AIR operates on a super capacitor as a substitute for a rechargeable battery, which enhances the environmental sustainability of the device.
Figure 1. SODAQ AIR
Active Mode
The AIR is designed to be used while cycling, so it comes with a bicycle mount to capture air quality in motion. In this mode, AIR will measure air quality every 10 secs. The AIR has a runtime of approximately 5 hours of continuous usage, after which the LED will pulse RED indicating that the device needs to be recharged.
Static Mode
SODAQ AIR needs to be connected to a power source and have the fixed magnet inserted to operate in static mode. Besides, placing it outside in a well-ventilated place will guarantee more insight into the actual air quality (placing it next to a kitchen or heating vents for instance will influence the readings). In static mode, AIR measures the air quality every 5 minutes. For static outdoor use, the AIR needs to be continuously connected to power and therefore needs to be protected from rain.
Viewing the Data
Anonymised data is displayed on a map at knowyourair.net. The data from a specific device can be seen on the same page, by entering the IMEI and CODE of the device. The data is downloadable as an open-access CSV file. An open API is under development in the COMPAIR project. The API will be further integrated into the COMPAIR end-user applications such as the policy monitoring dashboard.
Further instructions on how to use SODAQ AIR can be found in the User Guide.
Telraam S2
The Telraam S2 sensor is designed for easy installation. The device is packed in a custom built package, which includes all the individual components (sensor, power plug, cable and window mounting bracket) as well as a basic installation guide printed on the inside of the packaging.
Figure 2. Telraam S2 packaging
The installation is very easy and requires 5 simple steps:
Create an account on www.telraam.net and link the device to your account by scanning the QR-code when powering the device
Add the user’s address
Find a suitable window
Attach the device using the mounting bracket
Select the segment where the Telraam will be measuring traffic
Figure 3. Telraam S2 mounted on a window
Once installed, the user can interact with the device, exploring counting data on the LCD-screen, visualising data in the last 24 hours or 15 minutes in various level of detail. The user can toggle between the different screens by using the Telraam-button. The sensor LCD-interface also gives an indication of the signal strength of the data connectivity, showing the user if the device can send data to the database. Finally, users can disable the screen with a long-press on the button. This is to reduce light pollution at night. The LCD screen can be reactivated with a simple button press.
OnePlanet NitroSense
This air quality device was developed by OnePlanet Research Center consisting of two Nitrogen Dioxide (NO2) electrochemical sensors (NO2-A43F, Alphasense Ltd, UK), a Telaire T9602 sensor for temperature (T°C) and relative humidity (RH%), and a solar panel.
Figure 4. OnePlanet NitroSense
The device also includes an internal battery which is charged by the solar panel to allow continuous operation away from the electricity grid. The battery capacity is sufficient for 3-months without sunlight. The internal storage matches the timeframe with a capacity that can store data up to 3-months to prevent data loss. The device is capable of providing 3-minute monitoring resolution that is sent to the cloud in 1-hour intervals. The environmental sensors (NO2 and T°C/RH%) are exposed outside of the device to avoid measurement interferences inside the enclosure. Based on our research, the NO2 electrochemical cells of Alphasense are one of the NO2 sensors that can reach ambient ppb (parts per million) detection levels. Based on initial lab tests of OnePlanet Research Center, the device shows good performance in terms of linearity, accuracy and correlation. The datasheet of the NO2-A43F Alphasense electrochemical sensors outlines its specifications.
Being a pre-assembled device, when deployed for conducting citizen science experiments, the installation is fairly straightforward for any persons with an ability to handle the required tools and consumables listed below:
Required tools:
Ladder, minimum reachable height is 3m
Cordless drill (adjustable to 4Nm)
7mm socket
1/4” external square to 1/4" hexagon adapter
Consumables:
2x Hose clamp 60-180mm (Manufacturer part number: QIP180)
Figure 5. Hose clamps for OnePlanet NitroSense
With the above required tools and consumables, installation of the device includes the following steps:
Place the device within a radius of 3m of the intended location (of known GPS coordinates):
If available, use the existing infrastructure (e.g., lamppost, traffic sign, advertisement pole) to place the device. Determine in advance if permission is required for installation at the given location as some restrictions may apply
Place the device avoiding any obstruction e.g. tree
Place the device such that the solar panel on the device is pointing in the direction with the most sunlight
Firmly fix the device to the chosen infrastructure with two hose clamps. In sensitive areas, it is suggested that the device is placed out of reach e.g. 2.5m high
Once installed, take a photograph with GPS tagging enabled, of the device up close to include the device ID. Then take another photograph of the device - with GPS tagging enabled - from a distance to capture how the installation is positioned with respect to the reference station (if present) and the wider surrounding area
Capture ‘exact’ GPS coordinates below the device
On an excel file take note of the following:
Device ID
Installation Date
Location Name e.g. municipality name, cite name, nature name, company name
GPS Location (exact location)
Any remarks if applicable (e.g. height difference if more than 1 metre, infrastructure is unstable but works for now, difficult to get to the installation location, etc.)
Provide the excel file and photographs to OnePlanet for recordkeeping
It is important to note that at some point in time, like all devices, the Nitrosense devices will also require some maintenance depending on the age and status of the sensor device. While it may not apply for citizen science experiments, it is good practice to perform some common maintenance actions from time to time. These include sensor replacement, device reset due to loss of communication, battery inspection due to failure to charge because of the absence of sunlight or a fault in the charging electronics, device cleaning from being exposed to the outdoor environment, inspection in case the device has been tinkered with by unauthorised people without sensor knowledge, etc.
Of the three pre-assembled, self-installation sensors in COMPAIR, the OnePlanet NitroSense is the most complex. Given the limited number of this type of sensor and the level of complexity also with the installation, it is recommended to only promote self-installation of the OnePlanet NitroSense sensor with proper guidance.
Self-assembled sensors
A citizen science lab focusing on sensor assembly is more challenging for the participants but allows for a deeper citizen science experience. Typically, in such a setup, a physical workshop with a good preparation is mandatory. Also, the size of the sensor-assembly workshop should not be too large. In case only one or two researchers can conduct the workshop, the group of citizen scientists preferably should not exceed 30 participants.
Sensor-assembly workshops should focus on the sensor assembly. A generic introduction explaining sensor functionality is still useful, but the bulk of the time should go to assembling sensors.
The format of such a “sensor-assembly workshop” requires thorough preparation by the researcher. The researcher should be able to address any question of the citizen scientist, even if the researcher is not a sensor-specialist. That means, depending on the sensor used, the researcher needs to have previously aligned with the sensor designer/manufacturer to gain sufficient knowledge of the sensor used.
Sensor.Community
One of the air quality sensors being used to measure particulate matter (PM2.5 and PM10), is from Sensor.Community. This is a very basic and low-cost sensor with a long track record of being used in a citizen science context. More than 13000 such sensors are active worldwide within 57 ‘community labs’. Being a well-established sensor kit that has been used in many projects in the past, it was easily incorporated as one of COMPAIR’s self-assembled sensors. The assembly and installation guides are refined and improved from time to time.
The sensor setup deliberately has a strong “DIY-feel” and uses off-the-shelf and universally available low-cost components.
Figure 6. Sensor.Community Kit
The critical components are:
NodeMCU ESP8266 CPU/WLAN: the micro-processor and data communication component
SDS011 Fine dust sensor: the actual particulate matter sensor
A full assembly guide is available on Sensor.Community’s website. More simplified sensor assembly guides are available from derivative projects such as project “luchtpijp” in Belgium. There are also Dutch and English versions of the manual to self-assemble the sensor available online.
bcMeter
bcMeter (https://bcmeter.org/) is similar to Sensor.Community, with the main difference being that the device is measuring black carbon (bc) instead of PM2.5/PM10. However, the sensor technology used is different, a bit more complex and as such a bit more expensive and involves a higher level of technical complexity. However, the device is still low-cost and does not require specific technical skills, so it’s still within reach of use in COMPAIR.
The device consists of a custom 3D-printed casing with an aethalometer as the core sensing component. Sensor “governance” is done with a Raspberry Pi Zero, and an off-the-shelf hardware single-board-computer.
Figure 7. bcMeter
bcMeters are used in the COMPAIR project as a ready-built sensor, but it is possible to also organise sensor-assembly workshops to assemble the sensor, as all components can be sourced separately. The casing is 3D-printed and as such provides an additional opportunity for citizen science labs by organising printing of the casings by citizen scientists who have 3D-printers or with citizen scientists in a 3D-printing lab. 3D-files for case printing are available through the GitHub repository of bcMeter.
bcMeter is not as well established as Sensor.Community, but it does have full documentation of components, assembly guides and any other information that non-experts need to get started. The bcMeter-wiki is the go-to place to access all necessary materials. Due to the sensor technique used, regular maintenance is required which can be performed by the citizen scientist.
Telraam (v1)
In COMPAIR, we will test the user friendliness of the newly developed Telraam sensor “S2”, specifically to reduce the technology hurdle and to reach a larger audience. However, it is also possible to use the original, open source, Raspberry Pi-based Telraam sensor (v1) which can also be completely self-assembled. As with Sensor.Community and bcMeter, the Telraam v1 components are generic and can be sourced via different channels or purchased as a package. The core component is a Raspberry Pi 3 A+ and a camera module v2. Telraam can provide a 3D-printed camera holder, but also generic Raspberry Pi camera holders are available on the market. A citizen science lab involving Telraam v1, could involve assembly and software installation of the sensor in a workshop.
Figure 8. Telraam v1
The component list is as follows (starting top left, clockwise): black enclosure, Raspberry Pi 3 A+, Raspberry Pi power cable, SD card in adapter, Camera holder - either generic or specific 3D-printed version for Telraam, Raspberry Pi Camera module v2 + cable.
Figure 9. Telraam v1 component list
The Telraam FAQ provides a full step-by-step description of each step of the mechanical assembly process.
Apart from the assembly process, as with bcMeter, software installation can also be done by citizen scientists. In the case of Telraam v1, it involves flashing the SD-drive. As Telraam uses Raspberry Pi as a hardware platform, it’s easy to use existing tools to program or install software for Raspberry Pi’s e.g. Balena Etchers. The Telraam v1 software is open source and can be accessed via the Telraam GitHub. However, for non-programers, the software is also available as a simple image which just needs to be flashed using Balena. The procedure is well documented in an FAQ article on the Telraam-website.
More challenging options with Telraam v1 are possible by directly working on the code of the detection script, which is open source on GitHub. This might be of interest to developer communities looking for a new hackathon challenge.
As with bcMeter, there are some 3D-printed components in Telraam v1 (the camera mount) which allow citizen scientists who own a 3D-printer, to make their own Telraam sensor. The design files are available in Telraam’s Github repository.
Summarised in this FAQ article at Telraam are many of the DIY outdoor variants of Telraam v1 that have been attempted previously. This can be used as an inspiration to organise citizen science labs with the sole purpose of making modifications to the sensor that can be made suitable for outdoor use.
Figure 10. Telraam v1 outdoors
Conclusion on citizen science sensors used in COMPAIR
In this article we focused on the different sensors used in COMPAIR. Some are fairly easy to use while others are more complex. The latter require more specialised training and skills, but at the same time provide more opportunities for experimentation e.g. modifications of the device to perform measurements in different locations. The next article in the series will explain what data is collected by these devices and how we manage/use it to facilitate changes in policy and behaviour towards a more sustainable, climate-resilient future.