Sourced from a number of suppliers of varying scales based in Shenzhen, Sham Shui Po, the London Borough of Wandsworth, Ithaca, Snowdonia, and Denver, the kit developed and tested by Citizen Sense researchers during the “Air Walk: Environmental Monitoring in NXG” pilot project consisted of 5 air quality sensors, 1 air quality shield, 1 Shinyei Model PPD42NS Dust Sensor, 1 water sensor, 1 sound sensor, 3 Gas sensors (MQ2), 3 Gas sensors (MQ9), 2 Seeeduino V3.0 (Atmega 328P), 1 arduino board, 3 LED screens, 5 Grove Shields, 1000 resistors of varying ohms, a Universal 4 Pin 20cm Unbuckled Cable, 10 No2 sensors, 1 large gas sensor (gas sensed: unknown), 1 arduino ethernet, 3 lithium backpacks, 5 9-volt batteries, and 1 USB cable.
This list of sensors, cables, wires and other hardware forms part of the inventory of materials sourced for the air-sensing kits. As part of our methodological approach to the “Air Walk,” we assembled a series DIY sensing devices so that we might engage with the practices of citizen sensing through the sensors themselves. The process of making a number of kits for this “Air Walk” led us to an in-depth engagement with the proliferation of low-cost sensors, prototyping platforms, toolkits, devices, peripherals, mobile apps and code repositories. This approach is central to the practice-led research that we are undertaking through the Citizen Sense project, where through working with sensing kit that is meant to be relatively accessible, we can test the practices and claims made for citizen-sensing projects.
In order to sense air quality in New Cross Gate and Deptford, an area that typically exceeds EU air quality objectives for nitrogen dioxide (NO2), we developed one kit to sense NO2 levels along the walk, another to measure particulate matter 2.5 (PM2.5) particularly at the location of the SELCHP incinerator, and a third kit to display overall levels of gases, temperature and air quality (if you are interested in the how-to of these devices, schematics and code, we will be sharing this material in a blog post soon).
The kits we created from our inventory of materials became contingent on the sensors available and in circulation. However, the development of these kits allowed us another way to think though the potential sensing practices these assemblages perform, through the often hidden performances of hardware, software and production. By approaching these kits both as a counterpoint to official monitoring stations and as way to get inside monitoring practices by taking the instruments into our hands, “to hold them to one’s lips to experience them through sensation” (Barad 2007), a number of questions regarding the practices of prototyping portable devices from the assortment of products and instructions available became evident.
Citizen-sensing practices emerge as entanglements of DIY culture, environmental concerns, activism, tinkering, hobbyism, amateurism and startup culture. Through making kits we engaged with the processes of making the kits, spending hours trawling through forum comments, posting questions to e-lists, searching for specific sensors, and shopping for them online and in electronic markets. Historically, the phenomenon of sharing technologies for citizen science seems to be a key way in which technologies and devices developed. Air-sensing practices with mobile kits have an a ongoing and lengthy entanglement with citizen-led and collective organization. From our project desks cluttered with devices, wires and solder, the difference that becomes apparent is that in these earlier citizen-science projects, networks collectively organised around a few or even one sealed measuring device, which was shared across informal and formal networks. Measurements from these earlier mobile devices were also collected by experts to form “investigations,” such as in the Operation Smog event of 1956 in London, where 450 volunteers formed an informal network to share expensive equipment to measure smog levels in what was said to be the first full-scale investigation of London air quality.
Like the historical practices of Operation Smog, present-day citizen-sensing practices using air-sensing kits also materialise air measurements and investigations continuously through a series of intra-actions. However, because they involve assembling, adjusting and remaking a distributed array of technologies, the kits become not just a tool of investigation at sites of measurement, but also constitute sensory investigations into the practices of citizen sensing. Always in progress from their beta version, reformed and reforming through practice, citizen-sensing kits are prototypes of ongoing and emergent environmental sensation.
Following the promises of DIY sensing in blogs, re-blogs, tweets, links and workshop adverts, it would seem that the process of building an air-quality sensing kit should be low cost, achievable, and relatively easy–or at least it should be available in a kit form or as an “instructable.” However, despite the force of discussions and emergent designs of devices, there are surprisingly few projects that have emerged for the easy and effective mobile sensing of air quality. Many sensing projects that have arisen are in response to the “off-the-scale air pollution in China this year,” and are still in the prototyping stage. This has led to a gaggle of contingent devices emerging from DIY maker communities who take up the opportunity or provocation to test out new sensors supplied by component companies. What are the potentials of these DIY sensing practices that gather around one particular sensor? And how do sensing practices emerge that are contingent on the availability and circulation of sensors?
One of our kits for the “Air Walk” pilot project emerged from the Shinyei Model PPD42NS Dust Sensor supplied by Seedstudio. A small black box mounted on a plastic board, the PPD42NS Dust Sensor is designed to measure particulate matter levels in the air by counting the Lo Pulse Occupancy time (LPO time) in given time units, and is responsive to particulates with a diameter of 1µm or less. Another one of our kits developed for the “Air Walk” emerged from a small metal sensor consisting of a silver cylinder with a metal stretched gauze over its frame mounted on a plastic board labelled “air quality sensor,” also supplied by a Shenzhen-based supplier. The distributors Seeedstudio sell the “air quality sensor” as part of their “grove system” (a plug-in brick system that removes the need for soldering, wiring or working out resistances). Although designed for static indoor use, it has become a popular sensor for people to prototype outdoor air-quality measuring devices. The manufacturers state that due to the measuring mechanism, this sensor cannot output specific data to describe target gases’ concentrations quantitatively. Instead, it is designed as a sensor for applications that require qualitative results. The sensor documentation suggests this might be useful for auto-refresher sprayers (which are often used in homes or stores), and auto air cycling.
The “qualititative” description output by this sensor is printed in the serial communication from the device. As we needed our kit to be mobile and readable in the field, we printed the readings to an an LED screen, powered by a lithium backpack. Depending on the internal logic of the sensor it will write: “High pollution! Force signal active,” or “High Pollution!” or “Low pollution!” or “Fresh air” (note: no exclamation mark for this reading). These texts are based on a numerical reading that is output by the sensor, yet not mapped to any standard values. As this part of the code below exemplifies:
Serial.begin(9600); airqualitysensor.init(14); } void loop() { current_quality=airqualitysensor.slope(); if (current_quality >= 0)// if a valid data returned. { if (current_quality==0) Serial.println(“High pollution! Force signal active”); else if (current_quality==1) Serial.println(“High pollution!”); else if (current_quality==2) Serial.println(“Low pollution!”); else if (current_quality ==3) Serial.println(“Fresh air”); } }
The “if command” of the code shows the narrow limitations of the sensor and the numbers give the impression that there is equal space between High Pollution and the next category down of Low Pollution. In a recent forum discussion, a user of the “air quality sensor” highlighted a mistake in the sample code provided by the distributor, where the error/overflow in the AirQuality.h variable “temp” was causing the sketch to always show “high pollution” at certain temperatures.
There is also some ambiguity in the data sheet for the sensor, which describes the sensor as measuring harmful gases: carbon monoxide, alcohol, acetone, thinner, formaldehyde and the indeterminate category of gases that are labelled as “and so on.” Following the documentation back to the manufacturer, the “and so on” category to which the sensors are apparently calibrated remains relatively unexplained, even within the technical notes. The peculiar position of measuring “and so on” from a technical datasheet points towards these sensory material engagements as a moment of the unknown and partial. There is in this technical datasheet a glimpse of the sensor as performative, uncertain and always in process, yet still monitoring and marking environments.
Vicki Kirby, following Bruno Latour suggests, “data is indicative.” Sensors are no exception to these indicative inscriptions, as they often “throw up nodes of reference that effectively correspond” (Latour 24, cited in Kirby 81). However, how do we account for the inscription of the sensor itself, both into the data and through the devices and kits that assemble through these monitoring technologies? And most importantly for the purposes of the Citizen Sense project, how do these instruments and specific material arrangements intervene within the practices and politics of citizen sensing?
References:
- Karen Barad, Meeting the Universe Halfway: Quantum Physics and the Entanglement of Matter and Meaning (Durham: Duke UP, 2007)
- Vicki Kirby, Quantum Anthropologies: Life at Large (Durham: Duke UP, 2011)