The unexpected irony
Earlier this year, researchers uncovered a strong association between (paediatric) asthma and exposure to traffic related air pollution. It was further found that approximately 4 million new asthma cases could be the result of the aforementioned traffic related air pollution – of which 64% will occur in urban centers (Achakulwisut, Brauer, Hystad, & Anenberg, 2019). Despite the “fresh air” we are surrounded by during our bike commute in the Netherlands, further reports outlined by the “LongFonds” revealed that asthma is one of the more commonly faced chronic diseases faced by children in the Netherlands (Longfonds, 2019).
Albeit rather ironic, could it perhaps be that the “healthy” biking culture engrained in our society actually exposes us to such traffic related air pollution, especially in denser populated areas such as the Randstad (an area in the Netherlands that covers the four largest cities representing one of the more densely populated areas in Europe)? What role can technology play in identifying problem areas and perhaps even resolving such issues?
A fusion of two worlds
Investments in the internet of things, once a hype that struggled with feasibility, are now bearing their fruits as scalable and realistic projects begin providing solutions to societal challenges. One project undertaken by the Dutch province Utrecht in collaboration with Civity (company focusing on smart cities / data platforms) and SODAQ (internet of things hardware and software) involved creating air quality monitoring stations mounted on bikes (bicycles for the Americans reading). Yes, you read that correctly – bikes (CEF DIGITAL, 2019).
A dynamic element in a static world
How and why is mounting an air quality device onto a bike related to asthma? To answer that question, one needs to consider the current source of information for critical air-quality decisions. For a long time, established websites such as waqi.com have provided air quality information for specific locations (see figure 2). By utilizing a handful of measuring stations in a specific city, waqi is able to provide the air quality for the exact location where the stations are placed. One of the significant downsides to such a methodology is the static location of the station. For example, a biker on a busy road is exposed to different air quality than a static air quality station on top of a building in Amsterdam. Similarly, a biker exploring the local green park around his house faces different breathing conditions. It is evident what I am alluding to: the dynamic element of air quality measurements (by mounting them on bicycles) changes the way cities can be monitored. Instead of a handful of measurement locations in a specific city, the IoT addition could provide a few thousand locations, enriching the data collected and hence providing a stronger quantitative support for solutions to the air quality problem (figure 3).
Source: (The World Air Quality Index project)
The air quality devices are comprised of a two-part solution: hardware and the data platform. The hardware aspect solution consists of a custom development board (programmable through the Arduino IDE) with numerous onboard sensors (GPS, magnetometer, accelerometer, amongst others). This onboard accelerometer allows the device to register movement and only begin with air quality monitoring if it is in movement. Furthermore, the GPS tracks the location where the measurement was taken. The main feature, the air quality, is monitored through a PM 2.5 (particulate matter) sensor which accurately measures the amount of particulate mass concentration through laser scattering technology (SODAQ, 2018).
The information is then sent to a public dashboard every 10 seconds over the LTE-M network, a network developed to send data packages from IoT devices utilizing less energy than traditional cellular networks (a must for longevity of IoT devices). To properly visualize the data monitored, a publicly accessible dashboard displays all of the bike routes as well as an option to aggregate the data of numerous bikes in the same area for a clear overview.
The potential implementation
The Dutch choose their bikes as their mode of transportation 36% of the time (TNS Opinion & Social, 2014). Imagine if 1% of the Dutch population (17 million) was equipped with this air quality monitoring device (Worldometer). This would imply that 61,200 (36% of the 1%) people would be using their IoT-enhanced bicycles. If they would use their bicycle for 20 minutes a day on average, this would imply over 7,344,000 data points being created every single day (assuming the standard 10 second measurements).
Source: (Civity)
To link it back to the recent study published – this IoT application (along with the potential millions of data points that it brings) could create a much more in-depth mapping of air quality of cities (as demonstrated by the comparison of figure 2 and figure 3). This can allow for institutions to tackle problem areas (specific streets) or potentially test interventions they make. Think about all the possible quantifications that are possible with such dynamic monitoring – whether it’s understanding the impact of recently introduced “milieu zones”, or whether it’s understanding the impact of the nearby rum factory on the air quality on your street. The quantification of the environment will allow for institutions to utilize data-driven decisions changes to ensure the vitality of the Dutch biking culture (and save children from asthma, of course).
Sources:
Achakulwisut, P., Brauer, M., Hystad, P., & Anenberg, S. C. (2019, April 11). Global, national, and urban burdens of paediatric asthma incidence attributable to ambient NO2 pollution: estimates from global datasets. Retrieved from https://www.sciencedirect.com/science/article/pii/S2542519619300464
CEF DIGITAL. (2019, August 22). Context Broker to empower Dutch cyclists to choose healthier routes. Retrieved from https://ec.europa.eu/cefdigital/wiki/display/CEFDIGITAL/2019/07/31/Context Broker to empower Dutch cyclists to choose healthier routes
Civity. (n.d.). Civity Dataplatform. Retrieved from https://dashboard.dataplatform.nl/sodaq/v2/groene_fietsroutes.html
Longfonds. (2019, September 11). Longfonds en artsen luiden noodklok over vieze lucht. Retrieved from https://www.longfonds.nl/nieuws/longfonds-en-artsen-luiden-noodklok-over-vieze-lucht
SODAQ. (2018). Sniffer Bike. Retrieved from https://sodaq.com/projects/sniffer-bike/
The World Air Quality Index project. (n.d.). World’s Air Pollution: Real-time Air Quality Index. Retrieved from https://waqi.info/#/c/52.371/4.865/12z
TNS Opinion & Social. (2014). Quality of Transport. Quality of Transport. European Commission. Retrieved from https://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_422a_en.pdf
Worldometer. (n.d.). Netherlands Population (LIVE). Retrieved from https://www.worldometers.info/world-population/netherlands-population/
I really like your article and your take on how air quality sensors are not dynamic therefore may lack accuracy, but what about remote areas where no one ever bikes? For instance like industrial zones or wilderness. I guess in some areas we still need to use static sensors. But based on what you said, I think the combination of dynamic and static sensors is the best option to ensure accurate air quality measurements.
Hi Roman,
Thanks for the comment! I fully agree it should not be a full replacement of the current static technology but instead a combination of both sensors that should be mapped onto a single database for even more data points. There are evident pros and cons of the dynamic approach: the static sources have more sophisticated measuring technology (not just PM sensors, but also other air quality indicators). Moreover, the dynamic air quality sensors require NB-IoT connectivity. This implies that potential lack of cellular infrastructure in rural areas would be the first bottleneck to overcome (even before lack of bicycles).
That being said, the current article was written in the scope of air quality monitoring in the Netherlands in a culture where bicycles are used by almost everyone, everyday. Hence I tried to focus on the way current infrastructure (and cultural dispositions) could best empower the institutions with the ability to utilize data-driven decisions instead of on the broader scope of rural areas & areas without bicycles.
Regards,
Stijn
Hi Stijn,
Impressive about a topic far surpassing my expectations. Yet, while reading the article I could not help to wonder why the bike was the vehicle of choice in this study. Bikes indeed can access all city centres in the Netherlands, something cars are not always able to, but bikes are slow and only used for small distances, therefore bikes cannot map as much of a city as for example cars or city busses could. What is your view on the suitability of a hybrid model between cars, busses and bikes to map air quality in cities?
Hi Lars,
Thanks for the comment. I believe bicycles were chosen given the abundance of them in the Netherlands, and quite literally, in bigger cities almost every street will be biked on which can provide unique data that perhaps might not be obtainable by cars. However, I fully agree with the statement that bicycles have certain speed and distance limitations. But I do believe it is a great start that is successfully able to piggyback off of the Dutch culture.
This proposed hybrid model is very interesting to me, and certainly quite feasible. The PM (particulate matter) sensor I mentioned in the article is actually developed for high speeds and use in the automobile industry. This implies that these bicycle sensors could be mounted to cars. Ultimately, the goal should be full integration with other modes of transportation as well, such as the bus, but also tram and train, like you mentioned!