Sensors, Uncensored

640px-LilyPad_Arduino_Main_Board

How open sensor data can help enrich storytelling + bolster the media’s potential role in galvanizing civic engagement with the sciences.

By Lily Bui

Listen up, journalists.

Something interesting is happening, and you should be in the loop. Two words: “sensor journalism.”

The term is not entirely new, as it has circulated the blogosphere again and again. However, it’s been popping up more frequently in my e-mail stream and in conversations I’ve had with people on and offline, which kind of, sort of, really makes the science nerd in me tingle with excitement.

What is sensor journalism?

Sensor journalism refers to a method of generating or collecting data from sensors, then using that data to tell a story. You may think this sounds familiar, especially with the rise of data-driven journalism and the open data movement. However, as Kelly Tyrrell aptly puts it, “sensor journalism is the first cousin of data journalism.”

The distinction is this: instead of scraping data from the internet or existing databases, you are collecting the data (or enlisting the help of others to do so). Using sensors. In real-time.

I know that was a mouthful. So, let me show instead of tell.

My favorite example, to date, is WNYC’s Cicada Tracker. Lead by John Keefe, the project engaged WNYC listeners to build their own temperature sensors at home using instructions provided on the station website. The goal was to crowdsource temperature readings around the east coast to predict the emergence of the Magicicada brood. The data were then collected, visualized (beautifully) on a map, and used to tell a story.

The Cicadas Are Coming! from Radiolab on Vimeo.

Both scientific research and journalistic endeavor begin with the same thing: a question. In answering that question, for both science and journalism, crowdsourcing data allows the public to actively contribute to the investigation of the truth.

Who is gathering the data, and how accurate is it?

To reiterate, the idea of crowdsourced data collection is not new. Maker communities like Instructables, Spark Fun, Public Lab, MakeZine, and DIY.org have been around for a while and often focus on building tools (hardware and software) to make remote measurement possible. People who identify themselves as citizen scientists, hobbyists, or amateur scientists connected to communities like Cornell’s Ornithology Lab, CosmoQuest, SciStarter, Your Wild Life, and Zooniverse are also likely candidates for crowdsourced projects.

“This kind of technology is not for monitoring people,” said Travis Hartman in a recent interview with Current. “It’s for monitoring the environment we all share.” Hartman is a journalism grad student with an idea for a project to deploy a set of sensors throughout Columbia, Missouri, in order to study the city’s sound ecology.

As you’ve probably already picked up, Hartman is a j-student, not a scientist. Like him, many citizen scientists are non-experts (i.e. they don’t hold formal science degrees) but have an avid interest in science. That said, the argument against the quality or legitimacy of crowdsourced data does come up. And I’ll concede that it is, indeed, a valid one. How can we trust the data, even if we can track where it’s coming from? How do we know that build-it-yourself sensors are accurately calibrated and in working condition? Or worse–what if the data are biased or completely wrong? We can’t always know, but that shouldn’t deter us.

[AsideThere is also a phenomenon known as the "wisdom of crowds."]

PMF_SamilingUSGS1_l

An academic paper finds, “Most [citizen science] projects employ multiple mechanisms to ensure data quality and appropriate levels of validation. (Wiggins, et al., 2011)” Some citizen science project managers use crowdsourced data sets in an auxiliary manner to observe general trends rather than precise data points. Then, some projects use crowdsourced data comparatively, with reference to existing academic data. The basic message from the research community seems to be, We know. We’re aware. We’re working on it. We’ve found some solutions in the meantime.

Meanwhile, journalists are also finding ways to use crowdsourced data to contextualize and enrich stories rather than relying on them as a primary means of telling them (e.g. WNYC’s Cicada Tracker). At the end of the day, a healthy level of skepticism can only help advance current methods of crowdsourced data collection, as it suggests room for improvement.

How can open sensors benefit journalists?

When it comes to environmental monitoring, government agencies like the Environmental Protection Agency (EPA) do run regularly scheduled tests on air and water quality. However, this data is not always available to or accessible by the public. Also, if something were to affect the air or water quality between testing periods, and a government agency wasn’t aware of it, the public would otherwise be left in the dark about their environment. Bringing DIY-sensors into the picture could potentially democratize the process of monitoring your surroundings. For journalists, these data can offer insight into pertinent issues that eventually influence policy.

In making a case for why journalists should pay attention to and care about open sensor networks, Javaun Moradi writes (on his blog):

“It’s a responsibility that is every bit as noble as reporting and can achieve the journalism goals of informing the public, investigating corruption, speaking for the voiceless, and seeking truth. The other side benefit is that local media can deeply engage with their audience in new ways.” 

There’s also another dimension to this. While many citizen science projects are national and international (meaning anyone, anywhere can participate), some are local, focusing their research question on a specific region or city. As we all know, news can also be local–even hyperlocal. What could a more robust relationship between citizen science and the media potentially mean? A few things come to mind off the bat:

Tsuga_canadensis_adelges

  • raised public awareness and understanding of scientific research
  • a growing culture of civic engagement
  • deeper interaction with news audiences + richer storytelling

Local media outlets can also help connect the public to opportunities to take action in their own communities. WHYY-FM (an NPR affiliate station) in Philadelphia has launched a bi-weekly citizen science segment in partnership with citizen science site SciStarter, focusing on projects connected to their broadcast region. A recent story focused on how the public can help report sightings of the woody adelgid in Douglas firs to help scientists track the invasive species in or near Philly. Journalists and media outlets can help facilitate the discovery of these types of local opportunities.

Who else is doing this?

These are some more examples of crowdsourced science (some involving sensors, some not) in the news. No doubt, this is only a small slice of the pie. I sense (get it?) that there is much more brewing beneath the surface.

WBUR: Kite aerial photography to challenge construction permit on nuclear waste site
http://www.wbur.org/2013/07/10/pilgrim-nuclear-waste-permit

KPBS: Monitoring shark populations in San Diego
http://www.kpbs.org/news/2013/jun/04/sharks-attracting-attention-san-diego-waters/

KBIA: Urban sound ecology project
http://www.current.org/2014/02/grad-student-teams-up-with-missouris-kbia-to-measure-decibels-in-noisy-city/

Public Lab’s water monitoring tool, RIFFLE
http://publiclab.org/wiki/open-water
(There are many other Public Lab tools that can be used for environmental monitoring.)

“What Do Open Sensor Networks Mean for Journalism?” blog post by Javaun Moradi
http://javaunmoradi.com/blog/2011/12/16/what-do-open-sensor-networks-mean-for-journalism/

NBC: How citsci project JellyWatch identified species of salps clogging nearby nuclear reactor
http://usnews.nbcnews.com/_news/2012/04/27/11432974-diablo-canyon-nuclear-plant-in-california-knocked-offline-by-jellyfish-like-creature-called-salp?lite

I am enthralled by how many people are trying new things with sensor data. I love this spirit of experimentation that is circulating, and I hope that it’s contagious.

As we move forward into the future, networked sensors will likely become a more integrated part of our lives. With the improvement of wearable tech like Google Glass, the FitBit, Jawbone, and more, will the possibilities for sensor journalism shift from reporting on our environment to deeper stories on data we’ve collected about ourselves and each other? Granted, these projections are not without their caveats. The incipient ubiquity of networked sensors also raises important discussions about surveillance, privacy, and ethics.

For now, I’d love to hear from you! You’ve made it this far, so you’re clearly interested in this conversation too. (That means we just formed a sacred bond. Sorry, you’re stuck.)

Questions: What other sensor projects (related to journalism or otherwise) have you come across? What other applications do you see for open sensor data that people haven’t tried yet? Leave your thoughts in the comment below or tweet me @dangerbui.

Image: Wikimedia (top & bottom right), USGS.gov (top right)

About these ads

4 thoughts on “Sensors, Uncensored

  1. This is a very thoughtful and thought-provoking overview of a critical need for scientists who have historically relied on remote sensors for a variety of projects. For example, research scientists from the Potomac Environmental Research and Education Center at George Mason University in northern Virginia have studied the ecology throughout the Potomac River watershed. My PEREC research projects investigate the presence and concentrations of various pollutants in streams and rivers that enter the Potomac River ecosystem from myriad land-use practices across agricultural, suburban and urban regions.

    Simply knowing that a toxic pollutant is present in a stream at a certain concentration is only part of the story. We also need to know something about the circumstances in which it exists. Obviously, a stream that contains a high concentration of a pollutant–say atrazine, a widely-used herbicide–that has deep and rapid flow will have a greater affect on the Potomac River ecology than a low concentration in a small, trickling stream. For decades, the information about stream flow (discharge) and depth (gage height) have been available through the USGS network of sensor stations that monitor stream 24/7 throughout the US. In recent years, date at most sites can be accessed online in real time through the site http://waterdata.usgs.gov/nwis/rt.

    Combined, this data gives scientists and policymakers “loadings” data–kilograms per annum of each pollutant entering a water body.Many “supersites” have additional sensors that monitor other important parameters such as temperature, turbidity, pH, dissolved oxygen and specific conductance all of which provide context to the stream concentrations and clues to the origin, fate and transport of pollutants.

    Unfortunately, budget conditions in last several years have led to the following statement from USGS on the monitoring sites:

    “The U.S. Geological Survey (USGS) will discontinue operation of a number of streamgages nationwide due to budget cuts as a result of sequestration. Additional streamgages may be affected if partners reduce their funding to support USGS streamgages. The USGS is working to identify which streamgages will be impacted and will post this information as it becomes available.

    Some sites such as the one located on the South Fork of the Shenandoah River–the largest Potomac River tributary–have been scaled back from a supersite to only 2 parameters (waterdata.usgs.gov/va/nwis/uv/?site_no=01629500&agency_cd=USGS) while other sites have been deactivated completely such as Cedar Run, a stream draining a heavily agricultural region (http://waterdata.usgs.gov/va/nwis/nwisman/?site_no=01656000&agency_cd=USGS).

    These circumstances provide excellent opportunities for citizen science and crowd sourcing. Crowd-sourced funding could lead groups to “adopt a station” in order to restore its functionality. Citizen scientists could replace the sensors with targeted collection of stream data that is easily obtained with readily available hand-held sensors. Networks of groups could simultaneously collect bottles of stream water for lab researchers to analyze as well as site-specific parameters that provide the context to the story of our valuable water resource.

  2. Hi Lily: Nice roundup. I hope you’ll check out the two books Patrick Di Justo and I have co-authored: “Environmental Monitoring with Arduino” and “Atmospheric Monitoring with Arduino,” both Maker Media. We wrote them a few years ago intending, among other things, to inspire hands-on data gathering among our colleagues in environmental and science journalism. We’d love to hear of any persons and groups using or innovating off of these projects.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s