Open Knowledge Festival 2014

This blog post is written by Lily Bui, M.S. Candidate in Comparative Media Studies at the Massachusetts Institute of Technology, and it's cross-posted from the Tow Center for Digital Journalism blog.
Last July in Berlin, Lily Bui gave a session at the Open Knowledge Festival called Sensors, Uncensored. The session discussed different methods of using sensors to generate data in support of journalistic inquiry and public concerns. The attendees of this session organized a Google group to stay in touch, which has now culminated into a sensor journalism community of practice, which Lily discusses below. This groups is international and multidisciplinary, and they have recently collaborated on a pending Wikipedia article for sensor journalism.

As seen in the Tow Center's Sensors & Journalism report, the growing availability and affordability of low-cost, low-power sensor tools enables journalists and publics to collect and report on environmental data, irrespective of government agency agendas. The issues that sensor tools already help measure are manifold, i.e. noise levels, temperature, barometric pressure, water contaminants, air pollution, radiation, and more. When aligned with journalistic inquiry, sensors can serve as useful tools to generate data to contrast with existing environmental data or provide data where previously none existed. While there are certainly various types of sensor journalism projects with different objectives and outcomes, the extant case studies (as outlined in the Tow report) provide a framework to model forthcoming projects after.

But it may not be enough to simply identify examples of this work.

Invariably, just as important as building a framework for sensor journalism is building a community of practice for it, one that brings together key players to provide a space for asking critical questions, sharing best practices, and fomenting connections/collaborations. Journalism certainly doesn't happen in a vacuum; it is best served by collaborations with and connections to outside sources. A sensor journalism community has already begun to emerge on a grassroots level, spanning multiple countries, disciplines, and issues of concern. We might look to the nodes in this community to outline a protean map of stakeholders in the field:

Journalists.

Since public opinion can be driven by press coverage, and since storytelling is a central part of news, journalists, documentarians, and media makers with an interest in sensor data play an important role in shaping how the public perceives certain issues. More than that, the media also have the ability to highlight issues that may have slipped under the radar of policymakers. In this scenario sensor data could potentially serve as evidence for or against a policy decision. Most sensor journalism case studies so far have relied on normative forms (print, online) to convey the data and the story, but there is much room for experimentation, e.g. sensor-based documentary, radio, interactive documentary, data visualization, and more.

Educators.

In the classroom, there is an undeniable opportunity to cultivate a generation of journalists and media makers who are unintimidated by hardware and technology. Not only this — the classroom also becomes an ideal place to test technology without being beholden to the same restrictions or liabilities as professional newsrooms. Educators across the U.S. have begun incorporating DIY sensors into classroom projects (see Emerson College, Florida International University, and San Diego State University projects), the results of which touch on many of the same questions that professional journalists encounter when it comes to sensor tools. The teaching practices applied to sensor journalism can also be the foundations of training models for professional journalists and civic groups seeking to investigate issues.

Hardware developers.

Because hardware developers design and build the tools that journalists and others would potentially be using, they have a stake in terms of how the tool performs downstream of development. Journalists can also collaborate with hardware developers in identifying tools that would be most helpful: Journalists may have specific requirements of data accuracy, data resolution, range of measurement, or the maturity of their equipment. Likewise, hardware experts can recommend tools that provide access to raw data and transparent interpretation algorithms. On the flip side, some hardware developers, particularly in the open source community, may help identify potential environmental issues of concern that then inform journalists' research. Recently, a conversation about certification of sensors, which originated within the hardware development community, crystallized around the notion of how to instantiate trust in open sensor tools (or sensors in general) when used for various purposes, journalism included.This is telling of how an open dialogue between hardware developers and journalists might be beneficial to defining these initial collaborative relationships.

Researchers.

Since using sensor tools and data in journalism is new, there is still significant research to be done around the effectiveness of such projects from both a scientific/technological standpoint as well as one of media engagement and impact. Researchers are also best poised, within academia, to examine tensions around data quality/accuracy, sensor calibration, collaborative models, etc. and help provide critical feedback on this new media practice.

Data scientists, statisticians.

Since most journalists aren't data scientists or statisticians by training, collaborations with data scientists and statisticians have been and should be explored to ensure quality analysis. While some sensor journalism projects are more illustrative and don’t rely heavily on data accuracy, others that aim to affect policy are more partial to such considerations. Journalists working with statisticians to qualify data could contribute toward more defensible statements and potentially policy decisions.

Activists, advocates.

Because many open sensor tools have been developed and deployed on the grassroots level, and because there is a need to address alternative sources of data (sources that are not proprietary and closed off to the public), activists play a key role in the sensor journalism landscape. Journalists can sometimes become aware of issues from concerned citizens (like the Sun Sentinel's “Above the Law” series about speeding cops); therefore, it’s essential to cultivate a space in which similarly concerned citizens can voice and discuss concerns that may need further investigation.

Urban designers, city planners, architects.

Many cities already have sensor networks embedded within them. Some of the data from these sensors are proprietary, but some data are publicly accessible. Urban designers, city planners, and architects look to data for context on how to design and build. For instance, the MIT SENSEable City Lab is a conglomerate of researchers who often look to sensor data to study the built environment. Sensor data about environmental factors or flow can help inform city design and planning decisions. Journalists or media makers can play a role in completing the feedback loop — communicating sensor data to the public as well as highlighting public opinions and reactions to city planning projects or initiatives.

Internet of Things.

Those working in the Internet of Things space approach sensor networks on a different level. IoT endeavors to build an infrastructure that includes sensors in almost everything so that devices can interact better with people and with each other. At the same time, IoT infrastructures are still in development and the field is just beginning to lay its groundwork in the public consciousness. Imagine motion sensors at the threshold of your house that signal to a network that you're home, which then turns on the devices that you most commonly use so that they’re ready for you. Now imagine that on on a neighborhood or city scale. Chicago’s Array of Things project aims to equip the city with environmental sensors that can report back data in real time, informing residents and the city government about various aspects of the city’s performance. What if journalists could have access to this data and serve as part of a feedback loop back to the public?

By no means is this a complete map of the sensor journalism community. One would hope that the network of interested parties in sensor journalism continues to expand and include others — within policy, legacy news organizations, and more — such that the discourse generated by it is a representative one that can both challenge and unite the field. Different methodologies of collecting data with sensors involve different forms of agency. In some sensor journalism scenarios, the agents are journalists; in others, the agents are members of the public; and in others yet, the agents can be governments or private companies. Ultimately, who collects the data affects data collection methods, analysis of the data, and accessibility of the data. No matter what tools are used — whether they are sensors or otherwise — the issues that journalists seek to examine and illuminate are ones that affect many, and on multiple dimensions (individual, local, national, global). If we are truly talking about solving world problems, then the conversation should not be limited to just a few. Instead, it will take an omnibus of talent and problem solving from various disciplines to pull it off.

References

Pitt, Sensors and Journalism, Tow Center for Digital Journalism, May 2014
Chicago's Array of Things project
Sun Sentinel's “Above the Law” series

This blog post is written by Lily Bui, M.S. Candidate in Comparative Media Studies at the Massachusetts Institute of Technology, and is cross-posted from her personal blog.

This year, I had the privilege of presenting at the Open Knowledge Festival in Berlin, Germany, an international meeting of the minds that brings together advocates for open science, data, government, and societies. While still not completely recovered from jetlag, I wanted to make sure I wrote this down before it all disappeared into the ether! (Though, most of it is conveniently logged in the session's Etherpad.)

My Thursday session focused on sensor journalism — using sensors to generate data that could be used to support journalistic inquiry. Long story short, it is clear that for many common issues that journalists pay attention to (both local and global), there is a sensor that could potentially help generate and collect data.

Any emergent field deserves its share of constructive criticism to shape it and to keep it balanced as it evolves. So, in our workshop, we ran a handful of thought experiments that stimulated some critical thinking about the sensor journalism space.

Needless to say, it was incredibly inspiring to see people from so many different countries thinking collaboratively about both local and global problems that sensor tools + data might help solve. They held constructive debates and discussions, asked challenging questions, and in the end helped me gain a better understanding of what still needs to be done for the field.

In the course of merely one hour, a group of people who were initially unaware of sensor journalism transformed into one that was able to identify various world problems that sensor data could help investigate, as well as suggest approaches to doing so. Now, if that's not inspiring, I don’t know what is.

[The feedback I received from this conference also warranted the start of an e-mail group and official hashtag for #SensorJourno to connect those wanting to stay in the conversation. Please join in if it suits your interests!]

Format

There were four groups total, and each group was given a scenario involving sensors + storytelling for which they needed to find a solution and define a working strategy. One group in particular (Group D) was given the sole purpose of playing devil's advocate — to think about the risks that come with more experimentation with sensor-based reporting. You’ll find the scenarios below.

GROUP A:

Your city, Dendropolis, was once known for its abundant and lush forest. Over time, a company called TIMBER, INC. began to cut down the trees for paper production, construction materials, etc. Now, there is very little forest left. Recently, a law was passed to restrict deforestation in Zone X. However, an anonymous source leaves you a tip that TIMBER, INC. is secretly still cutting down trees in Zone X. So far, there is no way to prove this. How could you use sensors to monitor this kind of activity? What kind of sensors are they? How would you collect the data? How would you use the data to tell a story? Would the story be a video, podcast, newspaper article, blog post, phone app, game, etc.?

This group thought of creating a visible border around the protected area/trees, so if the border was breached, then an alert could be sent off to the correct authorities to act upon it. Another idea was to place noise sensors in the area so that if the sound of a chainsaw was detected, alerts would be sent to the proper authorities or anyone else who signed up to receive them.

[Incidentally, this strategy actually resembles that of the current project called Rainforest Connection.]

GROUP B:

You are part of a research group that has designed an open source sensor which vibrates a few hours before earthquakes. This technology could possibly work as an early warning system for people who live in areas that are at risk for earthquakes. However, the technology is useless unless the information it provides reaches people who need it. Design a plan for action that includes media outreach. How will you spread the word and make this sensor available to as many people who need it as possible? How will you teach people how to use it? How would you use the data to tell a story? Would the story be a video, podcast, newspaper article, blog post, phone app, game, etc.?

For the sensor itself, the group came up with a list of questions to start thinking about their deployment strategy:

– How do we get it to potential earthquake zones?

– How do we get information/alerts from teh sensor to the local people?

– What kind of information?

Then for installing the sensor in earthquake zones, the group decided that the best way forward might be to build and maintain relationships with local communities, NGOs with extensive distribution networks, fire departments/emergency first responders, and local news outlets. The sensors would be installed in public buildings in identified earthquake zones.

In order to generate feedback to the community if an earthquake is detected by the sensors, the group thought of using relationships with local radio or TV as well as a broadcast SMS alert to spread the word in case of emergency. They even thought of tying the alerts to Ushahidi, an open source crisis alert system, to collect data from the ground and broadcast it to people.

GROUP C:

Identify an issue that you can help track using smart phone sensors. What is the best way to report the data you collect — an app, game, etc.? After the data is collected, how can you convey it in a way that will make sense to people that the issue affects?

The phone sensors that this group identified were as follows: GPS, microphone, accelerometer, gyroscope, battery/ambient temperature, barometer, bluetooth, light. (There are actually many more.)

Inspired by a group member who was visiting Berlin from the Sahara, where sandstorms can be prominent, this group designed an app meant for parents who were concerned about air quality outside, especially while their children were playing outdoors. With this app, parents can submit a photo of air quality/conditions to an app, which will be aggregated in an open, public database. The data would be visualized in a way that would show where air quality/pollution was the worst so that parents could make better decisions about where to take their kids for recreation.

GROUP D:

Your role is to play devil's advocate. While sensor journalism project emerge, what possible risks need to be considered? What problems can you anticipate might arise from more experimentation?

– Limited accessibility & representativeness. Mostly who can afford the sensors will acquire them, which excludes many demographic groups.

– Transparency/privacy. People may not be aware of what data they are giving away.

– Quality of the collected data. Problems with calibration of sensor tools, improperly collected data, working with unreliable data.

– Generalization. Analysis of data from local samples cannot be generalized to make greater statements about the wider issue (e.g. data from monitoring the water quality of one private well does not reflect the water quality of an entire state, region, country, etc.).

– Lack of clarity as to who can start a sensor journalism initiative — journalists, scientists, both?

– Bias. Journalists can shift media attention toward specific issues tha thave reached a critical mass, while ignoring other issues that could require attention.