Research Notebook

The Rise of the Expert Amateur:
Citizen Science and Neo-Volunteerism


By Eric Paulos
  • If you cannot measure it, you cannot improve it. -- Lord Kelvin
In the Living Environments Lab, we conduct research focused on societal problems. More specifically, we focus on the critical intersection of human life, our living planet and technology.

One direction of our work is shifting mobile phones from mere "communication tools" to "personal-super-computer-radio-stations-with-sensors" that can enable everyday people to become "citizen scientists" as they collect and share measurements of their environment. This new usage model for mobile phones will lead to important contributions in four research areas that:
  • Improve the science literacy of everyday citizens through active participation in basic scientific data collection and use of scientific principles;
  • Provide professional scientists and stakeholders with access to richer data sets for modeling, analyzing and advancing both fundamental and applied knowledge about people and ecosystems;
  • Develop new usage models and user experiences for the mobile phone as a tool for promoting transparency and enabling grassroots participation in local community and civic government policy making; and
  • Create a greater public awareness and understanding of the relationships between humans and the natural environment particularly with regard to sustainability and environmental issues.
Through crowd-sourcing and new open-development platforms for mobile phones, we believe ordinary citizens around the world can become a central, driving element for collecting, reporting, interpreting and improving our health and our environment.

Challenges in citizen science

There are challenges that have to be overcome, of course. When collecting data, for instance, why and how will people be motivated to participate? How and what type of data will be collected? When will samples be taken?  How will problems of sensor accuracy, drift and calibration be addressed? What are reasonable types of sensors to use?

We must address ways the data is reported. How will we explain to our "citizen scientists" things like the range, accuracy and calibration of our sensors? Is the data better shared as raw numbers, or should it be interpreted and abstracted in a visualization such as a moving bar graph, meter or some other pictorial device?

How will the collected data be shared--on a personal mobile phone or in a public space? How will privacy be assured? How will our data be archived, preserved and authenticated, and what techniques will be used to insure that our data are valid?

Finally, what tools and techniques will facilitate change--productive debate and ultimate positive social benefit? How will people use the data to argue for and against various hypotheses?

We also have to study the best ways to invite and encourage active participation in both data collection, and the use of data to develop real solutions to human problems.

Data collection at the grassroots

Ordinary citizens have long been agents of change, in grassroots movements ranging from neighborhood watch campaigns to political revolutions. Using citizens to collect scientific data is less well known, but there are precedents.
One successful, long-running citizen data-collection effort has been the National Audubon Society's Christmas Bird Count, which since 1900 has performed an annual census of birds in the Western Hemisphere.

Other movements in citizen data-collection indicate that many people have an intense interest in participating in scientific surveys. In the "Great World Wide Star Count," observers around the world are encouraged to go outside, look skywards after dark, and report the count of stars they see. In effect, they're measuring light pollution. Another is "Project BudBurst," where people submit time-stamped images of when flowers in their city bloom. This helps collect data on climate and pollen counts.

Our work builds on these kinds of existing citizen science projects but adds the ability to collect environmental data using sensors equipped mobile phones.

Taxi Drivers in Ghana and the 'Prius Effect'

One of our early experiments was aimed at measuring air pollution from automobiles in Accra, the capital of Ghana. In 2007, we recruited taxi drivers and students to collect air quality data. The drivers and students were provided with sensors for the pollutants carbon monoxide, sulfur dioxide and nitrogen dioxide, along with a GPS unit (Figure 1). A series of visualizations (Figure 2) from the sensor data as well as interviews resulted in a series of design recommendations and significant behavioral changes in our participants to improve air quality.

Perhaps the most surprising results from this study were not that air urban quality varies wildly over time and space, but that the participants recruited to collect this dataset began to experience what we referred to as "the Prius effect." Drivers of Toyota's Prius automobile get feedback on their real-time fuel mileage, and according to published reports, many have altered their driving behavior in an attempt to "score" 100 miles-per-gallon. Our participants in Ghana--exposed to real-time values for air quality--likewise developed strategies to minimize their exposure to poor air quality by altering their routes and times of travel across the city of Accra.

A Pocket-Sized Solution

After reviewing the results of our Ghana experiment, we integrated a series of air-quality sensors with a mobile phone (Figure 3). This system measured and reported carbon monoxide, nitrogen dioxide, ozone, temperature and humidity and visualized the aggregate data on a public website.

While many research efforts in mobile sensing are focused on system-level challenges, such as integrating sensors and addressing hardware power issues, developing tools to share and visualize the data is a much larger challenge. Successful tools help persuade users to positively change their behavior while avoiding scare tactics.

A civic-citizen hybrid

Local governments and other agencies can play a role in data collection, and we  obtained first-hand experience during a collaboration with the City of San Francisco to collect air quality readings using municipal street sweepers. We installed sensors to measure CO, NOx, ozone, temperature, relative humidity and speed on 12 street-sweepers. As the vehicles cleaned city streets, we were able to obtain real-time readings of environmental conditions across the city. (Figure 4)

That approach may seem to move away from using data collected by average citizens, but we think it's a realistic "hybrid" method where "citizen participation" includes citizens as well as civic agencies and public infrastructure. Such approaches are also great scaffolding techniques towards larger-scale public data collection.

Moving citizen science indoors

Citizen science can also be used to study indoor air quality. In one experiment, we designed a system integrated with Apple's iPhone to measure and visualize particulate matter--the small particles of pollution found in air and a common problem in Pittsburgh. We also explored ways to use the visualizations to persuade people in a social network to modify behavior such as smoking.

Through a study of 14 households in Pittsburgh, we found that this system, called inAir, increased awareness of air quality, promoted behavioral changes that improved air quality, and demonstrated the persuasive power of sharing information and strengthening social bonds. (Figure 5) One of our participants, after viewing her air quality data compared with others, actually quit smoking!

This is consistent with studies that have shown people are heavily influenced by the behaviors and actions they expect or know others to be performing.

Monitoring water conservation

Water is our most precious and most rapidly declining natural resource. More than 1 billion people do not have access to safe drinking water and more than 2.5 billion lack adequate sanitation. World health agencies estimate that more than 5,000 people--many of them children--die every day from water-related illnesses.

But having access to safe, plentiful water is influenced by a complex set of variables, including political, industrial and environmental constraints. And it's not just the developing world that faces water scarcity. Rapidly growing metropolitan areas in the western United States also struggle to provide water to citizens with diminishing returns and steadily increasing costs.

We wanted to extend our environmental awareness research beyond air quality, so we applied similar techniques to monitoring water conservation using low-cost sensors (less than $40) with persuasive displays.

Prototypes with two display styles--numeric and ambient--showing individual and average water consumption were installed in public and private spaces such as faucets and showers. The numeric display presents current water usage to the nearest tenth of a gallon while the visualization presents the same information in the form of a "traffic light." (Figure 6)

 Our work demonstrated that showing individual users the extent of their water usage helped promote conservation in real ways. It also inspired users to become curious about water usage in other contexts, such as dishwashing and lawn watering.

Areas for Future Research

"Citizen science" efforts have the potential to provide a vast and exciting new dataset for scientists and urban planners to use in improving public health and quality of life.

Involving non-experts in both data collection and community data analysis also has the potential to radically change public perceptions about urban planning and approaches to problem solving within neighborhoods and across cities.

Inviting public participation builds trust and encourages elected officials to make decisions that are transparent, and in the best public interest. It also allows policy makers to share their best practices with other communities.

Finally, successful citizen science projects could effect positive changes in society and help citizens of all ages participate in democracy and improve their understanding of the environment. The potential for grassroots efforts to emerge from such work and produce solutions to current local and global health and environmental issues is very real. 

We hope our work will help motivate future research that empowers everyday people to learn, understand and improve their health and well-being, and broaden their awareness of their environment.

Eric Paulos is an assistant professor of human-computer interaction and joined the SCS faculty in 2008. Founder and director of CMU's Living Environments Lab, his research focuses on how technology intersects both human life and the living planet. He also serves as adjunct faculty in CMU's Entertainment Technology Center.

A native of California, Paulos earned his doctorate in electrical engineering and computer science in 2001 from the University of California at Berkeley. Prior to joining Carnegie Mellon, he served as senior research scientist at Intel Research in Berkeley, Calif., where he founded the Urban Atmospheres research group.

More information about citizen science and Paulos' other research is available at www.paulos.net.
For More Information: 

Jason Togyer | 412-268-8721 | jt3y@cs.cmu.edu