Soundscape

Citizen Science & Sonification

Soundscape is a citizen science app that turns noise pollution into short songs. Users collect noise levels of their city and that data gets aggregated for city officials and the public to utilize. I worked with two UX designers to create this experience.

Role
UI/UX Designer
Date
October 2015 to November 2015
While At
Carnegie Mellon Human-Computer Interaction Institute
With
Prerna Pradeep & Paige Pritchard

Problem & Setup

  • A given user, “Dale” in Baltimore’s IT department, is tasked with figuring out noise pollution levels across the city.
  • Seventh grade student, “Aya”, has an Android phone, uses Instagram frequently, and is interested in the helping science in some way.
  • We are contacted by Dale to design an app that helps him and users like Aya.

Solution & Application Prototype

A sound-level recording app that turns decibel levels into short tunes. User recordings help cities dynamically tackle noise pollution.

My Role

  • I, like my teammates, brainstormed the potential directions for the app.
  • In our research phase, I interviewed my sister to gain more insight to younger audiences’ perception of citizen science.
  • I ensured that we were compliant with Google’s Material Design.
  • I created the sonification mapping and audio files in GarageBand.

Process

When our team formed, we individually wrote down all of our thoughts about the problem and potential solution. This quick brainstorming led to a round of sharing and early design directions based on smartphone limitations and our own insights.

  • Since Aya uses an Android, we decided to use Google’s Material Design framework.
  • Motivation to use the app needs to be a focal point of our designs.
  • We can record decibel levels, but the user must initiating the recording (i.e. we cannot use passive recording).
  • Recording and storing audio files wouldn’t be useful for the first version of the app nor is it apparent that it would be useful. If we went that route, that kind of functionality needs more of focus on privacy.
  • Social media can provide insights to our design, but the next Facebook isn’t going to be a sound recording app. We will accent our designs with social patterns, but we won’t wrap a complex social network around our designs.

One of our storyboards exploring how a user might form a team. We imagined each user belonging to only one group. This was also when we were calling the project "SoundCheckr".

Design Path

Each phase of our design was only as complex as it needed to be. Work that gets progressively more complex allows you to focus on the level of detail at hand and get continual feedback on the elements you’re designing.

  1. Set initial design and research directions
  2. Research similar applications
  3. Speed-dating/scenarios
  4. Laid out our general flows
  5. Wireframes with minimal color usage (Balsamiq)
  6. Interaction/flow tested with selected wireframes (InVision)
  7. High fidelity mockups created (Sketch)
  8. Interaction/flow tested with all mockups (InVision)

Paige whiteboarding lo-fi wireframes. Before this, we used a wireframe planning technique employed by one of Basecamp’s designers (on my suggestion). This allowed us to outline our pages and flow as a simple, connected graph.
The basics of logging data. Left: when users record a decibel level, they also can mark if it was inside or a nuisance. Right: if a user shares an individual data recording to Instagram, we prepare the composite image for them. We refined the idea after looking at other people trying to share their life-logging data on Instagram by taking screenshots.

On Motivation

The hardest design decisions we made focused around defining the motivations for using Soundscape. From our experience with citizen science communities, we saw that motivation was either intrinsic to science (e.g. “I use X because it helps science”) or extrinsic to the result of science (e.g. “I use Y because it helps science which helps the world be better”). To gain some insight, I interviewed my sister about her app usage, social media usage, and motivations for studying science. From that conversation and our previous thoughts, our team compiled a list of potential motivating factors for Soundscape:

  1. Intrinsic for those who just like citizen science
  2. Authority-Based for those who get told to do it by a teacher
  3. Altruistic for those who like helping the scientists or the city
  4. Badge-Addicts for those who have an absurd appreciation for badges
  5. Social for those who use the app to help friends
  6. Artistic / Artistic-Social for those who like generative art from the measurements

Each motivation, taken as the locus of design, would lead to wildly different design patterns. We explored badges, social connections, and generative art paths the most.

Badges The design element we went back and forth on the most was badges. (Oh units of glory, why do you have to be so hard?) We initially wanted badges to help motivate users to continue logging data and complete actions we defined.

Our results indicate that badges are useful to motivate students while activity streams have the potential to activate students.

—Santos et al. in “Evaluating the Use of Open Badges in an Open Learning Environment”

Good thing we were thinking about an activity stream.

Findings reinforced previous research that digital badges function differently according to the type of learner. Results indicated a generally positive view of badges in English courses, though levels of intrinsic motivation to earn the badges increased for high expectancy-value learners only.

 — Reid, Paster, & Abramovich in “Digital badges in undergraduate composition courses: effects on intrinsic motivation”

Ok. So badges need to be able to be achieved for some level of internal motivation. Further research into this study shows that the badges should be related to performance rather than just external motivators.

DIY.org served as an inspiration for our badge system.

Ultimately, we decided to leave badges for a future version of the application. For the first version, badges themselves shouldn’t be what we rely on as the primary motivator nor did we want to spend our resources on some sort of badge management UI.

Social Groups We also explored how motivation might work with teams. Early on we decided that we wouldn’t create a social media app around sound pollution, but we could still learn from social media design decisions.

Our first proposal had exclusive and private groups users could join. Each user could join one group and users would only see the activity of the people in their group. We liked this decision for its simplicity and team-based dynamics. However, when thinking through use cases, it seemed as though the group-based interaction would be overly restrictive.

We abandoned this idea and re-focused our design to work better with existing social media applications. A more complex social aspect could be designed in a future version, but not version one.

Generative Art We ultimately decided to focus our efforts on creating generative art. User recordings could be motivational if they were turned into a song after a few recordings. Badges and social dynamics have been explored before. We wanted to explore a less-taken path.

Left: me and Prerna testing prototype soundscape noises. Right: refined play modal for soundscapes.

There are two times a user can share their activity on social media channels. Each time a user makes a recording they are prompted to share a composite image of their decibel level, location, and the scientists they are helping.

Then, after a user has 15 recordings, we create a soundscape: a 15 note tune that maps the user’s decibel levels to a scale of MIDI notes. By mapping the decibel readings to a musical scale, we guarantee that the random numbers from the data are represented as notes that work well together. The mapping can be done an infinite number of ways. But, for our prototypes of soundscapes, I chose the scale of C and played the same set of notes with different MIDI instruments in Garageband. Previously, I had used an life-logging app called Reporter that, among other things, records the decibel level when you make a data entry. My real data generated these final soundscapes.

Creating interesting, sharable content for the Soundscape users allows them to make a post on social media easily. Interest from their post can drive new users to our application.

Given a longer time frame, we would have reached out to other citizen science app users to discover their motivations via interviews and speed-dating exercises.

Back to Top