Beyond the Blue Dot

Can a map generate empathy by humanizing data through narrative? A prototype that turns census data into short stories about the people who live and work in a place.

Role: UX, visual design, prototyping, code
Product: Prototype in a Jupyter notebook
Timeline: Spring 2020, four months
Advisors: Anthony Vanky. Reader: Violet Whitney

Beyond the Blue Dot landing page with a map of New York — **Fig. 1**Choose a place and a type of person, and read their story.

01Context

Urban analysis tends to work top-down, in aggregates. This project asked a more human question: who lives here, who works here, and what do their lives look like?

A planner working on a transit project wants to disrupt people’s days as little as possible, so they need to know who lives and works nearby, how they get around and what they earn.

Getting that from GIS takes days. Community meetings help, but they don’t represent everyone. There was no quick way to see a place at the scale of one person.

Diagram comparing the old research process with the new one — **Fig. 2**Today, planners piece this together from meetings, surveys and days of data cleanup. The tool cuts it to three steps: pick a place, pick a person, read the story.

02Approach

Pick a location and a profile, and the tool builds a narrative about a typical or outlier resident or worker from national data, so it works anywhere in the US. The persona and flow came from conversations with practicing planners, who needed it to be fast, intuitive and easy to fit into their work.

Early prototypes tested a split screen of map and story. Testers found the split clear but the navigation hard to follow, and the stories not personal enough. The final version leaned into avatars for each profile, and a half-mile boundary, the planning standard for a comfortable walk, to frame each person’s day.

User persona for an urban planner — **Fig. 3**The persona, from conversations with planners.

User flow from location to persona to narrative — **Fig. 4**A deliberately short flow: location, person, story.

Wireframes of the tool’s screens — **Fig. 5**Wireframes, from landing page to generated story.

Diagram of the tool’s structure from introduction to data stories — **Fig. 6**How the pieces connect.

Screens from the first, intermediate and final iterations — **Fig. 7**First, intermediate and final iterations.

03Outcome

A working proof of concept, built in Python on New York’s Lower East Side, generating stories for residents from the Census Bureau’s American Community Survey, the LEHD workforce data, the National Household Travel Survey and the Google Places API.

Next steps would have added workers, travel data and natural language processing for richer stories. The name plays on the blue dot that stands for you in Google Maps. Here, the dot is a person.

Fig. 8The working prototype, generating a resident’s story.

Next project

Election data for 5M voters

Say hello.

Working on something for the built environment? I’d love to hear about it.

LinkedIn Resume (PDF)