The Healthy Cities project was conceived to address the frustration among city residents about how little they know about the "healthiness" of their neighborhood, and how difficult it seems to be to make a noticeable contribution. Initially, we conceptualized "healthiness" as environmental health, though we have significantly broadened our definition after interviews and surveys of Berkeley residents to include aspects of city health such as public school conditions, air quality, effective minimum wage, maintenance of houses and streets, unemployment, individual health, racial diversity, pedestrians, public events, and more. We are building public displays that make city health more visible and encourage change by making residents feel like their contributions are valuable. We also want to encourage community participation in the displays, either indirectly by seeing changed habits reflected in the displays, or directly by collecting data or posting comments.

Specifically, we interviewed six East Bay residents to get a qualitative understanding about their perceptions of a healthy city. We followed up with a survey of over 145 residents from the cities of Berkeley, Oakland, El Cerrito, and Richmond. The surveys were both on-line as well as on paper. The paper surveys were distributed to residents in front of public areas such as Post Offices, Farmers' Markets, etc. Our interviews suggested broad categories if issues that affect a person's perception of city health. The survey questions delved deeper into the broad categories that surfaced from the personal interviews.

This resulted in us picking top 10 categories that came up as being the most important to the residents to be reflected in a display about the health of their city. Among the top 10, we chose ones that recurred in the interviews as well as fared well in the questionnaires (4.0 or above out of 5). Within these top choices, we picked two that were feasible: Resource Management and Light Pollution.

We developed designs for public displays of both of these categories. The Resource Management tracking display would be local, exhibiting the effect of people recycling cans and seeing the results immediately. The Light Pollution display would be a global perspective, parsing through night images of the city of Berkeley to determine where electricity was used the most.

A preliminary interface has been designed for the Resource Management display, which will sense (through load cells) the change in weight whenever cans were recycled into a particular recycling bin. As the weight increases, a visual "meter" rises to give the audience a sense that they contributed to the "health" of the city. The interface being deployed on a Sony Clio currently has been made to work in a simulated environment, where the addition of a can is simulated with the clicking of a button on the touchscreen.

A preliminary interface has been designed for the Light Pollution display, which senses, via computer vision, the amount of light given off by lights in the city of Berkeley at night. Multiple web cameras are used to collect images every few minutes. These images are analyzed for brightness characteristics and aggregated across cameras. The resulting brightness information is overlaid on a map of Berkeley and presented on a screen to users.

Publications based on this project:
http://kettle.cs.berkeley.edu/ambient/10
Berkeley EECS Research Journal (BERJ) (To Appear)