Urban Sensing

From UrbanWiki

A participatory sensing system is one that allows individuals and communities to collect, share and organize information through data collection using cell phones and other mobile platforms, to enable them to make a case for change and to explore and understand their life and relationship with the environment. It leverages commodity hardware like mobile phones, which can capture images, audio, GPS and cell tower location, and can interface with other environment monitoring or health monitoring sensors to allow for direct or indirect, but in situ, measurements of phenomena of human concern such as public health, urban planning, individual wellness, civic concerns etc.

Participatory Sensing builds on CENS' experience in Embedded Networked Sensing, but is unique in that it is truly human in the loop sensing, where participants become part of the system and carry and wear sensors on them. Users also participate in the classification, audit and analysis of data.

Some Participatory Sensing applications are:

1)Diet Monitoring where participants use mobile phones to capture images of their food at meal time for dietary recall and discussion with their dietitians.

2) Walkability Study where participants living in a neighborhood use mobile phones to capture geo-tagged images of objects in the neighborhood that need attention to improve walkability in the area e.g.: broken sidewalks, difficult crossings etc. These can be used to prepare reports for a neighborhood council. Neighborhood Noise Mapping and Asset mapping are similar applications.

3) Personal Environment Impact Report is an application where users use the GPS on mobile phones to log their pollution impact and exposure with the environment.

Participatory Sensing offers a lot of promise but it has some very challenging problems - especially in the areas of privacy, participation, relevance and integrity.

1) Privacy is being addressed through selective sharing mechanisms including resolution control and also by having a safe zone for data before it is shared to the public.

2) Participation can be encouraged with feedback to the user and also by enabling humanly understandable visualizations of data.

3) Activity and location analysis enable us to understand the context in which data is collected.

4) Participation, performance, and mobility analysis helps us to get relevant data by fine-tuning our tasking techniques for participation.

Data quality, credibility, and feedback are key enablers for participation. We believe that having built-in methods to promote these initiatives are important in participatory sensing.