50 PERVASIVE computing Published by the IEEE CS n 1536-1268/09/$26.00 © 2009 IEEE CONTENT SHARING T oday’s mobile phones are smarter than ever: they now take and pro- cess pictures and videos, issue messages and email, access the Web, allow games on demand, and play music. More people around the world take their phones everywhere they go, using them in a variety of environments and situations to perform a whole range of different tasks. In India, for example, more people access the Internet from their phones than from a PC, a scenario that will certainly play out across the globe in the years to come. 1 Most mobile phones include a variety of sens- ing components. By expanding this capability, we can derive some interesting sensing modalities—for ex- ample, scrutinizing local envi- ronments to detect and reduce pollution or using medical ap- plications to tackle other prob- lems on a societal scale. In this article, we dis- cuss experiences and lessons learned from deploying four mobile sensing applications on off-the-shelf mobile phones within a recreational framework called Mob- Sens that contains elements of health, social, and environmental sensing at both individual and community levels. We describe the main components of our applications, which facili- tate logging and external communications. We also outline the challenges faced when building and testing these applications and describe our strategies for overcoming them. MobSens Prototypes Mobile sensing—also known as “participatory sensing,” 2 “urban sensing,” 3 or “participatory urbanism” 4 —enables data collection from large numbers of people in ways that weren’t previously possible. Using mobile phones has several advantages over unattended wireless sensor networks for environmental sensing applications: • Mobile phones can provide coverage where static sensors are hard to deploy and main- tain, and large numbers of cell phones already exist around the world, providing the physical sensing infrastructure. • Deploying the sensing hardware and pro- viding it with network and power requires signi icant effort in other sensor network- ing systems. The availability of more pow- erful operating systems and the transfer of standardized programming languages on ever-smaller computing platforms have spurred the recent development of software applications for mobile computers, includ- ing Symbian (www.symbian.com), Google Android (http://code.google.com/android), Microsoft Mobile (www.microsoft.com/ windowsmobile), and iPhone (www.apple. com/iphone). • Such systems can beneit from local com- munities as the driving element for environ- mental sensing. This approach, sometimes referred to as “citizen science,” uses mobile sensor technology to help individuals person- ally collect, share, compare, and participate Four mobile sensing applications that work on off-the-shelf mobile phones contain elements of health, social, and environmental sensing at both individual and community levels. Eiman Kanjo, Jean Bacon, and Peter Landshoff University of Cambridge David Roberts Nokia MobSens: Making Smart Phones Smarter Authorized licensed use limited to: University of Texas at Arlington. Downloaded on January 12, 2010 at 19:38 from IEEE Xplore. Restrictions apply.