1!After Motes and Multihop: Mobile Phones !and the Global Mobile Sensor Network!Andrew T. Campbell, Dartmouth College!Wireless sensor networks have driven many great innovations over the last decade - represents a very active area of on-going research!The mote have been a superb platform for research!But, challenges !remain!• Not ubiquitous!• Energy problem!• Scaling (cost and performance) problem!• Event unpredictability!• Donʼt have economy of scale!Importantly, sensor networks donʼt impact our everyday lives. Why?$People are out of the loop!Thatʼs all just changed because of this … !… or this !Embedded sensors:!• 3-axis accelerometer!• Proximity sensor!• Microphone!• Camera!• GPS!• Bluetooth!Solves many outstanding problems:!• ubiquitous, energy problem, ! has economy of scale, scales, etc.!Nokia 6210 Navigator ! or, in the future ..!Embedded sensors:!• 3-axis accelerometer!• Digital compass!• Microphone!• Camera!• GPS!• Bluetooth!2!The cool green “emotional” phone !Embedded sensors:!• 3-axis accelerometer!• Proximity sensor!• Digital compass !• Pollution/air quality sensor!• GSR “emotion sensor”!• RFID/NFC !• Microphone!• Camera!• GPS!• Bluetooth!Youʼll be able to answer lots of questions and build new cool$applications with these phones!What is my personal air quality like today? Or, the air quality of my neighborhood, $school, town, or city?!How stressed is the city this morning?!How do social “conversation” networks evolve?!Work on audio networks by Tanzeem Choudhury (Dartmouth)!Where are my friends and what are they doing right now?!3!I know what you are thinking!Sounds like an Orwellian nightmare!!Big challenges: trust, privacy, security are critical issues (David Kotz, Dartmouth)!Now imagine 1 billion “sensor !enabled mobile phones” !scattered across the planet!people are in the loop!This will lead to …!The global mobile sensor network!Societal scale !sensing!I know what you are thinking!You canʼt cover a volcano with mobile phones!!4!On second thoughts, phones are getting cheaper and we have a ready supply of adventurous!grad students ;-)!We are at a crossroads!My position for this talk$The mobile phone (and not the mote) will serve as the main platform for sensing innovation over the next 5 years.$Your mobile phone will sense your surroundings, learn your behavior (what you do, where you go and how you interact with people and your environment), and help you navigate your day.$Collectively, mobile phones will form societal scale sensor networks in support of community, urban, and global sensing applications and problem solving. $You donʼt have to believe me, $just wait for it to evolve!The new “mote” (minus the multihop problems)!5!We started in 2005 to study people-centric sensing!Characteristics of existing mote networks!• Small-scale, short-lived, mostly-static!• Application-specific!• Multi-hop wireless!• Very energy-constrained!• Mobility not an issue or driving factor!• People out of the loop!Characteristics of $People-Centric Sensing!• Large-scale, long-lived, mostly-mobile!• Application-specific!• Multi-hop wireless!• Very energy-constrained!• Mobility not an issue or driving factor!• People out of the loop!Characteristics of$People-Centric Sensing!• Large-scale, long-lived, mostly-mobile!• Application-agnostic!• Multi-hop wireless!• Very energy-constrained!• Mobility not an issue or driving factor!• People out of the loop!Characteristics of $People-Centric Sensing!• Large-scale, long-lived, mostly-mobile!• Application-agnostic!• No multi-hop wireless !• Very energy-constrained!• Mobility not an issue or driving factor!• People out of the loop!6!Characteristics of$People-Centric Sensing!• Large-scale, long-lived, mostly-mobile!• Application-agnostic!• No multi-hop wireless !• Periodic recharging !• Mobility not an issue or driving factor!• People out of the loop!Characteristics of$People-Centric Sensing!• Large-scale, long-lived, mostly-mobile!• Application-agnostic!• No multi-hop wireless !• Periodic recharging!• Mobility is a driving factor!• People out of the loop!Characteristics of$People-Centric Sensing!• Large-scale, long-lived, mostly-mobile!• Application-agnostic!• No multi-hop wireless !• Periodic recharging!• Mobility is a driving factor!• People in the loop!Characteristics of$People-Centric Sensing!• Large-scale, long-lived, mostly-mobile!• Application-agnostic!• No multi-hop wireless !• Periodic recharging!• Mobility is a driving factor!• People in the loop!• Security and privacy of data critical!People-centric sensing application domains!Andrew Campbell, et al, “The Rise of People-Centric Sensing”,! IEEE Internet Computing, July/August 2008"I know what you are thinking!7!Static sensor networks$donʼt scale to large $areas, sensing coverage $is costly, performance $doesnʼt scale either, $and events are $unpredictable in$time and space.!Public sensing gains $scalability and $sensing coverage $by using people $opportunistically as $mobile sensors !The beauty is that the infrastructure already exists (i.e., people and $the global cellular network)!People-centric sensing is based on an “opportunistic sensing paradigm” and an “interaction model” that captures interaction between people, and, between people and their surroundings!local peering!Internet!(backend data!sharing)!GIS and crowd !sourcing data!Emerging sensing paradigms!Participatory !sensing (UCLA)!Hybrid !approaches!Opportunistic sensing (Dartmouth)!Next frontier in sensing$is “people-centric” urban sensing!Ron Fricke, timescape is a day in the life of a city (edited version)8!Remainder of my talk!• Three people-centric sensing applications we developed!• BikeNet (personal/public sensing)!• CenceMe (social sensing app)!• SoundSense (personal sensing app)!• Need for open sensing/comms software for mobile phones!• Wrap up!The BikeNet system – sensor bikes!BikeNet - sensor bikes!Shane Eisenman, et al, “The BikeNet mobile sensing system for cyclist experience mapping”, ACM SenSys '07 We can answer many questions from sensor
View Full Document
Unlocking...