SATIRE: A Software Architecture for Smart AtTIRE R. Ganti, P. Jayachandran, T. F. Abdelzaher, J. A. Stankovic (Presented by Linda Deng)

Slides:



Advertisements
Similar presentations
Phil Buonadonna, Jason Hill CS-268, Spring 2000 MOTE Active Messages Communication Architectures for Networked Mini-Devices Networked sub-devicesActive.
Advertisements

Mobile GIS.
Global MP3 Geoffrey Beers Deborah Ford Mike Quinn Mark Ridao.
MP3 proposal. Template  Title  Your group name and group members  Application overview  Main functions  Detail description  Timeline and task assignment.
Prestige 2.0 Commercial Thermostat with Internet Connectivity.
Decentralized Reactive Clustering in Sensor Networks Yingyue Xu April 26, 2015.
What is TinyOS? An open-source OS Designed for low-power wireless devices, such as sensor networks, ubiquitous computing, personal area networks, smart.
DATA COLLECTION USING ZIGBEE NETWORK Timothy Melton Moscow, ID.
Wireless Sensor Networks Craig Ulmer. Background: Sensor Networks n Array of Sensor Probes ( ) n Collect In-Situ Data about Environment n Wireless.
CSE 5392By Dr. Donggang Liu1 CSE 5392 Sensor Network Security Introduction to Sensor Networks.
Autonomous Helicopter: James Lyden Harris Okazaki EE 496 A project to create a system that would allow a remote- controlled helicopter to fly without user.
MIAMI Medical Instrument Analysis and Machine Intelligence
Wireless Sensor Networks for Habitat Monitoring
What is a Wireless Sensor Network (WSN)? An autonomous, ad hoc system consisting of a collective of networked sensor nodes designed to intercommunicate.
The Internet & The New IT Infrastructure Chapter 9.
1 Introduction to Wireless Sensor Networks. 2 Learning Objectives Understand the basics of Wireless Sensor Networks (WSNs) –Applications –Constraints.
TRADING OFF PREDICTION ACCURACY AND POWER CONSUMPTION FOR CONTEXT- AWARE WEARABLE COMPUTING Presented By: Jeff Khoshgozaran.
Mica: A Wireless Platform for Deeply Embedded Networks Jason Hill and David Culler Presented by Arsalan Tavakoli.
Watchdog Confident Event Detection in Heterogeneous Sensor Networks Matthew Keally 1, Gang Zhou 1, Guoliang Xing 2 1 College of William and Mary, 2 Michigan.
IEEE Standardized radio technology for low power personal area networks Joe Polastre January 14, 2004.
Generic Sensor Platform for Networked Sensors Haywood Ho.
Hardware Design Experiences in ZebraNet [ZSLM04] - Princeton University Sensys ‘04 Presented By: Jay Taneja.
A Framework for Patient Monitoring A. L. Praveen Aroul, William Walker, Dinesh Bhatia Department of Electrical Engineering University of Texas at Dallas.
Analysis of the Performance of IEEE for Medical Sensor Body Area Networking ECE 5900 Computer Engineering Seminar Instructor: Dr. Chigan Huaming.
Wireless Sensor Networks for Habitat Monitoring Jennifer Yick Network Seminar October 10, 2003.
Intel ® Research mote Ralph Kling Intel Corporation Research Santa Clara, CA.
Security and Digital Recording System Students: Gadi Marcu, Tomer Alon Number:D1123 Supervisor: Erez Zilber Semester:Spring 2004 Final Presentation.
Radio-Triggered Wake-Up Capability for Sensor Networks Soji Sajuyigbe Duke University Slides adapted from: Wireless Sensor Networks Power Management Prof.
GPS. Meet the Bad Elf GPS Pro The Bad Elf GPS Pro offers the first Apple approved GPS with an informative user interface.
INTRUSION DETECTION SYSTEMS Tristan Walters Rayce West.
Building an Application Server for Home Network based on Android Platform Yi-hsien Liao Supervised by : Dr. Chao-huang Wei Department of Electrical Engineering.
Introduction To Wireless Sensor Networks Wireless Sensor Networks A wireless sensor network (WSN) is a wireless network consisting of spatially distributed.
Home Health Care and Assisted Living John Stankovic, Sang Son, Kamin Whitehouse A.Wood, Z. He, Y. Wu, T. Hnat, S. Lin, V. Srinivasan AlarmNet is a wireless.
Hardware implementation and Demonstration. Synapse RF26X We started off with Synapse RF26X 10-bit ADC Up to 2 Mbps Data Rate 4K internal EEPROM 128k flash.
Learning Micro-Behaviors In Support of Cognitive Assistance AlarmNet is a wireless sensor network (WSN) system for smart health-care that opens up new.
Clemson University Bluetooth™ Enabled Accelerometer Tracking (BEAT) Technology for Leg Ulcer Patients Aleksey Shaporev 1, Vladimir Reukov 1, Chengyi Tu.
Ambulation : a tool for monitoring mobility over time using mobile phones Computational Science and Engineering, CSE '09. International Conference.
Mobile Databases: a Selection of Open Issues and Research Directions Authors: Rachid Guerraoui et al. Sources: SIGMOD Record, 33(2), pp.78-83, 2004 Adviser:
Intelligent Shipping Container Project IMPACT & INTEL.
Electrical & Computer Engineering Presentation by: Andrew Frieden Matthew Weydt Nick Setzer Tyler Dunn PDR 2012 – Team Gong Self Sustaining Sensor System.
Presenter : Hyotaek Shim
WSN Done By: 3bdulRa7man Al7arthi Mo7mad AlHudaib Moh7amad Ba7emed Wireless Sensors Network.
ZebraNet Hardware Design Experiences in ZebraNet Presented by Zuhal Tepecik.
© 2005 Victor Shnayder – Harvard University 1 CodeBlue: A Wireless Sensor Network for Medical Care and Disaster Response Victor Shnayder Harvard University.
SDP 2010 Team Goeckel PDR Stephan Adams Ghaida Al-Jermozi Jessica Lau Daniel Smith UMass CampusVision.
EWatch: A Wearable Sensor and Notification Platform Paper By: Uwe Maurer, Anthony Rowe, Asim Smailagic, Daniel P. Siewiorek Presenter: Ke Gao.
TECHNICAL SEMINAR ON SMART QUILL Presented By:- your name.
SunSPOT Wireless Modules Gurdip Singh and Shravanthi Kallem Pervasive Sensor Network Laboratory Computing and Information Sciences.
Introduction to Wireless Sensor Networks
1 EnviroStore: A Cooperative Storage System for Disconnected Operation in Sensor Networks Liqian Luo, Chengdu Huang, Tarek Abdelzaher John Stankovic INFOCOM.
Infocom’07 Authors:Liqian Luo, Chengdu Huang, Tarek Abdelzaher John Stankovic Presented By Rohini Kurkal Under Guidance of Dr.Bin Tang.
Wireless Sensor Networks for Habitat Monitoring Intel Research Lab EECS UC at Berkeley College of the Atlantic.
Development of a wireless telemetry system for model rocketry.
6-10 Oct 2002GREX 2002, Pisa D. Verkindt, LAPP 1 Virgo Data Acquisition D. Verkindt, LAPP DAQ Purpose DAQ Architecture Data Acquisition examples Connection.
Advanced Computer Networks Topic 2: Characterization of Distributed Systems.
Computer Science 1 TinySeRSync: Secure and Resilient Time Synchronization in Wireless Sensor Networks Speaker: Sangwon Hyun Acknowledgement: Slides were.
System Architecture Directions for Networked Sensors Jason Hill, Robert Szewczyk, Alec Woo, Seth Hollar, David Culler, Kris Pister Presenter: James.
Spotlight: Personal Natural Resource Consumption Profiler Younghun Kim, Zainul Charbiwala, Akhilesh Singhania, Thomas Schmid, Mani B. Srivastava Networked.
Interfacing External Sensors to Telosb Motes April 06,2005 Raghul Gunasekaran.
Internet of Things. IoT Novel paradigm – Rapidly gaining ground in the wireless scenario Basic idea – Pervasive presence around us a variety of things.
Adaptive Sleep Scheduling for Energy-efficient Movement-predicted Wireless Communication David K. Y. Yau Purdue University Department of Computer Science.
Mechanical layer Logic layer Sensors layer Poll for sensor state Return polled state & also do interrupts in immergency situations Sensors checking suspension,
© 2008 RoboRubik11 Technical Communication Skills Practicum Constraint Analysis.
Introduction to Wireless Sensor Networks
Luis E. Palafox and J.Antonio Garcia-Macias CICESE – Research Center 2009 Proceedings of the 4 th international conference on Wireless pervasive computing.
Software Architecture of Sensors. Hardware - Sensor Nodes Sensing: sensor --a transducer that converts a physical, chemical, or biological parameter into.
Created by :- prashant more prashant more. INTRODUCTION Bluetooth is wireless high speed data transfer technology over a short range ( meters).
COMPSCI 110 Operating Systems
Telemedicine: Issues in Mote Based Remote Patient Monitoring
Tiers vs. Layers.
Presentation transcript:

SATIRE: A Software Architecture for Smart AtTIRE R. Ganti, P. Jayachandran, T. F. Abdelzaher, J. A. Stankovic (Presented by Linda Deng)

Goals Personal wearable monitoring devices – Archive user activities for later lookup – Monitor at-home patients Transparent/non-obtrusive Longevity of service Flexible/modular system architecture

Hardware considerations Sensing: Acceleration + GPS sensors – Activity + location monitoring – Can reconstruct activities w/ hidden Markov models Memory/storage: Less than 1 MB? – 1 MB can log 6 hours of disconnected 50 1 byte/sample – Logging in flash memory Persistence allows for operating at low power when no activity is recorded

Hardware considerations (cont.) Communication: Wireless – Collaboration b/w pieces of attire – Information upload: < 1 minute near access mote? Processing: Microcontroller sufficient Energy supply: Enough for a “season”? – AA batteries only good for a few days – But power-saving schemes + lithium batteries?

MicaZ chosen Off-the-shelf microcontroller-based device 2-axis accelerometer + pluggable GPS module 512 KB flash radio AA batteries But limited by off-the-shelf components…

Typical scenario

System architecture (mote) Application layer: Receives data from sensor layer Filter layer: Common data processing algorithms OS layer: Data storage, upload, synchronization

System architecture (PC) Parsing layer: Processing of raw sensor data Interpretation layer: Offline algorithms for parsed data Application layer: User interfaces for apps

Data collection and storage For walking, needed at least 12 Hz sampling  25 Hz for each accelerometer axis How to increase disconnected operation time? Truncate filter – Effectively reduce accelerometer range Stillness filter – Just record number of samples of stillness

Data upload Need sufficient rate, transparency, reliability Beaconing – Motes send periodic beacons – Base allows one mote to communicate – NACKs for reliability

Data synchronization Synchronize motes on different parts of body To maintain temporal correlation: beaconing – During reconstruction, same beacon # = same time Beacon values recorded in flash – But beacons occur much less frequently than data –  Acceptable overhead

Power management Typical mote lifetime = < 7 days? But we want it to last for 3 months! Duty cycle scheme – Sleep when still, wake periodically to check stillness Low-power operation possible 90% of time? Assuming 5% duty cycle during low-power op…  Battery life of 7 weeks ≈ 3 months? Same life for GPS if no fixes during stillness…

Reconstruction of activity logs How to determine type of activity performed? Feature vector approach inaccurate in practice Hidden Markov model – Machine learning technique requiring training data – Fairly accurate for predefined activities – But only for pre-trained “general” activities…

User interface Archive availability provided by Internet Web server + DBMS to answer queries – Centralized Or data can be hosted on home computer…

Privacy and security Secure communication channels? Authorization for accessing info archive? Not really addressed…

Evaluation Sending data rate = 30 Kbps – Due to serial port bottleneck b/w access mote + PC? – And lack of buffer space on jacket mote…  Read/writes to flash interleaved Truncate/stillness filters reduced flash needed If disconnected op is long, sampling rate becomes approximate – No effect from clock drift when near base Location tracking via GPS worked fine

Stationary activity identification 2 different users 10 minutes of each activity, 1 minute training

Non-stationary activity Just 1 user

ID without specific training “General” activities can be pre-trained Or user can specially train own jacket…

Experiment example

Conclusions Flexible/modular architecture for smart attire HMMs for activity identification Prototype with accelerometer and GPS sensors Bottlenecks of off-the-shelf hardware? Privacy?