0 2 4 6 8 10 12 14 16 18 20 10 -2 10 0 10 2 10 4 Frequency (Hz) abs(FFT(.)) Frequency plot, vertical sensors at L1-L5 V2 V4 V13 V7 V9 0 1 2 3 4 5 6 7 8 9 10 -8 -6 -4 -2 0 2 4 6 8 Time (sec) Acceleration (mg) Time plot, vertical sensors at L1-L5 V2 V4 V13 V7 V9 Wireless Sensor Networks for Structural Health Monitoring Overview • Ambient vibrations of the structure are monitored and used to determine the health status of the structure. • With a Wireless Sensor Network, low cost monitoring is possible without interfering with the operation of the structure. Accelerometer Board Challenges • High Fidelity Data • High Frequency Sampling with Low Jitter • Time Synchronized Sampling [FTSP] • Large-scale Multi-hop Network [Mint] • Reliable Command Dissemination [Broadcast] • Reliable Data Collection [Straw] Silicon Designs 1221L ADXL 202E ADXL 202E Silicon Design s 221L Range -2G ~ 2G -0.1G ~ 0.1G System noise fl oo r 200(μG/ √Hz) 30(μG/√Hz) Price $10 $150 • Two measurement axis each with two accelerometers • Thermometer, 16bit ADC, Low-pass filter • On-board Digital Signal processing • Calibration for manufacturing variation and temperature - 0.99 0.19 - 0.73 1.00 0.74 First Vertical Mode of Vibratio n Estimated results match with a FE model of the bridge (SAP) Deployment at the Footbridge Berkeley SF Bay mid-span quarter-span 5 9 Base Station 260ft 16ft 2 7 1 13 10 3 8 4 12 11 14 Sukun Kim*, Shamim Pakzad + , David Culler*, James Demmel*, Gregory Fenves + , Steve Glaser + , Martin Turon # * EECS, UC Berkeley + CEE, UC Berkeley # Crossbow Software Architecture Best-effort Single-hop Communication Broadcast MintRoute FTSP Low-level FLASH BufferedLog Straw Sentri (Application Layer) • When sampling, only necessary components are turned on to reduce jitter • Straw provides reliable data collection • Selective-NACK is used – complexity is drawn from the sender (mote) to the receiver (PC) • Rate-based control • Pipelining increases channel utilization Deployment at the Golden Gate Bridge 1 st mode 2 nd mode 3 rd mode Vertical Frequency (Hz) 1.35 1.79 11.47 Damping Ratio 0.055 0.02 0.043 Horizonta l Frequency (Hz) 2.37 7.87 11.91 Damping Ratio 0.26 0.16 0.123 • Nodes on the main span and the south tower • Distance between nodes on the west span is either 100ft or 50ft • Exposed to strong and salty wind and fog 8 nodes 56 nodes 1125 ft 4200 ft 500 ft 246 ft SF (south) Sausalito (north) Node, Battery, Antenna Rusting of C-clamp Base station in Tower 0 100 200 300 400 500 600 -20 0 20 Tim e (sec) Accel(m g) Tim e and Frequency plots,V erticalsensors, s284n62 45 50 55 60 65 -5 0 5 Tim e (sec) Accel(m g) 0 5 10 15 20 25 30 35 40 45 50 0 1 2 frequency (H Z) P S D (m g/Hz) 0 0.5 1 1.5 2 2.5 0 1 2 frequency (H Z) P S D (m g/H z) 0 100 200 300 400 500 600 -50 0 50 Tim e (sec) Accel(m g) Tim e and Frequency plots,V erticalsensors, s284n45 45 50 55 60 65 -10 0 10 Tim e (sec) Accel(m g) 0 5 10 15 20 25 30 35 40 45 50 0 2 4 frequency (H Z) P S D (m g/Hz) 0 0.5 1 1.5 2 2.5 0 2 4 frequency (H Z) P S D (m g/Hz) (a) Vertical, Quarter span North of the South Tower (b) Vertical, Quarter span South of the North Tower Bandw idth versus H op C ount 0 200 400 600 800 1000 1200 1400 0 10 20 30 40 50 H op C ount Bandw idth (B/s) Vibration Data from the Footbridge