1. ©2012 University of CaliforniaPrepublication Data March 2012Berkeley Sensor & Actuator Center Spring 2012 Wireless Sensing Applications for Critical Industrial Environments Fabien Chraim Prof. Kris Pister

2. ©2012 University of CaliforniaPrepublication Data March 2012Berkeley Sensor & Actuator Center Spring 2012 2/12 Richmond Refinery Wireless Umbrella

3. ©2012 University of CaliforniaPrepublication Data March 2012Berkeley Sensor & Actuator Center Spring 2012 3/12 Smart Fence – Proposed System Architecture 50uA / 20mA 10uA / 1 mA 10uA / 2mA

4. ©2012 University of CaliforniaPrepublication Data March 2012Berkeley Sensor & Actuator Center Spring 2012 4/12 Sequential Hypothesis Testing Used in radar, clinical trials, channel selection… Decision making as samples are collected Two hypotheses –H0: no (or low) activity –H1: high activity [1] A.Wald and J.Wolfowitz, “Optimum character of the sequential probability ratio test,” Ann. Math. Statist., vol. 19, pp. 326339, 1948. [2] Fellouris, G.; Moustakides, G.V.;, “Decentralized Sequential Hypothesis Testing Using Asynchronous Communication,” Information Theory, IEEE Transactions on, vol.57, no.1, pp.534-548, Jan. 2011 0 660ms 0 210ms H1 transmitted No transmission H0 transmitted 0 660ms

5. ©2012 University of CaliforniaPrepublication Data March 2012Berkeley Sensor & Actuator Center Spring 2012 5/12 Sequential Hypothesis Testing – Example Outputs 0 8 seconds 0 1.6 seconds

6. ©2012 University of CaliforniaPrepublication Data March 2012Berkeley Sensor & Actuator Center Spring 2012 6/12 Smart Fence - Continuation Solar energy scavenging Enclosure redesign Air quality monitoring (H2S)

7. ©2012 University of CaliforniaPrepublication Data March 2012Berkeley Sensor & Actuator Center Spring 2012 7/12 Valves

8. ©2012 University of CaliforniaPrepublication Data March 2012Berkeley Sensor & Actuator Center Spring 2012 8/12 Valve Position Monitoring Why? –Detect abnormalities (alarms) –Better understanding of pressure gradients (plant-wide) Goals –Instrument a variety of valves (ball, butterfly, gate, globe) –Report position wirelessly –Accuracy of +/- 5 degrees (or 5%) –No calibration / Easy installation –Lifetime of ~5 years –Cheap

9. ©2012 University of CaliforniaPrepublication Data March 2012Berkeley Sensor & Actuator Center Spring 2012 9/12 Valve Position Monitoring – Experimental Setup

10. ©2012 University of CaliforniaPrepublication Data March 2012Berkeley Sensor & Actuator Center Spring 2012 10/12 Valve Position Monitoring – Data (quadrature)

11. ©2012 University of CaliforniaPrepublication Data March 2012Berkeley Sensor & Actuator Center Spring 2012 11/12 Valve Position Monitoring – Data (rotary)

12. ©2012 University of CaliforniaPrepublication Data March 2012Berkeley Sensor & Actuator Center Spring 2012 12/12 Project Outlook Large fence deployment (Intrusion + gas sensing) On-site valve testing Motor Vibration sensing (pumps, small electric motors…)