The PermaSense ProjectLow-power Sensor Networks

for Extreme Environments

Jan Beutel, ETH ZurichNational Competence Center in Research –

Mobile Information and Communication Systems

PermaSense – Alpine Permafrost Monitoring

• Cooperation with Uni Basel and Uni Zurich

PermaSense – Aims and Vision

Geo-science and engineering collaboration aiming to:– provide long-term high-quality sensing in harsh

environments– facilitate near-complete data recovery and near real-time

delivery– obtain better quality data, more effectively– obtain measurements that have previously been impossible– provide relevant information for research or decision

making, natural hazard early-warning systems

To Better Understand Catastrophic Events…

Eiger east-face rockfall, July 2006, images courtesy of Arte Television

PermaSense Deployment Sites 3500 m a.s.l.

A scientific instrument for precision sensing and data recovery in environmental extremes

PermaSense – Key Architectural Requirements

• Support for ~25 nodes

• Different sensors– Temperatures, conductivity, crack

motion, ice stress, water pressure– 1-60 min sensor duty-cycle

• Environmental extremes– −40 to +65° C, ΔT ≦5° C/min– Rockfall, snow, ice, rime,

avalanches, lightning

• Near real-time data delivery

• Long-term reliability– ≧99% data yield– 3 years unattended lifetime

Relation to other WSN projects• Comparable to other environmental

monitoring projects– GDI [Szewczyk], Glacsweb [Martinez],

Volcanoes [Welsh], SensorScope [Vetterli], Redwoods [Culler]

• Lower data rate

• Harsher, higher yield & lifetime

• Data quality/integrity

PermaSense Technology

PermaSense – System Architecture

BackendSensor network Base station

• Powerful embedded Linux (Gumstix)

• 4 GB storage, all data duplicated

• Solar power (2x 90W, 100 Ah, ~3 weeks)

• WLAN/GPRS connectivity, backup modem

PermaSense – Base Station Overview

Base Station with Stacked Mikrotik WLAN Router

Gumstix Verdex

TinyNodeGSM

WLAN Router

EMP Protectors

IP68 Enclosure and Connectors

PermaSense – Sensor Node Hardware

• Shockfish TinyNode584– MSP430, 16-bit, 8MHz, 10k SRAM, 48k Flash– LP radio: XE1205 @ 868 MHz

• Waterproof housing and connectors

• Protective shoe, easy install

• Sensor interface board– Interfaces, power control

– Stabilized measurements

– 1 GB memory

• 3-year life-time– Single battery, 13 Ah

– ~300 µA power budget

Precision Sensing – Architectural Considerations

• One platform vs. family of devices? Make vs. buy?– Capabilities for multiple sensors– Extreme low-power– High quality data acquisition (ADC resolution on MSP430 not sufficient)– Reliability in extreme environment

• Based on existing platform, Dozer low-power protocol– 0.167% duty-cycle, 0.032mA

• Modular, accommodating different sensors on one platform• Single, switchable serial bus architecture• Strict separation of operating phases (TDMA)

[Burri – IPSN2007]

Precision Sensing – Sensor Interface Board I

Extension: One Serial Bus

• (Power) control using GPIO• Optimized for low-power duty cycling

Precision Sensing – Sensor Interface Board II

External Memory

• Data buffering• End-to-end validation

Precision Sensing – Sensor Interface Board III

Power Control and Protection

Design for Testbed Integration

Dozer Low-Power System Software Integration

• Dozer ultra low-power data gathering system– Beacon based, 1-hop synchronized TDMA– Optimized for ultra-low duty cycles– 0.167% duty-cycle, 0.032mA

• System-level, round-robin scheduling– “Application processing window” between data transfers and beacons– Custom DAQ/storage routine

time

jitter

slot 1 slot 2 slot k

data transfercontention

window beacon

timeslot 1 slot 2 slot k

Application processing window

[Burri – IPSN2007]

Validation Using Simultaneous Power Traces

PermaDozer – Power Performance Analysis

1/30 sec1/120 sec

148 uA average power

PermaSense Sensors

• Sensor rods (profiles of temperature and electric conductivity)

• Thermistor chains

• Crack meters

• Water pressure

• Ice stress

• Self potential

• Data: Simple sensors, constant rate sampling, few integer values

PermaSense Sensors – Sensor Rod Example

Sensors Rod Data –Temperature and Resistivity

PermaSense Sensors – Crack Meter Example

Crack Meter Data – Happy Geo-Science Partners

Of Testing, Bugfixing and Learning it the Hard Way…

Power Optimization – A Squeeze with Implications

• Regulator uses 17uA quiescent current• Bypass used to shutdown regulator -> ~1uA in standby

• No Bypass increases ADC accuracy: std dev 0.8844 -> 0.0706

The Result – Power Quality Increases Data Accuracy

AfterBefore

Of Sensors, Placement and Power Consumption

Testing – Physical Reality Impacting Performance

Watchdog Resetstime

Storage Duration

DAQ Duration

Ambient Temperature

• Networked testbed: Deployment-Support Network

• Power profiling

• Power trace verification [Woehrle – FORMATS2009]

• Temperature cycling

• Sensor calibration

• Rooftop system break-in

PermaSense – Test and Validation Facilities

[Beutel – EWSN2007]

But Nothing Can Replace Real Field Experience

Battery Voltage & Temperature

Lost PacketsSystem Outages

Node Health Data – Indispensable for Analysis

Nodes Reset

System Outages# Lost Packets

Network Disconnect

Understanding the Exact Causes is Hard...

• Missing EMP protector – broken radio chip?• If so is it snow/wind or lightning induced?• Long-term software bug – buffer overflow; timing?• Misunderstood “feature” of the implementation?• Hardware breakdown?• Energy-supply dependent?• Spring-time behavior based on the environment?• Cosmic rays at 3500 m a.s.l. ?• Can external effects trigger the reed contact on the reset?• …

Probably not feasible to test/reproduce this in the lab!

Installation – Timeconsuming and Demanding

PermaSense installation 2007, images courtesy of Arte Television

Pro’s and Con’s of Solar Panels

Demo – Live Data at http://tik42x.ee.ethz.ch:22001

Key PermaSense Challenges

System Integration Correct Test and Validation

Actual Data Interdisciplinary Team

PermaSense Achievements – Current Status

• Dozer integration successful– Best-in-class low power– DAQ vs. COM power consumption– Extreme installation effort (time)– Relative relaxation of multihop requirement

• Continuous data since mid July 2008

• Media attention

• First joint geo-science publications

[Gruber etal. – NICOP2008]

148 uA average power

Acknowledgements

• All PermaSense collaborators – Uni Basel, Uni Zurich, EPFL, SLF, ETH Zurich, …

• Further reading– J. Beutel, S. Gruber, A. Hasler, R. Lim, A. Meier, C. Plessl, I. Talzi, L. Thiele, C. Tschudin, M Woehrle, M.

Yuecel: PermaDAQ: A Scientific Instrument for Precision Sensing and Data Recovery in Environmental Extremes. IPSN 2009.

– J. Beutel, S. Gruber, A. Hasler, R. Lim, A. Meier, C. Plessl, I. Talzi, L. Thiele, C. Tschudin, M Woehrle, M. Yuecel: Operating a Sensor Network at 3500 m Above Sea Level . IPSN 2009.

– A. Hasler, I. Talzi, J. Beutel, C. Tschudin, and S. Gruber, “Wireless sensor networks in permafrost research -concept, requirements, implementation and challenges,” in Proc. 9th Int’l Conf. on Permafrost (NICOP 2008), vol. 1, pp. 669–674, June 2008.

– M. Woehrle, C. Plessl, J. Beutel and L. Thiele: Increasing the Reliability of Wireless Sensor Networks with a Unit Testing Framework . EmNets 2007.

– J. Beutel, M. Dyer, M. Yücel and L. Thiele: Development and Test with the Deployment-Support Network . EWSN 2007.

– K. Aberer, G. Alonso, G. Barrenetxea, J. Beutel, J. Bovay, H. Dubois-Ferriere, D. Kossmann, M. Parlange, L. Thiele and M. Vetterli: Infrastructures for a Smart Earth - The Swiss NCCR-MICS Initiative. Praxis der Informationsverarbeitung und Kommunikation, pages 20-25, Volume 30, Issue 1, January 2007.

– N. Burri, P. von Rickenbach, and R. Wattenhofer, “Dozer: ultra-low power data gathering in sensor networks,” in Proc. 6th Int’l Conf. Information Processing Sensor Networks (IPSN ’07), pp. 450–459, ACM Press, New York, Apr. 2007.

Project Webpage: http://www.permasense.ch