permasense iii sensor networks in extreme environments jan beutel, stephan gruber, christian...

PermaSense III Sensor Networks in Extreme Environments

Jan Beutel, Stephan Gruber, Christian Tschudin, Lothar Thiele

Objectives

• Establishing a highly reliable and dependable wireless infrastructure for sensor and actuator networks in extreme environmental conditions– Terrain movement, micro-seismics, temperature– Long periods of unsupervised operation– Limited power sources– Maximizing data yield

• Combining sensor and actuator functionality

• Paving the road for future applications in early warning

New Horizons

• Multitude of (new) sensors– Precision movement detection– Resistivity tomography– Micro-seismic events– Imaging

• Large variety of time scales– Real-time actuation and sensing for fast processes– Long term for slow processes

• Remote locations– No possibility of physical repair/update– No infrastructure– Sensor node in a pocket

Understanding the local and global perspective

Trift glacier, images courtesy of VAW, ETH Zurich

And when catastrophes happen

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

Tomorrow’s challenge – improving today’s practice

• Seismic/resistivity tomography

actuation andsensing

Current Status and State-of-the-Art

• PermaSense in MICS phase 2– Operational field sites since 07/2008– 15 nodes, 150 uA power consumption, 2 min sampling rate– Constant rate, simple sensors, limited context

• Sensor network research– Post “gold-rush” era

• The field is consolidating• Foundations have been made• Early (mis)conceptions have been understood and are being revised

– Sustainability and real technology transfer– Many open (hard) problems

• Permafrost monitoring and modeling– Manual interaction in field studies, fragile measurement systems– Quality data (to connect field/lab/simulation) missing for harsh

conditions

Comments from the proposal review

• The proposal could be stronger with meaningful involvement of more theorists to pursue the research questions that arise in deployments.

Practice Theory

Grow the team? Diversify?

Shift focus?

• System design – architecture– Design of extremely reliable systems by

constructive/compositional methods

• Resource allocation and arbitration– Locality of processing, storage– Communication requirements

• Efficient signal processing algorithms– Local processing to filter out interesting events– Remote triggering of sensors and actuators

• Natural processes and hazards– New insight through better data – Automated analysis and alerting

Th

eory

Colla

bora

tors

Scientific challenges – (MICS) theory collaborations

Thiele, Beutel, Mattern

Observability by Design

M. VetterliA. Geiger, Geodesy

ETH

Climate, geo-hazard, civil engineering

Impact of PermaSense

• Scientific– Driver application for more theoretical oriented technology

research– Technology showcase and catalyst for multidisciplinary work – New opportunities for environmental research, unprecedented

data

• Societal– Hot topic: Global warming, climate change, natural environment– Economic relevance: Natural hazard prevention, (re-)insurance

business– Media attention

– Joint geo-science publications

[NICOP2008]

Cooperation

• Close relation to– Swiss Experiment– Phase 3 proposals

• Environmental Monitoring and SensorScope II• Eternal Sensor Data Store• Observability by Design

• Based on proven cooperation between– University Zurich (Stephan Gruber)– University Basel (Christian Tschudin)– ETH Zurich (Jan Beutel, Lothar Thiele)

• Continuing to leverage MICS developed technology

• Co-funding and close collaboration with the Federal Office of the Environment (FOEN)