Wireless Sensor Networks for Habitat Monitoring Alan Mainwaring, Joseph Polastre, Robert Szewczyk, and David Culler, June 2002 Presented by Team 8: Feng Kai and Michael Thomsen, September 2004

Outline Habitat monitoring Network Requirements System Architecture Hardware and Design Results

Why? Why are we interested in Habitat Monitoring? Focus attention on a REAL network Some problems have simple concrete solutions, while others remain open research areas Application driven approach separates actual problems from potential ones, and relevant issues from irrelevant ones Collaboration with scientists in other fields helps define broader application space as well as scientific requirements 1 – REAL network requirements, problems and solutions

Scientific Motivation Questions What environmental factors make for a good nest? How much can they vary? What are the occupancy patterns during incubation? What environmental changes occurs in the burrows and their vicinity during the breeding season?

Scientific Motivation Methodology Characterize the climate inside and outside the burrow Collect detailed occupancy data from a number of occupied and empty nests Validate a sample of sensor data with a different sensing modality Augment the sensor data with deployment notes (e.g. burrow depth, soil consistency, vegetation data) Try to answer the questions based on analysis of the entire data set

Scientific Motivation Problems Seabird colonies are very sensitive to disturbances Solution Deployment of a sensor network The impact of human presence can distort results by changing behavioral patterns and destroy sensitive populations Repeated disturbance will lead to abandonment of the colony

Field Stations Great Duck Island James San Jacinto Mountains Reserve Almost 1 km2 (237 acre) remote island Coast of Maine Large Breeding colonies of Leech’s Storm Petrels and other seabirds James San Jacinto Mountains Reserve Small 0,1 km2 (29 acre) ecological preserve Near Idyllwild, California (about 100 km east of LA) Studying nest boxes, and ecosystems over time

Application Requirements General Requirements Internet access Hierarchical network (local networks over longer distances) Sensor network longevity (at least 9 months) Operating off-the-grid (battery or solar cells) Remote management (personnel just available 2-3 months) Inconspicuous operation (should not disrupt natural processes or behaviors under study) System behavior (stable, predictable and repeatable behavior) In-situ interactions (during deployment and maintenance) Sensors and sampling (light, temperature, IR, humidity, pressure) 1 – need for remote intercations support 2 – explain (p2.b.2)

System Architecture Patch Sensor Node Network (power) Sensor Patch Gateway (low power) Sensor Node (power) Transit Network Base-station (house-hold power) Base-Remote Link Data Service Internet Client Data Browsing and Processing

Sensor Nodes Sensor Nodes Separated into two logical components: Computational module (MICA) Sensing Module (Weather board) Sensor nodes communicate and coordinate with one another Battery powered Small (5 x 3.8 x 1.25 cm3) Mechanically robust Weather-proofed (but ventilated to avoid data distortion)

Sensor Nodes Computational Module Mica Platform Atmel Atmega 103 micro controller @ 4 MHz 512 kb Flash memory 916 MHz, 40 kbps radio 2 AA batteries w/boost converter

Sensor Nodes Sensor Module Mica Weather Board Temperature Photoresistor Barometric pressure Humidity Passive IR (Thermopile) Designed to coexist with other sensor boards Low variation when interchanged with other sensors of same model IR sensor and thermistor and photo resistor to detect cloud cover IR sensor: detect occupancy, measure temp of nearby object (bird in nest)

Power Management Sensor Node Power Limited Resource (2 AA batteries) Estimated supply of 2200 mAh at 3 volts Each node has 8.128 mAh per day (9 months) Sleep current 30 to 50 uA (results in 6.9 mAh/day for tasks) Processor draws apx 5 mA => can run at most 1.4 hours/day Nodes near the gateway will do more forwarding 75 minutes

Power Management Compression Does the compression require more power than transmission? Even if it does it may be worthwhile to compress if data must pass through many nodes 2 - 4 times reduction by combining: Delta compression and Standard compression algorithm: Is it worthwhile to implement compression?

Gateway Function Hardware Relay sensor patch network data to base station Coordinate the activity within the sensor patch Provides additional computation and storage Gateway to base station distance apx 100 m Hardware Strong-Arm based embedded system (CerfCube) Runs an embedded version of Linux Compact Flash memory 2.5 W supplied by solar cells and lead-acid battery inches

Base Station Base Station Laptop: Coordinates the sensor patches Provides database service Connects to the Internet: Great Duck Island: Two-way satellite James Reserve: T1 line Base Station

The Gizmo In situ Interaction (Planned) PDA-sized device Communicate directly with the sensor patch Direct communication to the mote Provides fresh readings Interactively control the network (sampling rates, power management etc) Useful during initial deployment and retasking of the network

Communication Routing Routing directly from node to gateway not possible Approach proposed for scheduled communication: Determine routing tree Each gate is assigned a level based on the tree Each level transmits to the next and returns to sleep Process continues until all level have completed transmission The entire network returns to sleep mode The process repeats itself at a specified point in the future

Network Retasking Network Retasking Initially collect absolute temperature readings After initial interpretation, could be realized that information of interest is contained in significant temperature changes Full reprogramming process is costly: Transmission of 10 kbit of data Reprogramming application: 2 minutes @ 10 mA Equals one complete days energy Virtual Machine based retasking: Only small parts of the code needs to be changed Network can be reprogrammed once per day of all 9 months if no other task is done

Conclusion Paper conclusion Future Two small scale sensor networks deployed at Great Duck Island and James Reserve (one patch each) Results not evaluated No calibration was done, development of auto-calibration procedure suggested Future Develop a habitat monitoring kit

Problems 2002 Deployed Network No new science about petrel breeding behavior, many insights into how to build sensors that would yeild that science Base Station laptop must run unattended (problem with unscheduled system reboots (mostly fixed)) Temperature Sensors: Measured temperature inside the enclosure, rather than ambient temperature Good correlation with Cost Guard measurements on cloudy days Batteries: Only operational down to 2.5 V (expected down to 1.6V)

Problems 2002 Deployed Network Rain affecting the humidity sensor and short-circuiting the battery So far only mild challanges – low data rates and not really extreme environment

Additional References R. Szewczyk, J. Polastre, A. Mainwaring, D. Culler: ”Lessons from a Sensor Network Expedition”, UC Berkerly, Jan 2004 J. Polastre: ”Design and Implementation of Wireless Sensor Networks for Habitat Monitoring”, Research Project, 2003 www.greatduckisland.net

2003 Network New design Lithium battery based New enclosure Burrow, Ground, D-Cell, 2002 Ground