P1252122134kPUsF

1
Wireless Sensor Networks for Habitat
Monitoring BY Alan Mainwaring
and David Culler Intel Research, Berkeley Intel
Corporation Joseph Polastre, Robert Szewczyk and
David Culler EECS Department University of
California at Berkeley John Anderson College of
the Atlantic Bar Harbor, Maine
                       Wireless biological
sensors placed in nests
A student uses the 'petrel peeper', a portable
infrared video system, to inspect a burrow.
Students at Maine's College of the Atlantic are
using the data to learn more about Storm Petrels
in their native habitat.
2
OUTLINE

1. Requirements for Habitat Monitoring are
   Established
2. Design Requirements for Hardware, Sensor Network
   and Capabilities for Remote Data Access and
   Management are Determined
3. A System Architecture is Proposed to Address
   these Requirements
4. A Specific Instance of the Architecture is
   Presented for Monitoring Seabird Nesting
   Environments and Behavior
5. Results and Recommendations are Discussed

3
Habitat Monitoring Questions

• What is the usage pattern of nesting burrows over
  24-72 hour cycle when one or both members of a
  breeding pair may alternate incubation duties
  with feeding at sea?
• What changes can be observed in the burrow and
  surface environmental parameters during the
  course of the approximately 7 month breeding
  season(April-October)?
• What are the differences in the
  micro-environments with and without large numbers
  of nesting petrals?

4
Habitat Monitoring Requirements

• Minimal disturbance in monitoring
• Simple, Easy deployment
• Economical Method for Conducting Long Term
  Studies

FOR EXAMPLE ...
5
10.

• Existing Land-Atmosphere Observation Systems
• Requires local power utilities
• Requires miles of power cables
• Expensive(100k)
• Takes weeks to deploy
• Requires flat locations
• Measurements are limited to tower footprints

6
Remote Sensing Requirements

• Internet Access
• Hierarchical Network (wireless capability)
• Sensor Network Longevity (9-12 months)
• Operating off-the-grid (bundled energy supplies)
• Management at-a-distance (PDA Query a Sensor,
  Adjust Param, Locate Devices)
• Inconspicuous operation
• System Behavior
• In-situ Interactions
• Sensors and Sampling
• Data Archiving

7
Proposed System Architecture
8
Implementation Strategy
9
Sensor Network Node
10
Sensor Board
11
Energy Budget
Panel Size in2 Total Watt Hours per Day x
____1_____ Peak Winter Hours
.065W / in2
12
Expected Lifetime
13
GREAT IDEAEXCEPT THE SIZE OF THE MICE WAS TOO
LARGE TO FIT IN PETREL BURROWS!
14
(No Transcript)
15
Patch Gateway
FIRST CHOICE CerfCube Strong Arm embedded
System Running Linux and 802.11b single hop w/
CompactFlash 802.11b adapter 1GB Storage and
Solar Panel 2.4GHz antenna w/Range of 1000
feet HOWEVER 802.11b requires bidirectional
link in MAC and has TCP/IP Overhead And had 2
required 2 orders of magnitude more power than a
mote
16
Base Station Installation (DBMS) User Interfaces
including a PDA
17
RESULTS AND RECOMMENDATIONS
18
OTHER APPLICATION SERVICES
LOCALIZATION, TIME SYNCRONIZATION AND SELF
CONFIGURATION
19
DATA SAMPLING AND COLLECTION
20
Communications

• Power Efficient Communication Paradigms
  must include routing algorithms, medium access
  algorithms and managed hardware access tailored
  for efficient network communication while
  maintaining connectivity when required to source
  or relay packets. Future above ground nodes will
  have harvesting capabilities to enable node hop
  routing

21
Health Status Monitoring
Diagnostics such as voltage at periodic rates as
opposed to only during transmission (or
Intelligent Schemes)
22
LATER ADVANCES
23
NEW WEATHER BOARD DESIGN

• Mica SensorboardThe mica sensorboard can have
  these sensors
• temperature
• photo
• magnetomer
• accelerometer
• microphone
• sounder (buzzer)

24
MICA 2
25
CALIBRATION
26
NEW PACKAGING