GLACSWEB Sensor Networks & GLACSWEB Kirk Martinez IAM Group ECS
What are Sensor networks? large numbers of sensors integrated into one information system sensor data must reach a server the networks are not typically tcp/ip –ad-hoc networking is needed
Why sense the environment? Studying global environmental changes requires massive data collection… Coarse but large systems exists for weather prediction but deep oceans, ice caps, microclimates are less covered
Challenges reusable cheap components, self- configuring robust communications autonomous and adaptive behaviour extensible ad-hoc networking low power to enable data fusion, data must be accessible: need a semantic sensor web
Who is doing this too? Jet Propulsion Lab –use the term "sensor webs"
Nextwave Envisense Centre Glaciers - Glacsweb Flooding - Floodnet Coastal erosion - Seacoas
Glacsweb Aims and Objectives To monitor glacier dynamics as a contribution to the study of ‘Global Warming’ and glaciology To develop a pervasive sensor network To produce generic components and expertise useful in other environments
GLACSWEB Pervasive computing Supraglacial data Subglacial data
Test Site
map
Field site
expected movement in a year Base Station reference station Ice Sediment 13m 10m 7m 3m
Sometimes the only way to get your equipment in is to employ a helicopter….
Understanding the Subglacial Environment Ground Penetrating Radar Drilling & till sampling Borehole camera
GPR interpretation T1 T2 T0 T3 L1 L2 L3 10m0 N Subglacial Channel? Ice flow “strong” reflection
hot water drilling
Ice hole depth 10m0 N Ice flow 72m Water drained 70m 76m 80m Water filled 55m 65m 82m 56m 62m 63m 75m>38m GPR survey 50MHz “strong” reflection Subglacial Channel Probes in ice Probes in till Till sampled
System overview
Probes Plastic case (10cm long) PIC microcontroller Radio Transceiver A/D and amplifiers Batteries Sensors: tilt, temp, pressure Real time clock
Probe subsystems
Probe pressure tests in Oceanography
Base station Measure snow levels, temp, box tilt, bat V Radio links to Ref station and probes DGPS Large power supply and solar panel
Base Station
Reference station Small Linux server ISDN line Backup of all data Gets all DGPS data Deposits data on Southampton server
3D Positioning (1 Probe) Distance from PR to RC = time taken for signal to reach RC Position of PR calculated via trilateration
PCBs
HoursProbeBase StationReference Station 0000Data logging Data logging Data logging Data loggingGPS logging 1600Communication Data transfer 2000Data logging-- System Timeline System on times:Data logging- 11 seconds GPS logging- 20 minutes Communication- 180, 300, 600 seconds (PR, BS, RS) Data transfer- when completed
testing and construction of the prototype probes
probe deployment in water-filled hole
probe deployment in part water-filled hole
Preliminary results
Battery Voltage (Base) Solar panel seems to have provided more power than was used
Temperature & tilt of the base station
Summary & Conclusions Designed, built, tested and deployed 9 sensor probes, Base Station and Reference station Some things are difficult to calculate beforehand Lots of ad-hoc design and tuning is needed during installation
Future development Smaller probe electronics Lower frequency probe communications Probe positioning system Web site data plots
Credits Royan Ong, Joseph Stefanov, Al Riddoch, Harvey Rutt, ECS.Royan Ong, Joseph Stefanov, Al Riddoch, Harvey Rutt, ECS. Jane K. Hart, Sue Way, GeographyJane K. Hart, Sue Way, Geography Ian Marshall, BTExact;Ian Marshall, BTExact; Nathan Boyd and John Argirakis, InteliSYSNathan Boyd and John Argirakis, InteliSYS Funded by:Funded by: the Royal Society & DTI's Nextwave programme. the Royal Society & DTI's Nextwave programme.