InNetwork Adaptation of Sensor Node Location and Energy William J' Kaiser UCLA Electrical Engineerin

1
Center for Embedded Networked Sensing
MWH Meeting 02.09.07
Jeff Goldman, Tom Harmon, Bill Kaiser
2
what is embedded networked sensing?
embedded in the physical environment (soil,
canopy, groundwater)
networked sharing information (data, system
status) via wired or wireless connections.
sensing measurement instruments (sensors,
transducers)
an internet of sensors
3
what do we do at CENS? create programmable autonom
ous, distributed, multi-function, multi-user,
observatories to address compelling science and
engineering issues
4
why are we here?

• Technology past a maturation threshold
• Developing strategic alliances in environmental
  science, engineering, and stewardship
• Require tight coupling between technology and
  application
• Shared interest areas and competencies
• Mutual benefits

5
mutual benefits
CENS expand to more, new proving grounds research
support student internships new research
opportunities
MWH, its clients partners create competitive
advantage efficiencies augment RD
capacity access to graduates leverage state
support tax benefits IP rights
6
whats in store today?
Bill Kaiser High-level look at sensing
technology NIMS technology Rapid deployment
version Aquatic version Tom Harmon Application
of NIMS technology Local example San Joaquin
River Other potential problems that can be
addressed
7
Embedded Networked Sensing Systems
New York Times NEON
Wireless Multihop Seismic Array
Terrestrial Ecosystem Sensing
Subsurface Contaminant Sensor Arrays
8
Contaminant Flux Measurement

• Directly measure contaminant flux in large scale
  river
• Identify sources and sinks
• Point source and non-point source contributions
• Measurement
• Map 3D flow field in river cross-section
• Map contaminant concentration
• Integrate product to compute flow
• Requires high spatial resolution
• Non-uniform contaminant distribution
• Complex flow and mixing

Confluence of Merced and San Joaquin Rivers
9
System Requirements

• Sensor Payload
• Support standard water quality sensor packages
• Support angle-resolved flow measurements
• Sensor Location Control
• Control sensor position in river span and depth
• Sampling
• Must also include spatially-resolved sample
  collection
• Autonomous
• Must sample continuously and with limited
  operator burden
• Rapidly deployable

10
NIMS Systems

• Introduce cable infrastructure to enable
• Large sensor payload
• Precise location control
• Autonomous operation
• Networked Infomechanical Systems (NIMS)
• Developed in 2003
• Deployed in multiple environments
• Fixed Based Version
• Rapidly Deployable Version
• Vendor
• Kaifuse Inc.
• Network Access
• Local standard WiFi
• Remote access integrated with SensorBase database
  system

11
NIMS Components
suspension cable
shuttle
Control Module
horizontal drive
vertical drive
Control Module Wireless Embedded Computer
12
NIMS Sensor Payloads

• Water Quality (standard sensor options)
• Conductivity
• Temperature
• pH
• Turbidity
• Ammonia
• Nitrate (ISUS Spectroscopic)
• Dissolved Oxygen
• Hydrologic
• 3-Axis velocity
• pitch/roll/yaw corrected
• Sensor payload depth
• Sounding
• Fluorescence
• Chlorophyll
• Physical Sampling
• Standard syringe canister sampling
• Under Development
• Tracer Systems

13
NIMS Deployments

• Ecosystems
• Mt. San Jacinto Reserve
• White Mountains
• La Selva, Costa Rica
• Urban Streams
• Medea Creek, Los Angeles County
• Lake Systems
• Lake Fulmor, Mt. San Jacinto Reserve
• River Systems
• San Joaquin

14
NIMS Operations at San Joaquin River
15
NIMS Operations at San Joaquin River
16
NIMS-AQ Systems

• Selected for wide span operations
• Buoyant payload support
• Standard NIMS Platform
• Embedded computing
• Wireless access
• Rapidly deployable
• Light weight, small team
• Horizontal and vertical actuation
• Standard sensing and sampling payloads

17
Applications perspective

• Current practices
• Time series at fixed stations
• Intensive synoptic sampling campaigns
• Model parameterization often limited to 1D by
  available data

18
NIMS RD on the San Joaquin River

• Automated mapping of coupled flow velocity
  water quality
• Resolving fluxes and gradients
• Quantitative mass balances
• Actuated/autonomous sample collection

19
Example Application San Joaquin-Merced
Confluence Tests (Oct 2005, Aug 2006)

• Accuracy of velocity field
• Gradient mapping
• Salt load assessment
• Mass balance over river reach
• Physical sampling

20
Pre-deployment characterization

• Kayak-mounted system (Valeport Echosounder)
• DGPS coordinate with depth
• Real-time plotting software

Oct 2005 NIMS transect
21
San Joaquin-Merced River Confluence

• 55 m span NIMS RD
• Sontek ADV
• Hydrolab sonde
• 1 low res scan
• 2 high res scans

22
System performance

• October 2005
• 50m span, 6000 samples of 8 variables in 100
  minutes
• (comparable experimental effort 15m stream, 350
  sample points, 1 variable (velocity), 2 weeks)
• August 2006
• Successful testing on adaptive sampling
  algorithms
• Successful testing of user-actuated physical
  sampling
• 2007?
• Looking for bigger challenges in creating an
  end-to-end technology

23
Velocity field calibration results
Harmon et al. Environ. Eng. Science, in press,
24(2), 2007 (March issue)
10x vertical exaggeration in plots
24
Coupled velocity-conductance readings (integrated
to yield a total salt load)
San Joaquin side
Merced side
Harmon et al. Environ. Eng. Science, in press,
24(2), 2007 (March issue)
25
Other data are undergoing analysis
San Joaquin
Merced

• Velocity
• Conductance
• Nitrate
• pH
• temperature

Setup time 2 h Sampling time 1-2
h Tear down time 2 h (pre-NIMS AQ, which is
faster)
26
Precision mass balances flow and chemical mass
Mass in Mass out groundwater loss (gain)
27
Multi-scale mixing and re-aeration
Dual-Scale Mixing/tracer characterization
Continuous injection of rhodamine dye solution
de-oxygenated water
NIMS AQ transects
28
Other applications

• Actuated or event-triggered physical sampling
• Example (1) groundwater seepage suspected, (2)
  allow temperature or nitrate sensors to trigger
  sampling events, (3) send sample for stable
  isotope analysis

29
California applications alone

• Tighten the salt balances (San Joaquin main stem,
  Calsim II salinity-flow ratings curves)
• Other contaminants (e.g., nutrients)
• Quantify groundwater loss/gain along river
  systems (e.g. San Joaquin)
• Detailed characterization of hypoxia and
  experiments to alleviate it (Stockton ship
  channel)
• Detailed assessment of environmental flow,
  riparian restoration, etc. (SJR releases,
  tributary salmon habitat restoration efforts,
  Klamath releases)

30
what now?

• Provided a look at technology and applications
• Understand the needs of MWH
• Identify common ground related to NIMS and other
  sensing methods
• Pursue joint ventures
• Leverage state funding via Industry-University
  Cooperative Research Program, etc.