vemco acoustic telemetry overview - imos

VemcoAcoustic Telemetry Overview

VemcoAcoustic Telemetry Overview

AgendaAgenda

� Single Frequency Telemetry – Rationale and How it works

� Equipment Overview and Representative Deployments

� Code Maps & Worldwide Unique IDs

� Data Rate and Collisions




� Data Rate and Collisions� Data Rate and Collisions

� Detection Performance & Range Limits

� False Positives

� Vemco User environment (VUE) Software

� Future Product Directions



� False Positives



More Detail on Topics Covered?More Detail on Topics Covered?

� At www.vemco.com/resources

� White Paper providing an overview of VR2 related products, performance to be expected and providing guidance to users

� FAQs – continually being expanded

� At www.vemco.com/resources

� White Paper providing an overview of VR2 related products, performance to be expected and providing guidance to users

� FAQs – continually being expanded� FAQs – continually being expanded

� Application Notes to come

� Equipment manuals

� Newsletter

� Contact Customer Support

� FAQs – continually being expanded

� Application Notes to come

� Equipment manuals

� Newsletter

� Contact Customer Support

AgendaAgenda









� False Positives





� False Positives



Communication RealitiesCommunication Realities

� Only two reliable methods of encoding information:

1. Presence or absence of a particular frequency

2. Time of pulse arrival

� Blanking requirement

� Only two reliable methods of encoding information:

1. Presence or absence of a particular frequency

2. Time of pulse arrival

� Blanking requirement

Our PhilosophyOur Philosophy� Create robust, scalable technology based on single

frequency telemetry with a view to

� Reliable data gathering and sharing

� Ease of deployment and operation

� Suitable for almost all environments

� Create robust, scalable technology based on single frequency telemetry with a view to

� Reliable data gathering and sharing

� Ease of deployment and operation

� Suitable for almost all environments

� Affordable

� Maintain compatibility across products and avoid unnecessary obsolescence

� Exploit Multifrequency/Spread Spectrum solutions as an enhancement to address specific requirements

� Affordable

� Maintain compatibility across products and avoid unnecessary obsolescence

� Exploit Multifrequency/Spread Spectrum solutions as an enhancement to address specific requirements

Single Frequency Coding MethodSingle Frequency Coding Method

� Unmodulated, fixed frequency pulses with information coded in spacing between pulses

“Delay” between successive

� Unmodulated, fixed frequency pulses with information coded in spacing between pulses

“Delay” between successive � “Delay” between successive Bursts to accommodate multiple transmitters

� Simple “Continuous” coding used only for Tracking

� “Delay” between successive Bursts to accommodate multiple transmitters

� Simple “Continuous” coding used only for Tracking

Why 69 kHz?Why 69 kHz?

Frequency (kHz) Tag Diameter (mm) Range Possibilities in Ocean (Metres)

<50 > 20 mm Several km

50 – 100 7 – 16 ~ 300 to 1000

>100 kHz < 7 mm < 300 metres

0

500

1000

1500

0 100 200 300

Frequency (kHz)

Ra

ng

e (

me

tre

s)

Typical Range Curves

142, 155 and 165 dB Transmitters• “Brick wall” at High

Frequency

• Choose Frequency as low as

possible consistent with size

requirement

Frequency Choice Fresh WaterFrequency Choice Fresh Water

Frequency Range

100 2827

150 1803

200 1409

300 785

400 450

Distilled Water Range (142 dB Transmitter)

� Range can be enormous in good conditions

� Still a penalty for going higher than necessary – Vemco using 180 kHz for new 6 mm family

� CAUTION long range can cause problems in high echo environments

� Range can be enormous in good conditions

� Still a penalty for going higher than necessary – Vemco using 180 kHz for new 6 mm family

� CAUTION long range can cause problems in high echo environments

Distilled Water Range (142 dB Transmitter)

Range Limitations – Overview Range Limitations – Overview �Factors under control of equipment designer:

– Signal Level and Energy, Frequency

– Receiver Bandwidth, Sensitivity, SNR to detect, etc.

�Factors under control of equipment designer:

– Signal Level and Energy, Frequency

– Receiver Bandwidth, Sensitivity, SNR to detect, etc.

�Factors not under designer’s control:

– Noise Levels: Sources include Wave action, rain, animals, boats, echo sounders

– Propagation conditions – absorption (Salt Water), temperature gradients, entrapped air bubbles, suspended material, biomass, etc.

�Factors not under designer’s control:

– Noise Levels: Sources include Wave action, rain, animals, boats, echo sounders

– Propagation conditions – absorption (Salt Water), temperature gradients, entrapped air bubbles, suspended material, biomass, etc.

Determining Appropriate Transmitter Power

Determining Appropriate Transmitter Power

� Can calculate range in Normal Ocean Conditions for as function of transmitter power and receiver characteristics

� BUT Range testing essential if conditions

� Can calculate range in Normal Ocean Conditions for as function of transmitter power and receiver characteristics

� BUT Range testing essential if conditions � BUT Range testing essential if conditions expected to differ from Normal Ocean Conditions

� Vemco provides an On Line Range Calculator to give users a starting point

� BUT Range testing essential if conditions expected to differ from Normal Ocean Conditions

� Vemco provides an On Line Range Calculator to give users a starting point

Fresh Water ConsiderationsFresh Water Considerations

� Absorption negligible so that very large range theoretically possible – over 6 km for 142 dB Tag!

� BUT other factors usually make range less – often much less – than in the ocean. These include:

� Man made noise (Dam Turbine, Sport fishing echo sounders,

� Absorption negligible so that very large range theoretically possible – over 6 km for 142 dB Tag!

� BUT other factors usually make range less – often much less – than in the ocean. These include:

� Man made noise (Dam Turbine, Sport fishing echo sounders, � Man made noise (Dam Turbine, Sport fishing echo sounders, etc.)

� Absorption due to turbulence and sediment

� Much more difficult (and often seasonally affected) to predict than in the ocean – in situ measurements are essential

� Man made noise (Dam Turbine, Sport fishing echo sounders, etc.)

� Absorption due to turbulence and sediment

� Much more difficult (and often seasonally affected) to predict than in the ocean – in situ measurements are essential

AgendaAgenda









� False Positives





� False Positives



“VEMCO 69 kHz Network” Overview“VEMCO 69 kHz Network” Overview

� Proven to work in virtually all environments

� Hundreds of thousands of unique ID codes with Error Checking

� Installed Base of Compatible Listening Stations

� Proven to work in virtually all environments

� Hundreds of thousands of unique ID codes with Error Checking

� Installed Base of Compatible Listening Stations � Installed Base of Compatible Listening Stations > 8,000 – all receivers detect all Vemco tags

� Creates huge potential for infrastructure and data sharing – e.g. POST, OTN, AATAMS, etc.

� Installed Base of Compatible Listening Stations > 8,000 – all receivers detect all Vemco tags

� Creates huge potential for infrastructure and data sharing – e.g. POST, OTN, AATAMS, etc.

“VEMCO 69 kHz Network” Transmitters“VEMCO 69 kHz Network” Transmitters

� Multiple trade offs of range, size and life (including multiyear)

� Sensor options – pressure and temperature today

� Onboard intelligence including Processing of Sensor Data, ON/OFF cycles, etc.

� Multiple trade offs of range, size and life (including multiyear)

� Sensor options – pressure and temperature today

� Onboard intelligence including Processing of Sensor Data, ON/OFF cycles, etc.

� Sufficient ID space to provide worldwide unique IDs

� Manufacturing enhancements for repeatability, quick order fulfillment, etc.

� Prices have held steady or been reduced despite a significant fall in the US dollar

� Sufficient ID space to provide worldwide unique IDs

� Manufacturing enhancements for repeatability, quick order fulfillment, etc.

� Prices have held steady or been reduced despite a significant fall in the US dollar

Representative TransmittersRepresentative Transmitters

V6 is first of 180 kHz Transmitter FamilyV6 is first of 180 kHz Transmitter Family

V9, V7 and V6 Coded Pingers V13PT and V9PT Sensor Tags

V16 High Power, Long Life Tags

Active and Passive TrackingActive and Passive Tracking

Active Tracking

� Receiver with Directional Hydrophone mounted on boat

Gather detailed

Active Tracking

� Receiver with Directional Hydrophone mounted on boat

Gather detailed

Passive Tracking

� Fixed receivers (Monitoring Stations) which log detections of tags present

Long term data on

Passive Tracking

� Fixed receivers (Monitoring Stations) which log detections of tags present

Long term data on � Gather detailed information on a single fish by following it

� Limited scope – time and area

� Gather detailed information on a single fish by following it

� Limited scope – time and area

� Long term data on many fish potentially over large areas

� Extended scope – time, species, geography

� 100% duty cycle

� Long term data on many fish potentially over large areas

� Extended scope – time, species, geography

� 100% duty cycle

Positioning SystemsPositioning Systems

� Grid of Monitoring Stations to combine features of Active and Passive Tracking

� Fish position detected accurately by triangulation

� Grid designed so signal detected by at least 3

� Grid of Monitoring Stations to combine features of Active and Passive Tracking

� Fish position detected accurately by triangulation

� Grid designed so signal detected by at least 3 � Grid designed so signal detected by at least 3 Receivers

� Accuracy driven by knowledge of receiver positions & time of arrival of transmitted signal

� Stationary – unlike Active Tracking

� Limited Geographic Extent – unlike Passive Tracking

� Grid designed so signal detected by at least 3 Receivers

� Accuracy driven by knowledge of receiver positions & time of arrival of transmitted signal

� Stationary – unlike Active Tracking

� Limited Geographic Extent – unlike Passive Tracking

VR100 Digital Tracking ReceiverVR100 Digital Tracking Receiver

� Current 100 kHz Configuration detects tags simultaneously on up to 8 frequencies

� Outperforms conventional receivers

� Current 100 kHz Configuration detects tags simultaneously on up to 8 frequencies

� Outperforms conventional receivers

�

receivers

� Decodes tags you can’t hear

� Simultaneous Detection of near and far transmitters without gain adjustment

�

receivers

� Decodes tags you can’t hear

� Simultaneous Detection of near and far transmitters without gain adjustment

Other Uses of VR100Other Uses of VR100

� Diagnostic Tool during study design

Can log all detections along with receiver location, signal strength and timing of individual pulses and echoes

� Basis for Data Acquisition system used to

� Diagnostic Tool during study design

Can log all detections along with receiver location, signal strength and timing of individual pulses and echoes

� Basis for Data Acquisition system used to � Basis for Data Acquisition system used to provide more detailed assessment of acoustic conditions during study design phase

� Basis for Data Acquisition system used to provide more detailed assessment of acoustic conditions during study design phase

VRAP Positioning SystemVRAP Positioning System

� Targets Residency Studies

� Real Time Precise Positioning

� Targets Residency Studies

� Real Time Precise PositioningPositioning

� Extent Limited by Transmitter Range

� Typically 1 to 2 metre accuracy

Positioning

� Extent Limited by Transmitter Range

� Typically 1 to 2 metre accuracy

VR2W Monitoring ReceiverVR2W Monitoring Receiver

Bruce et al, CSIRO Proceedings VR2

Workshop, Catalina

Island, 2005

• Life > 1 Yr• Logs 1,200,000 Detections• Bluetooth Download• $1,200

Remainder of Presentation will

focus on VR2-like Receivers

VR3 VariantsVR3 Variants

VR3s offer VR2 Functionality with Remote Communications

VR3s offer VR2 Functionality with Remote Communications

VR3 with Underwater ModemVR3 with Underwater Modem

KEY FEATURES

� Data Transfer Rate 1200 Baud

� Slant range up to 350 metres

� Internal tilt sensor allows monitoring during deployment

KEY FEATURES

� Data Transfer Rate 1200 Baud

� Slant range up to 350 metres

� Internal tilt sensor allows monitoring during deployment

Initial Argos Satellite Units Initial Argos Satellite Units

Dagorn & Holland, 2004

Seychelles, France & USA

Example Deployments of VR2s and VR3s

Example Deployments of VR2s and VR3s

Simple Technology enables wide range of deployments with opportunities for data sharing

Simple Technology enables wide range of deployments with opportunities for data sharing

Sacramento RiverSacramento River~ 200 VR2 array~ 200 VR2 array

http://californiafishtracking.ucdavis.edu/index.html

POST “Ocean Track” CurtainsPOST “Ocean Track” Curtains

Tagging CentersTagging Centers

Future CurtainsFuture Curtains

Existing CurtainsExisting Curtains

50 Species

5000 Tags

POST Receivers

Puget Sound

220 VR2 Receivers in Puget Sound

NWIFC

WDFW

NOAA

Puyallup Tr/Corps

UW/KCnty/Seattle/Corps

Seattle/Corps

NOAA/Seattle/Corps

AgendaAgenda










� False Positives





� False Positives



Tag Coding OverviewTag Coding Overview

� More than one scheme is used – parameters include length of first interval and number of intervals

� More than one scheme is used – parameters include length of first interval and number of intervalslength of first interval and number of intervals

� Each Coding scheme is referred to as a Code Space

� Receivers are configured with a Code Map which defines the Code Spaces they can detect – up to four for legacy VR2s

length of first interval and number of intervals

� Each Coding scheme is referred to as a Code Space

� Receivers are configured with a Code Map which defines the Code Spaces they can detect – up to four for legacy VR2s

Risk of Duplicate CodesRisk of Duplicate Codes

� Historically: Started with 256 Codes and then moved to 4k with geographic separation

� Popularity of technology was increasing the risk of iD overlap

� Any potential for confusion of ID of fish

� Historically: Started with 256 Codes and then moved to 4k with geographic separation

� Popularity of technology was increasing the risk of iD overlap

� Any potential for confusion of ID of fish � Any potential for confusion of ID of fish detected threatens integrity of all research results (except closed systems)

� By 2006 it became clear that status quo was not an option

� Any potential for confusion of ID of fish detected threatens integrity of all research results (except closed systems)

� By 2006 it became clear that status quo was not an option

Parameters of SolutionParameters of Solution

� Commitment to issue codes that are unique worldwide starting in 2007 Season.

� Maintain concept of 69 kHz worldwide tracking network – i.e. all receivers detect all tags worldwide

� Commitment to issue codes that are unique worldwide starting in 2007 Season.

� Maintain concept of 69 kHz worldwide tracking network – i.e. all receivers detect all tags worldwideworldwide

� Demand met by 68k IDs in 2007 with the ability to expand as necessary

� Introduced S64K to provide more unique IDs for Sensor tags

worldwide

� Demand met by 68k IDs in 2007 with the ability to expand as necessary

� Introduced S64K to provide more unique IDs for Sensor tags

MAP 110 – Standard Receiver Code Map MAP 110 – Standard Receiver Code Map

Code Space

Tag Type

SYNC Information Transmitted

A69-1008 R256 401.3 Legacy. Available for future use once VR1 use ends

A69-1206 R04K 380 Reserved for future use once existing tags in field expire

A69-1105 S256 360 8 Bit Sensor Data & 8 Bit ID Code

A69-1303 R64k 320 16 Bit ID Code some of which are assigned to Sensor Tags

� Will my Receiver Detect a particular Tag?

� Yes, as long as its Code Space is listed in Receiver’s Code Map

� VUE makes this simple

� Receivers configured to MAP110 will detect all Vemco Tags in the world

– Exception: A small number of tags used in legacy or closed situations with corresponding non standard Code Maps

� Will my Receiver Detect a particular Tag?

� Yes, as long as its Code Space is listed in Receiver’s Code Map

� VUE makes this simple

� Receivers configured to MAP110 will detect all Vemco Tags in the world

– Exception: A small number of tags used in legacy or closed situations with corresponding non standard Code Maps

Sensor Tags

S64k Sensor Codes – How it WorksS64k Sensor Codes – How it Works

� Tag alternately transmits S256 code as now and an R64k which uniquely identifies it

� Simple Migration situation:

� Simple matter to match sensor data and unique Tag

� Tag alternately transmits S256 code as now and an R64k which uniquely identifies it

� Simple Migration situation:

� Simple matter to match sensor data and unique Tag � Simple matter to match sensor data and unique Tag ID

� Price: Half data update rate for a given Delay

� If ratio of sensor tags to pingers is low, can simply halve Delay for sensor tags with little impact on collision statistics

� Simple matter to match sensor data and unique Tag ID

� Price: Half data update rate for a given Delay

� If ratio of sensor tags to pingers is low, can simply halve Delay for sensor tags with little impact on collision statistics

64k Sensor IDs – High Residency Situations64k Sensor IDs – High Residency Situations

� General Approach: For each S256 code detected searches for occurrence of a 64k ID which has been assigned to a sensor tag. This determines which fish the sensor data belongs to

� Works perfectly unless two tags simultaneously appear with S256 ID code. Note, however, that:

� General Approach: For each S256 code detected searches for occurrence of a 64k ID which has been assigned to a sensor tag. This determines which fish the sensor data belongs to

� Works perfectly unless two tags simultaneously appear with S256 ID code. Note, however, that:

�

with S256 ID code. Note, however, that:

� Such occurrences will be rare since Vemco will continue to provide as much geographic separation as possible between duplicate S256 Codes

� One still knows which fish were present. Sensor data will just be unavailable during time of simultaneous residence

�

with S256 ID code. Note, however, that:

� Such occurrences will be rare since Vemco will continue to provide as much geographic separation as possible between duplicate S256 Codes

� One still knows which fish were present. Sensor data will just be unavailable during time of simultaneous residence

AgendaAgenda










� False Positives





� False Positives



Sharing Bandwidth & CollisionsSharing Bandwidth & Collisions

Randomization of delay means that if two transmitters collide at a particular time, they will not on the next transmission – critical to the ability to deal with multiple transmitters

Randomization of delay means that if two transmitters collide at a particular time, they will not on the next transmission – critical to the ability to deal with multiple transmitters

Residency and Transmission DensityResidency and Transmission Density Pulse Burst Time

Transmission Density = Residency x ––––––––––––––––––––––

Pulse Burst Time + Delay

Transmission Density

Percent ”Clear” Transmissions

0.01 99%

0.1 76%

0.25 50%

Examples for Burst & Delay Burst

Times = 3 & 60 seconds

Residency Density

4 0.1

� “Clear” Transmissions provide a rough, estimate of percent that will be detected

� Choose Delay carefully – shorter is not better!

� “Clear” Transmissions provide a rough, estimate of percent that will be detected

� Choose Delay carefully – shorter is not better!

0.25 50%

0.5 23%

1 5%

1.5 1%

4 0.1

20 0.5

40 1.0

Detection Performance AnalysisDetection Performance Analysis

� Vemco provides an On Line Collision Calculatorto assist users

� Uses some some simplifying assumptions –validated with detailed simulations

� Vemco provides an On Line Collision Calculatorto assist users

� Uses some some simplifying assumptions –validated with detailed simulations

� Example results follow -- all results to 95% confidence level (i.e. stated performance or better 19 times out of 20)

� Example results follow -- all results to 95% confidence level (i.e. stated performance or better 19 times out of 20)

Residency and Detection TimeResidency and Detection Time

0

10

20

30

40

2 4 6 8 10 12 14 16 18 20

Tim

e t

o D

ete

ct

All

(M

inu

tes)

0

60

120

180

240

300

360

20 25 30 35

Tim

e to D

ete

ct A

ll

Limit can be moved out by increasing DelayLimit can be moved out by increasing Delay

2 4 6 8 10 12 14 16 18 20

Number of Transmitters Present Number of Transmitters Present

Average Delay = 60 sec. Expansion towards Limit

Effect of DelayEffect of DelayMaximum Residency Number for

all Tags to be Detected Delay

(Minutes) Minutes to

Detect 3 Resident Tags Within 1 Hour Within 3 Hours

.5 1.9 15 20 1 2.5 23 35

1.5 3 29 43

2 3.6 32 51

3 4.6 38 64 4 5.5 40 74

5 6.4 42 82

� TRADE OFF: Number of Tags that can be handled

Versus

Time detect a small number of Resident Tags.

� Vemco R64k transmitters can be very successfully applied to most situations as long as one pays attention to selection of Tag Delay

� TRADE OFF: Number of Tags that can be handled

Versus

Time detect a small number of Resident Tags.

� Vemco R64k transmitters can be very successfully applied to most situations as long as one pays attention to selection of Tag Delay

5 6.4 42 82

7.5 8.5 42 96 10 10.5 391 104

Average Time between Detections for Each TagAverage Time between Detections for Each Tag

10

20

30

40

50

60M

inu

tes

30 seconds

60 Seconds

90 Seconds

120 Seconds

240 Seconds

Delay

� Previous slides focused on how long to detect all tags

� Detection rate gives average update rate for sensor tags

� Previous slides focused on how long to detect all tags

� Detection rate gives average update rate for sensor tags

0

10

2 8 14 20 26 32 38 44 50

Number of Transmitters Present

AgendaAgenda










� False Positives





� False Positives



Role of Range TestingRole of Range Testing

� User range testing should focus on areas and conditions of planned deployment

� Critical that this be done if conditions are suspected to be different from those in open ocean

� User range testing should focus on areas and conditions of planned deployment

� Critical that this be done if conditions are suspected to be different from those in open ocean ocean

� Testing needs to be carefully planned in order to produce relevant results

ocean

� Testing needs to be carefully planned in order to produce relevant results

Ideal – but time consuming – Range TestingIdeal – but time consuming – Range Testing

� Objective is to find RMIN – the range at which virtually all transmissions will be detected under all anticipated conditions

� Choose a starting range and place Tag (one, fixed delay) in a fixed location and at a depth where fish are anticipated to swim

– Allow receiver to detect long enough that representative conditions

� Objective is to find RMIN – the range at which virtually all transmissions will be detected under all anticipated conditions

� Choose a starting range and place Tag (one, fixed delay) in a fixed location and at a depth where fish are anticipated to swim

– Allow receiver to detect long enough that representative conditions – Allow receiver to detect long enough that representative conditions are encountered – days or weeks and other representative transmitter depths and directions if this might be significant

– If detection rate 100% or very close, use this as RMIN or try further out

– If detection rate below 100%, move in closer.

� If results disappointing, keep in mind that receiver deployment location or depth might be problem

– Allow receiver to detect long enough that representative conditions are encountered – days or weeks and other representative transmitter depths and directions if this might be significant

– If detection rate 100% or very close, use this as RMIN or try further out

– If detection rate below 100%, move in closer.

� If results disappointing, keep in mind that receiver deployment location or depth might be problem

Short and Long Term Range TestsShort and Long Term Range Tests

0.00

0.20

0.40

0.60

0.80

1.00

1 3 5 7 9 1 1 1 1 1 2 2 2 2 2 3 3 3 3 3 4

Dete

ction P

robabilty

Range

RangeRMIN

Typical Short Term 0.20

0.40

0.60

0.80

1.00

Dete

ction P

robabilty

� Almost always need to derate range from short term test to get RMIN

� Long Term Detection Probability of 50% does not mean that 50% of the transmissions of a fish passing through will be detected

� Almost always need to derate range from short term test to get RMIN

� Long Term Detection Probability of 50% does not mean that 50% of the transmissions of a fish passing through will be detected

Typical Short Term

Result0.00

0.20D

ete

ction P

robabilty

RangeRMIN

0.40

0.60

0.80

1.00D

ete

cti

on

Pro

ba

bilit

y

20 km/hr

30 km/hr

40 km/hr

Mean Wind Speed

(σ = 0.5 x Mean)153 dB Transmitter

Impact of Noise VariabilityImpact of Noise Variability

0.00

0.20

0 200 400 600Range (Metres)

De

tec

tio

n P

rob

ab

ilit

y

40 km/hr

� Wind Noise is often the dominant influence – used here to illustrate

� Significant derating required if high noise levels anticipated

� Wind Noise is often the dominant influence – used here to illustrate

� Significant derating required if high noise levels anticipated

Typical Results from Shad Bay TestTypical Results from Shad Bay Test

0.6

0.7

0.8

0.9

1

Pro

ba

bilit

y o

f D

ete

cti

on

Weather less severe

0.7

0.8

0.9

1

Pro

babili

ty o

f D

ete

ction

� Transmitter: 142 dB re 1 µPa @ 1 metre

� ~ 1 Week of Data for each data point

� Working Range ~ 400 metres for this location and Tag type

� Transmitter: 142 dB re 1 µPa @ 1 metre

� ~ 1 Week of Data for each data point

� Working Range ~ 400 metres for this location and Tag type

0.4

0.5

225 275 325 375 425 475 550 600

Range (metres)

Pro

ba

bilit

y o

f D

ete

cti

on

0.5

0.6

Pro

babili

ty o

f D

ete

ction

275 Metre detail shows dips during

weather events

Gate and Array Design Considerations

Gate and Array Design Considerations

Brief Overview here – see White Paper for Details and MethodsBrief Overview here – see White Paper for Details and Methods

Site Fidelity ApplicationSite Fidelity Application

� Size of “site” varies with time depending on reception conditions (usually Noise)

� Size of “site” varies with time depending on reception conditions (usually Noise)

Maximum

Range

� Sentinel Tags can be useful in estimating size at any given time

� Sentinel Tags can be useful in estimating size at any given time

Working RangeWorking

Range

This one only seen in

lower noise conditions

Passage at a GatePassage at a Gate

� Time fish within range depends on range at that time, swimming speed and path

� Time necessary for Detection depends on Delay and Number of Fish Passing

� Time fish within range depends on range at that time, swimming speed and path

� Time necessary for Detection depends on Delay and Number of Fish Passing

"Typical"

� Overlap as shown necessary to ensure detection under all conditions

� High Detection Efficiency implies many fish detected by both

� Sentinel Tags help estimating performance at any given time

� Overlap as shown necessary to ensure detection under all conditions

� High Detection Efficiency implies many fish detected by both

� Sentinel Tags help estimating performance at any given time

Working RangeWorking Range

Typical

Range

Working

Range

PLWC

AgendaAgenda










� False Positives





� False Positives



False PositivesFalse Positives

How they occur and how to deal with them

How they occur and how to deal with them

Transmission Errors and False PositivesTransmission Errors and False Positives

� Regardless of communication scheme, transmission errors will occur

� Trade off between successful detections and false positives between two extremes:

� Regardless of communication scheme, transmission errors will occur

� Trade off between successful detections and false positives between two extremes:

� Aggressive Signaling: Large number of Transmission Errors

� Conservative Signaling: small number of Transmission Errors

� Data rate = Transmission rate – Error Rate

� Aggressive Signaling: Large number of Transmission Errors

� Conservative Signaling: small number of Transmission Errors

� Data rate = Transmission rate – Error Rate

Sync Max Int

Transmitted

Signal

Interference

Interference causes Failed DetectionInterference causes Failed Detection

Interference

What

Receive

sees

Sync SyncSync

Colliding Transmissions – Invalid SequenceColliding Transmissions – Invalid Sequence

Sync Max Int

Transmitter 1

Transmitter 2Transmitter 2

What

Receiver

Sees

Sync Sync Sync Sync

Colliding Transmissions – “Valid” SequenceColliding Transmissions – “Valid” SequencePurpose of Error Checking is to

Reject these – effective but not 100%

How Prevalent are False Positives?How Prevalent are False Positives?Average Number of Hours between False Positives Number of

Resident Tags Delay = 60 Delay = 120

5 12.6 20.5

10 7.1 8.9

20 1.5 1.1

30 1.4 0.6 30 1.4 0.6

Actual False Codes generated can come from a

very small set

Typical Results from Busy Receiver in FieldTypical Results from Busy Receiver in Field

Hourly Statistics

Avg Min Max Totals

Tags Present 9.5 6 13 24

Detections 161 18 202 3875

� Detection Rate and False Positive Rate Consistent with predictions for this and other days BUT

– One false Positive ID occurred 7 times over a 45 day period

– Another appeared twice in just a few hours

� Detection Rate and False Positive Rate Consistent with predictions for this and other days BUT

– One false Positive ID occurred 7 times over a 45 day period

– Another appeared twice in just a few hours

False Positives 1

Acceptance Tests and False PositivesAcceptance Tests and False Positives

� Valid Detections

� Multiple detections close together

� How close depends on Delay and number of Tags Present – i.e. larger intervals if lots of collisions

� Valid Detections

� Multiple detections close together

� How close depends on Delay and number of Tags Present – i.e. larger intervals if lots of collisions

� False Positives

� Single Detections or long intervals between repeats

� Acceptance Testing involves separation of Detections on the basis of above Characteristics

� False Positives

� Single Detections or long intervals between repeats

� Acceptance Testing involves separation of Detections on the basis of above Characteristics

Automating the ProcessAutomating the Process

� Used POST Data as test platform

� Now a prototype in house service

� Moving to application note and hopefully a customer tool

� Used POST Data as test platform

� Now a prototype in house service

� Moving to application note and hopefully a customer tool

� Philosophy: Conservative first scan accept test and more scrutiny on remainder

� Philosophy: Conservative first scan accept test and more scrutiny on remainder

Rigor of Acceptance & False PositivesRigor of Acceptance & False Positives

Total Number of Tag Receiver pairs 2661

More than one Detection 2291

A: At least two intervals < 4 minutes 1995

(5)

B: At least two intervals < 30 minutes of

which one is less than 4 minutes

2069

(8)

C: At least one interval < 30 minutes 2259

(14) C: At least one interval < 30 minutes

(14)

D: At least one interval < 12 hours 2275

(24)

First Scan Test

1. At least one interval < 30 minutes

AND

2. More intervals < 30 minutes than > 12 Hours

First Scan Test

1. At least one interval < 30 minutes

AND

2. More intervals < 30 minutes than > 12 Hours

Summary of Results for one SeasonSummary of Results for one Season

Detections 326,358

Accepted on First Scan 325,765

Not Accepted on First Scan 593

Finally Rejected 232 (0.07%)

Tag/Receiver Combinations 2661Tag/Receiver Combinations 2661

Accepted on First Scan 2245

Accepted on further analysis 243

Finally Rejected 173

with more than one Detection 18

Careful Examination of First Scan Accept showed no signs

of any False Positives

Can we Accept Single Detections?Can we Accept Single Detections?

� Not safe if any significant number of collisions occurring at receiver

� Reasonably safe if detailed examination shows no other tags at receiver at time of detection

� Not safe if any significant number of collisions occurring at receiver

� Reasonably safe if detailed examination shows no other tags at receiver at time of detection

� Plan to incorporate into tools ability to determine safeness of accepting single detections

� Planned new coding schemes will virtually eliminate false positives

� Plan to incorporate into tools ability to determine safeness of accepting single detections

� Planned new coding schemes will virtually eliminate false positives

AgendaAgenda










� False Positives





� False Positives



Vemco User Environment (VUE) Software

Vemco User Environment (VUE) Software

VUE is intended to give users a uniform easy to use interface as well as provide support for Application

Databases

VUE is intended to give users a uniform easy to use interface as well as provide support for Application

Databases

Legacy PC SoftwareLegacy PC Software

Problem: Basic unit is a file for each upload.Problem: Basic unit is a file for each upload.

Problems Introduced by File Orientation




1. Handling the Header

2. Duplicate Detections

1. Handling the Header

2. Duplicate Detections

Problem Introduced by HeaderProblem Introduced by Header

� Gets in the way of a smooth import __ Access or Excel

� Need to separate Header from Detections BUT

� Gets in the way of a smooth import __ Access or Excel

� Need to separate Header from Detections BUT from Detections BUT maintain the connection – otherwise Detection is ambiguous

from Detections BUT maintain the connection – otherwise Detection is ambiguous

Problem with Duplicate DetectionsProblem with Duplicate Detections

� Unless Memory is erased after each upload, importing of successive downloads creates duplicate connections

� BUT erasing memory each time is Risky!

� Unless Memory is erased after each upload, importing of successive downloads creates duplicate connections

� BUT erasing memory each time is Risky!

� The more receivers and uploads involved, the messier this gets.

� The more receivers and uploads involved, the messier this gets.

Legacy SW – Dealing with File OrientationLegacy SW – Dealing with File Orientation

� In small studies, one can deal with these manually

� Some Researchers have automated this

� VUE “Platform” Philosophy

� In small studies, one can deal with these manually

� Some Researchers have automated this

� VUE “Platform” Philosophy� VUE “Platform” Philosophy

� Deal with issues such as this which are common to all users allowing them to concentrate on unique aspects of their application

� VUE becomes “Front End” for single and multi user databases, registries, etc

� VUE “Platform” Philosophy

� Deal with issues such as this which are common to all users allowing them to concentrate on unique aspects of their application

� VUE becomes “Front End” for single and multi user databases, registries, etc

VUE Creates a Database of Detection Records VUE Creates a Database of Detection Records

Detection Record uniquely describes each Transmitter detection by a receiverDetection Record uniquely describes each Transmitter detection by a receiver

Detection Definers Additional Information

Date/Time (of the detection Sensor Readings and Units

Code Space & ID Transmitter Serial NumberCode Space & ID Transmitter Serial Number

Receiver Serial Number Receiver Station Name

Receiver Lat and Long

Signal Conditions (Checksum Errors, Signal density, etc.)

Etc.

VUE StructureVUE Structure

VR2Vemco User

Vemco

Database

(VDB)

VR2W Plug InVR2W

Future Receiver Support

� All future Monitoring Receivers

� VR100

Future Receiver Support

� All future Monitoring Receivers

� VR100

VR2 File

Plug In

VR2

File

Vemco User

Environment

(VUE)

Vemco

Receiver

Logs

(VRL)

Export

Data

VR2 Plug InVR2

VUE FeaturesVUE Features

� Configure and upload from multiple receivers of varying kinds

� Assigns station or location information to receivers and or detections

� Removes duplicate detection records

� Instantly sorts detections by receiver, station or

� Configure and upload from multiple receivers of varying kinds

� Assigns station or location information to receivers and or detections

� Removes duplicate detection records

� Instantly sorts detections by receiver, station or � Instantly sorts detections by receiver, station or transmitter

� Handles duplicate tag IDs

� Imports VR2 legacy files

� Correction for Time drift and incorrect clock settings

� Graphing of detections across receiver or transmitters

� Instantly sorts detections by receiver, station or transmitter

� Handles duplicate tag IDs

� Imports VR2 legacy files

� Correction for Time drift and incorrect clock settings

� Graphing of detections across receiver or transmitters

Detection Database (.VDB File)Detection Database (.VDB File)

� Created by importing .VRL files of desired Uploads

� Can contain millions of detections from multiple receivers –limited only by computing resources

� Can create any number of independent database files

� (Default) Receiver uploads automatically imported to currently

� Created by importing .VRL files of desired Uploads

� Can contain millions of detections from multiple receivers –limited only by computing resources

� Can create any number of independent database files

� (Default) Receiver uploads automatically imported to currently � (Default) Receiver uploads automatically imported to currently open database

� Instantly sort detections by receiver, transmitter or station

� Duplicate Detections – due to fully or partially duplicate uploads – automatically discarded. This eliminates a big issue that exists with legacy file-based approach

� (Default) Receiver uploads automatically imported to currently open database

� Instantly sort detections by receiver, transmitter or station

� Duplicate Detections – due to fully or partially duplicate uploads – automatically discarded. This eliminates a big issue that exists with legacy file-based approach

Export of Detection DatabaseExport of Detection DatabaseDate/Time,Code Space,ID,Sensor 1,Units 1,Sensor 2,Units

2,Transmitter Name,Transmitter S/N,Receiver Name,Receiver

S/N,Station Name,Station Latitude,Station Longitude

2006-10-05 23:40:30.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-05 23:43:34.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-05 23:45:28.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-05 23:47:40.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-05 23:51:06.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-05 23:54:29.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-05 23:57:41.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-06 00:00:02.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

Date/Time,Code Space,ID,Sensor 1,Units 1,Sensor 2,Units

2,Transmitter Name,Transmitter S/N,Receiver Name,Receiver

S/N,Station Name,Station Latitude,Station Longitude

2006-10-05 23:40:30.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-05 23:43:34.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-05 23:45:28.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-05 23:47:40.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-05 23:51:06.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-05 23:54:29.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-05 23:57:41.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-06 00:00:02.000,A69-1105,1,4.4,meter,,,,,,5673,,,, 2006-10-06 00:00:02.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-06 00:04:08.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-17 01:31:57.000,A69-1105,22,38.1,meters,,,,,,5673,,,,

2006-10-17 01:32:26.000,A69-1105,49,0,meters,,,,,,5679,,,,

2006-10-17 01:32:38.000,A69-1105,49,0,meters,,,,,,5679,,,,

2006-10-17 01:32:51.000,A69-1105,49,0,meters,,,,,,5679,,,,

2006-10-17 01:33:04.000,A69-1105,49,0,meters,,,,,,5679,,,, 2006-10-17 01:33:16.000,A69-1105,49,0,meters,,,,,,5679,,,,

2006-10-06 00:00:02.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-06 00:04:08.000,A69-1105,1,4.4,meter,,,,,,5673,,,,

2006-10-17 01:31:57.000,A69-1105,22,38.1,meters,,,,,,5673,,,,

2006-10-17 01:32:26.000,A69-1105,49,0,meters,,,,,,5679,,,,

2006-10-17 01:32:38.000,A69-1105,49,0,meters,,,,,,5679,,,,

2006-10-17 01:32:51.000,A69-1105,49,0,meters,,,,,,5679,,,,

2006-10-17 01:33:04.000,A69-1105,49,0,meters,,,,,,5679,,,, 2006-10-17 01:33:16.000,A69-1105,49,0,meters,,,,,,5679,,,,

Suitable for import to user Application

Database or spreadsheet

Putting it all TogetherPutting it all Together

Host and Archive

Project Database

VRL Files

Field Laptops

Detection

Database (.vdb) Export

RX Uploads

Possible Export to

Multiuser

Tag Registry, Oceanographic

Data, etc. (if relevant)

VRL File

ArchiveVRL Files

Other Study Data (Species,

Surgery Technique, etc, etc.)

RX

Uploads

Multiuser

Database

Demonstration with POST DataDemonstration with POST Data

OBJECTIVE

Demonstrate VUE as easy-to-use front end for Application Databases – including

for Legacy Data

OBJECTIVE

Demonstrate VUE as easy-to-use front end for Application Databases – including

for Legacy Data

Data QA IssuesData QA Issues

� Extensive Array with multiple users as well as many detections of “Non POST” transmitters

� Create reports on “Non POST” Tags as first step to identifying these fish – recruiting tool for POST

� Extensive Array with multiple users as well as many detections of “Non POST” transmitters

� Create reports on “Non POST” Tags as first step to identifying these fish – recruiting tool for POSTPOST

� Essential to have the ability to quickly identify potential false positives

POST

� Essential to have the ability to quickly identify potential false positives

Import of Detection DataImport of Detection Data

� Imported all unprocessed receiver logs (~800) since inception (2004) consisting of producing:

– 280 receivers

– 3500 Transmitters

– 3.5 million Detections

� Imported all unprocessed receiver logs (~800) since inception (2004) consisting of producing:

– 280 receivers

– 3500 Transmitters

– 3.5 million Detections

� Single VUE Data base immediately eliminated all duplication of detections and uploads – a task which required a large effort from POST/Kintama

� Import took approximately 1 hour

� Single VUE Data base immediately eliminated all duplication of detections and uploads – a task which required a large effort from POST/Kintama

� Import took approximately 1 hour

Connecting other Information to DetectionsConnecting other Information to Detections

� Tag Data – Ownership, Release, Date, Life, etc.

� Assignment of Receivers to Line, Receiver Location, etc.

� Spreadsheet data on tag serial numbers,

� Tag Data – Ownership, Release, Date, Life, etc.

� Assignment of Receivers to Line, Receiver Location, etc.

� Spreadsheet data on tag serial numbers, � Spreadsheet data on tag serial numbers, calibration, receiver location, etc. – especially useful when database contains detections from VR2 log files

� Typical queries follow

� Spreadsheet data on tag serial numbers, calibration, receiver location, etc. – especially useful when database contains detections from VR2 log files

� Typical queries follow

Time of Arrival at LinesTime of Arrival at Lines

POST and Foreign Tags – Tag Info SuppressedPOST and Foreign Tags – Tag Info Suppressed

vemco acoustic telemetry overview - imos

Documents