Wildlife telemetry and real-time data analysis A summary of the - - PowerPoint PPT Presentation

▶
wildlife telemetry and real time data analysis
SMART_READER_LITE
LIVE PREVIEW

Wildlife telemetry and real-time data analysis A summary of the - - PowerPoint PPT Presentation

Sep 07, 2022 245 likes •462 views

Wildlife telemetry and real-time data analysis A summary of the current state of consumer available technologies and recent innovations John Grant, Sigma Eight Inc. Dec 3, 2018 Conclude Observe Results are compared with A certain behaviour

slide-1
SLIDE 1

Wildlife telemetry and real-time data analysis

A summary of the current state of consumer available technologies and recent innovations John Grant, Sigma Eight Inc. Dec 3, 2018

slide-2
SLIDE 2

How do we collect the data?

Conclude Results are compared with the hypothesis Test The hypothesis is tested in a controlled environment Hypothesize A research asks ‘why?’, and proposes an explanation Observe A certain behaviour is seen in nature or in a subject

Scientific Method

slide-3
SLIDE 3

Acoustic Cost: $$ + Salt/fresh water + < 1g sizes + 3D locations + No antenna

  • Diving often involved
  • Hard to automate
  • Poor in turbulent

water

  • Cannot track by

land/air

  • Aquatic animals only

Acoustic Cost: $$ + Salt/fresh water + < 1g sizes + 3D locations + No antenna

  • Diving often involved
  • Hard to automate
  • Poor in turbulent

water

  • Cannot track by

land/air

  • Aquatic animals only

Current Tracking Technologies

GPS Cost: $$$ + High accuracy location + Precise timestamps

  • Usually combined

with a transmitter for retrieval

  • High power

consumption

  • Not available in < 1g

GPS Cost: $$$ + High accuracy location + Precise timestamps

  • Usually combined

with a transmitter for retrieval

  • High power

consumption

  • Not available in < 1g

VHF (Radio) Cost: $ + Freshwater/land + < 1g sizes + Low resolution locations + Easy-moderate to setup + Easy-moderate to automate

  • Antenna
  • Poor in deep or conductive

water

  • No high resolution

locations

slide-4
SLIDE 4

VHF Transmitter

  • Consists of 3 basic parts
  • Emits a signal on some time

interval

  • Emits on a specific

frequency

Battery PCB Antenna

slide-5
SLIDE 5
slide-6
SLIDE 6

Data Characteristics

Timestamp Receiver ID Antenna Frequency Type Code Power (dBm) 2018-12-02 10:12:27 4 1 150.560 BEEPER 1002

  • 94

... ... ... ... ... ... ... E x a m p l e R

  • w
slide-7
SLIDE 7

How do we process the data?

Conclude Results are compared with the hypothesis Test The hypothesis is tested in a controlled environment Hypothesize A research asks ‘why?’, and proposes an explanation Observe A certain behaviour is seen in nature or in a subject

Scientific Method

slide-8
SLIDE 8

Current Techniques

  • Data is often gathered manually
  • Sites visited occasionally → potential for

long dropouts

  • Sites visited often → can be expensive
  • Data wrangle in spreadsheets,

databases, or with R scripts

  • Scrutinize the data

? ? ? ?

slide-9
SLIDE 9

Solution: The internet!

  • Put receiver online
  • Respond to dropouts as

required, instead of a set timeline

  • Centralize the database for

all receivers

Internet

slide-10
SLIDE 10

Conclude Results are compared with the hypothesis Test The hypothesis is tested in a controlled environment Hypothesize A research asks ‘why?’, and proposes an explanation Observe A certain behaviour is seen in nature or in a subject

Scientific Method

Automate

slide-11
SLIDE 11

Can MITAS efficiently transmit receiver data

  • ver the internet?
slide-12
SLIDE 12

MITAS Experiment

Procedure:

  • 3 receivers were setup in our office in

Aurora, Ontario, Canada

  • 2 transmitters was placed nearby,

emitting every 1.1s at a moderate signal strength

  • A database server was setup in

Sydney Australia

  • Experiment was run for 1 week

Data

slide-13
SLIDE 13

Each transmitters average contact power grouped by receiver over the 1 week experiment.

Results

slide-14
SLIDE 14

99.81 ± 0.04%

Transmission efficiency

slide-15
SLIDE 15

Results Explained

Factors for data loss:

  • Random network outages
  • Power outages
  • Strong bursts of noise

Remedies:

  • Have wired connections when

possible

  • Ensure there is a backup source of

power, and/or voltage monitors

  • Pick frequencies that are not busy
slide-16
SLIDE 16

Demo

slide-17
SLIDE 17

Next Steps

  • Ability to add tags

○ Sort data by weight, age, gender, size… ○ Emails/notifications when a tag is present at receiver ○ Migration paths

  • Daily reports

○ Detection rates of beacons ○ Status reports on receivers

  • Share data

○ Add other researchers to your project with certain permissions ○ Publicize data

slide-18
SLIDE 18

All the data generated in this experiment is

  • pen source

https://github.com/SEI-John/MitasEfficiencyExperiment

slide-19
SLIDE 19

Questions?

slide-20
SLIDE 20

Data Wrangling

Grolemund, G., & Wickham, H. (n.d.). Retrieved November 27, 2018, from https://r4ds.had.co.nz/wrangle-intro.html