PROBLEM STATEMENT:
Developing solution to locate Red Imported Fire Ants using multispectral airborne sensors and location based Artificial Intelegence

SOLUTION:
Red Imported Fire Ants (RIFA) are one of the most invasive species to reach
Australian shores posing a significant risk to the natural ecology since they were
first detected in 2001. RIFA have the potential to cause more damage to the natural
ecology than other known pests such as foxes, wild dogs and cats, camels, rabbits,
and cane toads combined. RIFA has cost billions of dollars to other economies,
hence the need to act fast in Australia. On-ground human and canine teams are used
to locate and eradicate RIFA nests. However, these teams are resource-intensive
and difficult to scale across large areas of land. Therefore, Outline sought to
investigate and develop a remote sensing solution using aerial imagery and
AI-generated insights to complement existing in-field surveillance efforts.
Outline, in conjunction with Biosecurity Queensland, launched a pilot program using
aerial remote sensing techniques to identify RIFA nest locations. A field investigation
spanning thousands of hectares was conducted, identifying a unique heat marker
that could be used to establish the location of the RIFA nests. Traditional thermal
cameras have been tested before on this heat marker. However, these systems
generated too many false positives due to the proximity of other infield features
generating similar signals. To overcome this, Outline developed a custom GTechTM
7-band camera system that captured complimentary data inputs to enable an AI
model to differentiate between noise and the true heat marker from the RIFA nests.
These AI predictions were verified by on-the-ground teams with feedback fed back
into the model for training purposes. This combined effort created a strong feedback
loop to continually improve the model, delivering commercially acceptable results
in less than two years.
A world-first ultra-high resolution multi-spectral and thermal aerial surveillance
camera system was developed, in addition to an advanced geospatial AI model.
Outline was successfully able to detect the locations of RIFA nests with confidence,
over a significant geographic area of 13,000ha, scaling up to 50,000ha in 2021.
Following this project, Outline won the Asia Pacific Award for Spatial Excellence 2020,
which acknowledges innovation and excellence in the Surveying and Spatial industry.
DURATION:
2018 to date
HSEQ STATS:
130 Aerial survey hours; 1,200 person hours worked
PROJECT STAFF:
2 geospatial AI Specialists
1 hyperspectral analyst
2 data science professionals
3 photogrammetrists
1 operations manager
1 project lead
KEY ACHIEVEMENTS
Interested in working with us
Need straightforward advice on the best image based spatial solution for your business, talk to us!