Every year, natural disasters destroy homes, ravage our environment, and take lives. Australia is unfortunately exposed to almost every kind of natural disaster: drought, floods, bushfire, cyclones, earthquakes – you name
it, and we’ve probably experienced it. As our technology advances, more attention turns to the solutions we can implement in order to improve emergency responses.
MSP was excited to participate in SmartSat CRC’s first ever Ideation Challenge – Firefly.
The challenge: Rapidly conceive a payload for natural disaster preparation, response or recovery and demonstrate on a stratospheric balloon.
Commencing in 2020, the sprint-style project focussed on the development of a quick initial solution to a complex problem. Over a two-month period between August and September 2020, MSP’s Firefly team worked quickly to bring together a solution that addressed both the SmartSat CRC challenge, and a challenge being faced in the area of disaster response.
Bridging the gap between research and reality
Keeping channels of communication open during a natural disaster is imperative, however bushfires often result in power being cut and telecommunications failing. This can leave residents of high-risk areas either reliant on analogue radio to hear any important announcements, or even in a communication blackout.
Whilst researching what path to take in tackling the SmartSat CRC challenge, the team homed in on how bushfire smoke can affect telecommunications. Whilst they found a lot of research relating to smoke attenuation, they didn’t see this translate into many real-world applications. We aim to bridge that gap.
Our solution: To provide a real-time service that predicts the extent to which communication will be impeded in areas impacted by natural disaster.
MSP has been developing a machine learning model that predicts communication blackout areas caused by bushfire smoke. Through the use of real-time multispectral imaging combined with the knowledge of cell tower locations, our market survey suggests that our proposed service is a novel solution that facilitates an active approach to finding and responding to blackout areas.
How our members have applied themselves
As always, developing our members is at the heart of everything MSP does. Through the Firefly project, our members have had the experience of working a unique challenge with potential real-world implications. A range of skills had to be quickly learnt, or enhanced, to fulfil the challenge requirements in such a short period of time. This includes:
· Machine Learning with Python, including image classification, image segmentation
· Cloud platforms such as AWS and GCP
· Hardware such as the Nvidia Jetson Nano
· Creating datasets from satellite imagery
· ArcGIS and QGIS
Aside from the technical growth, the members also had to be very strict in time management whilst working cohesively to ensure a standout product was created in the constrained time.
Our team is excited to take this project to the next level in 2021 - watch this space for some exciting updates about our Firefly project coming soon!
Did this article interest you? Check out our projects page to see what else MSP has in the pipeline.