The Plan

BushfireNet (BushfireNet.com) is our end to end solution to the challenge set by the Wildfire XPrize (track B).  Our team started out in Australia and hence we talk about bushfires rather than wildfires.

Don't reinvent the wheel!

We don’t have time to reinvent the wheel if we are to have a solution ready for next fire season.  We're harnessing the hard work already done so far in the field and integrating it under the umbrella of BushfireNet.com

In order to compartmentalise the system we’ve broken the problem into 

  1. Detect: 
  2. Decide? 
  3. Do!
  4. Communicate and 
  5. Operate. 

The first three of these domains give us the iterative D cubed decision loop.

  1. Detect: - Bushfire Long Range Early Warning System (BLREWS) [ blrews.com ]
  • Tethered lighter than air (LTA) aerostats provide an “eye in the sky” with an extended range for detection based on the work done by Skysentry powered by Helikites
  • Robot kites provide a high wind alternative to the aerostats to extend coverage
  • Powered tethered drones provide another path to elevated observation.
  • Mobile fixed wing autonomous drones on patrol over a 100km2 area enable a closer view.
  • Patented fire detection sensor work provides a solution for both air and ground detection.
  • Traditional ground based deployable fire watch tower infrastructure (long range cameras - visual and thermal) gives an all weather capability.
  1. Decide? Bushfire AI
  • At the heart of BushfireNet decision making is the Bushfire AI backend system.
  • The various AI components of the system fall under this team
  • A key part of our AI development strategy is to train our BushfireNet AI on a real life bushfire visual data set that we will have the ability to collect on an ongoing basis with the creation of controlled test fires (in off season) around the Good Good Run test facility.
  1. Do!

Depending on the distance of the fire from the nodes, the following system components respond:

  • Bucket brigade (BB) [ bucketbrigade.au ]
    • A modernised “bucket brigade” approach to the problem of fast suppression that makes use of multiple smaller drones deployed throughout the region
    • Working with existing commercial drones and custom builds.
    • Using the Elide Fire Ball as a payload ( https://elidefire.com.au/ )
    • The BB drones will also serve as spotter/scouts to aid in validation of the nature of a fire.
  • Mobile Bushfire Artillery (MBA) [ bushfireartillery.com ]
    • Fight fire with fire! Artillery fire..
    • A mobile, but pre-positioned, ground launcher that “fires” unpowered retardant payloads (Elide fire ball) at a target location. Under the control of BushfireNet.com (and thus human monitoring).
    • Uses the eye in the sky to confirm range free of humans and assets along ballistic lines of fire.
    • Air powered launch avoids the fire risk of other alternatives 
  1. Communicate

The system depends on interconnection between components and makes use of the following communications infrastructure:

  • Mobile Phone network integration (Telstra)
  • Starlink (Portable)
  • WiFi
  • GoTenna (Danger Solutions)
  1. Operate
  • BushfireNet.com 
    • Our “Skynet” for wildfires
    • Provides the human on the loop interface.
    • Our team has extensive experience in complex commercial and government enterprise IT systems development and deployment.
  • At the heart of our mobility approach is trailer based deployment.
  • Compliance! Health and Safety management 
  • Regulatory compliance (local government planning approvals, aviation, firefighting, native title, national parks etc) 
  • USA vs Australia challenges.

Good Good Run Test Facility 

One of our key unique capabilities is our exclusive access to a private 13 km2 test facility (1/100th scale) in regional NSW Australia with an environment similar to that required by the competition guidelines where we have permission to light fires, operate facilities and drones, and deploy equipment on a property wide level. 

This allows us to conduct ongoing systems development, testing and iteration and collect the necessary training data for our AI model development.