The ‘Black Summer’ fires of 2019-2020 killed 33 people, and killed or displaced an estimated 3 billion animals; destroyed 3000 homes and 16 million hectares of land, representing $200 billion in damage - almost 10% of Australia’s GDP. The scale and complexity of the bushfires and weather systems during the Black Summer combined with the complex web of stakeholders, and data paralysis induced by 250+ siloed systems firefighters have to consult pushed people, processes, and existing technology beyond their limits. These challenges are only expected to get worse as climate change further impacts our environment and communities.
A future where we can detect new ignitions via social media, predict a bushfire’s spread in a matter of minutes, track firefighting resources in realtime, and share knowledge at a national scale is no longer a pipe dream. Firestory is delivering this to Australian fire agencies today. Firestory's cloud-native architecture allows it to ingest almost limitless amounts of data and distil it into a single source of decision-making intelligence on any device. The platform provides users with a simple to use, familiar experience, surfacing intelligence through a 3D spatial map, chronologically ordered incident feed or ‘Firestory’s’ and through push notifications.
The Black Summer fires highlighted the need for firefighting agencies to make significant changes in the way they fight fires - both locally and nationally. Firestory provides State Fire Agencies (SFA) with a standardised API driven platform with highly customisable workflows and feature sets, meaning user experience is not sacrificed for the sake of agency interoperability. Firestory is creating a national bushfire dataset that provides firefighters with enhanced situational awareness whilst reducing the cognitive overload associated with decision making - allowing SFA’s to respond with more confidence and accuracy which will ultimately reduce the impact of bushfires.
Core to Firestory’s cloud platform is ‘Firestory’, a 4D graph data structure that chronologically connects all the geospatial and time series data, metadata, predictions, social media, images, and videos related to an incident, redefining how we can visualise and interrogate events. Users are able to scrub through time to better understand how an incident unfolded, whilst allowing the platform to learn and improve its predictions and recommendations. Firestory’s enriched data provides greater flexibility in how we spatially communicate an incident to users, and enables Firestory to remain accessible in a browser even when the device is offline - a problem SFA’s had previously been unable to solve. Firestory’s mapping solution has been built from the ground up using vector 3D mapping tiles; vector tiles allow the system to selectively load in data relevant to the incident ensuring high-throughput, low-latency caching. In contrast, existing solutions load entire state or national datasets causing significant lag issues. Firestory uses machine learning image and object recognition, topic classification, and location pattern detection techniques to derive intelligence from non-traditional intel sources such as social media. The platform analyses and validates photos, videos, and posts before surfacing them to firefighters - providing them with greater situational awareness.