How many times have we received a bushfire warning on our phones during the Australian summer? Every summer in Australia, bushfire alerts appear on our phones, reminding us how common and dangerous these events have become. Existing platforms such as VicEmergency and Fires Near Me Australia, developed by rural fire services, provide valuable push notifications and live incident maps that keep communities informed about active fires.
However, these systems primarily report fires after they have already been detected or reported. For rural communities located far from monitoring infrastructure, this can mean that warnings arrive only once a fire has already spread and begun causing significant health, environmental, or infrastructure impacts.
This inspired us to explore how technology could shift bushfire response from corrective measures to preventative action, using early environmental signals such as changes in air quality to identify potential bushfires before flames become visible.
Aerosafe provides a seamless integration of hardware and software. Environmental sensors placed around regional victoria in urban and forested areas continuously collect data on CO, CO₂, VOCs, temperature, humidity, and particulate matter (PM). This data is transmitted to a development board, which sends the information to our mobile app. The app uses a data-driven detection model to process the environmental data, update a live map of the user’s surroundings, and issue alerts when conditions indicate a potential bushfire risk.
Based on the time constraints of this hackathon, we created our application using simulated sensor data, as it would not have been feasible to fully construct and deploy a physical prototype within the 48-hour timeframe. Instead, we focused on designing the system architecture and demonstrating how the platform would function in a real-world deployment. The proposed hardware system consists of solar-powered environmental sensor nodes equipped with a ESP32 Ultra‑Low Power Development Board and an TP4056 Li‑ion Battery Charging Module . These nodes would continuously monitor environmental conditions such as volatile organic compounds (VOCs), temperature, and humidity to identify early indicators of potential fire ignition. For the software side, we built our application using React Native, go, etc to rapidly prototype the user interface and simulate incoming sensor readings. Sensor data would be ingested and analyzed using Elasticsearch, enabling real-time data indexing, monitoring, and historical analysis. Using this data, the platform can detect abnormal environmental patterns, trigger alerts, and visualize risk levels across sensor locations. We also modeled a full 3D prototype of the physical sensor device, including the solar panel, environmental sensors, and enclosure, using SolidWorks to demonstrate how the hardware would be deployed in real environments.
One of the main challenges was researching and selecting appropriate sensors and sustainable power solutions for the device. We explored different environmental sensors capable of detecting early indicators of fire, while also considering how the system could be powered long-term using solar energy in remote areas. Another challenge was designing a system capable of handling large amounts of data from many IoT sensors deployed across bushfire-prone regions. Simulating multiple sensor nodes and managing the flow of environmental data required careful consideration of how the platform would ingest, store, and analyze real-time readings.
We also faced a learning curve when building the mobile and web interface, as our team had limited prior experience with mobile development. Developing the application and designing an intuitive dashboard required experimenting with new tools and frameworks. On the hardware side, creating a physical representation of the device in SolidWorks was challenging. We had to learn how to model and assemble the different components, including the enclosure, solar panel, and sensors, which required several iterations before achieving a workable design. Finally, we built our website using Framer despite having no prior experience with the platform. Learning how to structure the layout, connect pages, and present our project clearly took additional time, but it ultimately helped us communicate our idea more effectively.
One of our biggest accomplishments was being able to design and build an entire concept. That is including a working app prototype, a website, and a complete 3D model of our device all in under 24 hours. We spent the first day of the hackathon planning, researching the problem space, and deciding on the most effective approach before beginning development. Once we started building, our team worked closely together to learn new tools and solve technical challenges as they came up. Many of us were working with unfamiliar software, but by collaborating, sharing research, and helping each other troubleshoot, we were able to make rapid progress in a short amount of time.
We are also proud that our project was driven by a shared interest in environmental protection and finding ways technology can positively impact communities. Creating a solution aimed at improving early wildfire detection made the project especially meaningful for our team, and we are proud of what we were able to accomplish together within the limited timeframe of the hackathon.
Throughout the hackathon, we learned a lot about designing a complete system that combines hardware, software, and data analysis. Researching environmental sensors and sustainable power solutions helped us better understand how IoT devices could be deployed in remote areas for real-world monitoring. We also gained experience working with new tools and technologies. Building the application required us to quickly learn new development platforms and think about how sensor data could be visualised and interpreted in a meaningful way. Additionally, using SolidWorks to model the device gave us insight into the process of designing and assembling hardware components. Another key takeaway was learning how to manage large streams of environmental data and think about how systems like Elasticsearch could be used to analyse patterns, monitor conditions in real time, and detect anomalies.
Most importantly, we learned the value of collaboration under time pressure. By sharing knowledge, dividing tasks, and supporting each other while learning unfamiliar tools, we were able to turn an initial idea into a complete concept within a very short period of time.
The next step for AeroSafe is moving from simulation to real-world deployment. We plan to build a physical prototype of our solar-powered sensor node and begin testing it in controlled environments before deploying it in bushfire-prone regions. On the data side, we aim to expand our platform by integrating advanced analytics using Elasticsearch to process large volumes of environmental sensor data collected over time. By analyzing historical trends in temperature, humidity, air quality, and VOC levels, the system could identify patterns associated with early fire conditions and detect anomalies in real time.
In the future, this data could also be combined with external factors such as weather conditions, seasonal changes, and long-term climate trends related to global warming. Using machine learning and forecasting tools within the Elastic ecosystem, AeroSafe could move beyond simple detection and begin predicting periods of elevated wildfire risk.
Ultimately, our goal is to scale AeroSafe into a distributed network of low-cost, solar-powered sensors that provide early warnings to communities and emergency services, helping reduce response times and mitigate the impact of bushfires.
Built With
- docker
- expo.io
- go
- mapbox
- postgis
- postgresql
- react-native
- timescaledb
Log in or sign up for Devpost to join the conversation.