Bovisa 451 | Spot That Fire!

The Challenge | Spot That Fire!

Build a crowdsourcing tool for citizens to contribute to early detection, verification, tracking, visualization, and notification of wildfires.

WiFire

Wifire is a system that combines crowdsourced and satellite data with AI models to detect wildfires and alert people in danger. An intuitive UI lets citizens see the real-time and forecasted evolution of the fire, along with advices on how to behave.

The project originates from the necessity of facing and raising awareness about the phenomenon of wildfires, which have increased in the last years both in terms of quantity and severity.

A fast intervention in handling those fires and securing the surrounding population could avoid many disasters and environmental issues, particularly in places with a high fire hazard severity or where the practice of burning fields for agricultural reasons is largely spread.

The team brainstormed to find a solution to both raise awareness about the problem, improve the current emergency handling and enable citizens to play an active role in facing it, coming up with a cloud-powered app called Wifire.

At a conceptual level, the app acts as a transmitter and receiver of reports about ongoing fires, alerting the people which happen to be in danger via push notifications.

Each user who spots a fire can report it by simply opening the app and filling out a form that collects the main characteristics of the fire (e.g. what is burning, how it smells, how is the smoke). The report is then sent to the cloud service which opens a "possible report" and starts reviewing it to determine if there is actually a fire.

The review process uses data provided from NASA sources (e.g. FIRMS and AIRS) and satellites which are eventually combined into an AI model which uses image recognition and pattern analysis to determine whether the report is fake or not, using computer vision techniques to spot wildfires in satellite images and pattern recognition to analyze how the reports are distributed during an emergency and train an ML model.

Reports can also be verified by authorities (e.g. Fire Departments, Governments), which automatically mark the report as true, or by a peer-review process with the users who opted in a moderation program: (a group of users can decide to check incoming reports, like Google Maps's local guides initiative).

Once a fire is verified, the service sends a push notification to all the users in the area of the emergency, alerting them about the danger and providing a detailed page to track the evolution of the fire, see reports from other users and authorities and see the AI-predicted evolution of the fire.

Some of the challenges faced in developing the project are:

- how to validate reports received from users and how to build a trusting mechanism

- How to integrate various data providers into our service

- How can we make the app and the underlying service useful and engaging for users


This is the list of papers we took information from:

- Artés, T., Cortés, A., & Margalef, T. (2016). Large Forest Fire Spread Prediction: Data and Computational Science. Procedia Computer Science, 80, 909-918.

- FIRESENSE, F. D. Management Through a Multi-Sensor Network for the Protection of Cultural Heritage Areas from the Risk of Fire and Extreme Weather Conditions. FP7-ENV-2009-1244088-FIRESENSE, 2009,/http://www. firesense. euS.

- Iot, Matters. Technology helps first responders fight California fires. Network World, 10/24/2017, www.networkworld.com/article/3234655/internet-of-things/technology-helps-first-responders-fight-california-fires.html

- Zhong, X., Duckham, M., Chong, D., & Tolhurst, K. (2016). Real-time estimation of wildfire perimeters from curated crowdsourcing. Scientific reports, 6, 24206.

-Yin, J., Lampert, A., Cameron, M., Robinson, B., & Power, R. (2012). Using social media to enhance emergency situation awareness. IEEE Intelligent Systems, 27(6), 52-59.

- Wes Siler. Now you can forecast fires like the weather, 2018, www.outsideonline.com/2328341/how-forecast-wildfire

- Goodchild, M. F., & Glennon, J. A. (2010). Crowdsourcing geographic information for disaster response: a research frontier. International Journal of Digital Earth, 3(3), 231-241.

Link to cloud service github repo: https://github.com/lorenzofar/spaceapps-backend (still has work to do)

NASA Logo

SpaceApps is a NASA incubator innovation program.