Geoflare

Empowering collaborative wildfire response with real-time strategy and action

Comment

⚡️ Try out the app @https://geo-flare-tpxz.vercel.app/

Background

As climate changes exacerbate wildfires into a global issue, response systems have failed to catch up: remaining slow and unreliable. In 2021, Wildfires burned over 7 million acres of land with 4.5 million people at risk of fire-related property damages.

At the same time, the government has tried to solve this issue by developing web apps tracking broad, geospatial locations of objects. However, these maps fail to convey case-by-case wildfire risks, leaving individuals uncertain about their safety. The unintuitive software limits the use cases of this app, making the barrier-to-use extremely high. The lack of effective government response is also exacerbating housing shortages by limiting the building materials of homes to those that meet wildfire building codes.

We noticed that the issue with wildfires wasn't a lack of resources, but rather an over-reliance on outdated, unsustainable methods of measuring fire safety. Government solutions merely identify the issue space, but fail to convey strategic actions to combat them.

We wanted to create a real-time, action-based application empowering collaboration between firefighters and operators. At the same time, we wanted to create a quick response-feedback mechanism that would effectively address ever-changing, unpredictable wildfire patterns.

This is why we built GeoFlare: a real-time platform that leveraged geospatial computer vision and artificial intelligence to rapidly process environmental data and provide immediate, strategic actions to best tackle it.

What is GeoFlare?

GeoFlare is a collaborative ecosystem empowering real-time decision-making during live wildfires. We were inspired by strides made in military defense startups like Palantir and Anduril to develop a comprehensive, action-based platform to allow firefighters to focus on saving lives, not planning strategy.

Here's how it works:

  1. Identification: Live risk assessment based on objects identified in satellite imagery. We leverage computer vision to place boundary boxes around critical elements such as bush clusters that could create unanticipated impediments to a firefighter.

  2. Live Chat: Live strategy co-pilot chat bot allows you to interact with a live chat that recommends strategies based on the threats identified in the computer vision detection. Alt text

  3. Strategy: Path optimization by searching through an environment, leveraging information collected from computer vision, finds the most optimal route to a specific house address. It also measures the "priority level" of each home page on the severity of their wildfire situation.

How GeoFlare was built

GeoFlare was built with a team consisting of a full-stack developer, front-end developer, product designer, and ML researcher. This diverse team makeup allowed us to explore different perspectives on developing our product for different use cases.

GeoFlare's design was developed through a full-stack design-thinking framework from ideation to secondary market research (need-finding), user flow diagramming, low-fidelity wireframes, and high-fidelity prototypes. Our designer also developed a fully functional landing page for future scaling of our product. Everything was designed on Figma

GeoFlare's tech was designed through a discussion of core technologies that we were each most uniquely strong in. The front end was created using React, SASS (for the landing page), Tailwind.css, Typescript, and various other software. The backend was developed with Node.js with Gemini/pytorch integrations for the ML model and the chatbot interface.

On the machine learning side, we also custom-trained our model for the most accurate, threat-based object identification consistent with the danger-related information we wanted to pass to the users.

UI/UX

User Flows

Flows to understand how users naturally flow through the page and how we can adapt our product to the quick thought-processing mechanisms.

Lo-Fi Wireframes

Low Fiedlities helped us map out the structural layout of our idea. It also helped with quick iterations and jumping across different idea concepts.

Branding

Futuristic, geo-spatial branding was based around popular defense companies that we were fans of, specifically Palantir and Anduril (also first time trying out this kind of UI!)

High Fidelity and Prototypes

We made a landing page along with final prototyped versions of the dashboard web app.

Challenges we ran into

Design: designing for geospatial UI, computer vision, and dynamic animation states was a first for me. Designing for dynamic dashboard systems combining static UI and live transformations of data in real-time was also challenging, but helped refine the product concept. - Jasmine

Takeaways

Accomplishments that we're proud of

Building a functioning MVP with a full, conceptual dashboard and landing page.

What we learned

Building a multi-dimensional, action-based platform based on real-time strategy and updates.

What's next for Geoflare

Firefighters encounter plenty of different and unpredictable experiences. Given this, we plan to expand our technologies to include more datasets. This will account for any possible scenario a firefighter may encounter by incorporating resources and recommendations tailored to each suggested strategy. For example, training additional datasets including unpredictable and rare scenarios can provide firefighters in these scenarios with strategic predictions of never-before-encountered situations.

+ 1 more
Share this project:

Updates