Designing to save our forests from forest fires: Edge Deployment of AI to predict forest fires and fires spread patterns.

Designing to save our forests from forest fires: Edge Deployment of AI to predict forest fires and fires spread patterns.

Designing to save our forests from forest fires: Edge Deployment of AI to predict forest fires and fires spread patterns.

In collaboration with

In collaboration with

In collaboration with

✨tldr.

✨tldr

We designed a robotic system + dashboard suite to predict wildfire and fire spread patterns by capturing environmental data and images of the forest floor.

We designed a robotic system + dashboard suite to predict wildfire and fire spread patterns by capturing environmental data and images of the forest floor.

TIMELINE

1 Year
March 2024 - March 2025

ARTEFACTS

Robotic System
Dashboard Design WebApp
Remote Control WebApp

MY ROLE

Team Lead
Product Designer
Researcher
Usability Tester

TREX - Terrain Resource Evaluation Expert

Overview

Design Question

Design Question

How can we design and develop a versatile, power-efficient, and robust device that effectively integrates multiple sensors to acquire precise data for distinct applications (such as tracking a moving object from a set height or imaging ground-level vegetation along a defined path), while accommodating specific operational parameters like movement mechanisms, environmental conditions, and sustained power requirements?

How can we design and develop a versatile, power-efficient, and robust device that effectively integrates multiple sensors to acquire precise data for distinct applications (such as tracking a moving object from a set height or imaging ground-level vegetation along a defined path), while accommodating specific operational parameters like movement mechanisms, environmental conditions, and sustained power requirements?

Solution

Solution

TREX - Terrain Resource Evaluation Expert is an autonomous, multi-sensor robotic platform that's easy to set up and deploy. It autonomously collects precise data, including images and environmental readings, greatly reducing the time and effort of manual collection. The system is designed to be power-efficient and robust, ensuring reliable performance in challenging environments while providing a consistent data stream for advanced AI analysis.

Key Metrics & Impact

Key Metrics & Impact

95% images captures in focus

95% images captures in focus

95% model accuracy as compared to human readings

Problem and Context

Wildfire fuel measurement is slow, labor-intensive, and inconsistent. Manual stick counting methods require on-site effort, are prone to human error, and vary by terrain.

Fuel technicians typically work in teams to manually count measurement sticks and keep an accurate tally

The team gathers to evaluate the data's accuracy and consistency, working together to minimize subjectivity.

Fuel technicians manually count the sticks

Slope measurements may vary based on heights

i am open to full time work.

i am open to full time work.