Germaine Ng

Weather Box
AI-driven, Real-time Weather Intelligence for Wildland Fire Operations

Problem
Identifying changing environmental conditions is a critical task for Fire Behavior Analysts (FBAN) and Incident Meteorologists (IMET). To take the necessary measurements, field observers have to hike many miles into remote locations, where traditional automated weather stations cannot be deployed.

THE SOLUTION
Turning manual observation into connected intelligence.
COLLECT->TRANSMIT -> RELAY -> VISUALIZE -> DECIDE
We designed Weather Box as an integrated system connecting:

System Overview

It is a system that includes weather data (temperature, relative humidity, wind speed and wind direction) collection devices, LoRa Mesh Network to send weather data back in low signal coverage areas, a relay and two webapps for users to access data both in the field and in the office.

Functional Test
Conducted functional test to ensure technology addressed the real workflow rather than simply adding another data collection tool.
Prototype
Iterated prototype based on the feedback from usability tests.
Usability Test 1.0 & 2.0
Evaluated the proposed system through usability testings, using the results to refine the overall experience.
SME Interview
Conducted 4 SME interviews from DNR and NOAA to define the problem scope.
Secondary Research
Conducted secondary research and competitive analysis.
Approach
My biggest takeaway:
Great technical products start with understanding the mission, the people, and the problem- not the technology.
What I Bring
My role
Product Strategy
User-Centered Design
Technical Fluency
Cross-Functional Leadership
Problem definition · Requirements · Prioritization · Systems thinking
UX research · SME interviews · Usability testing · Product design
IoT · Sensors · LoRa networking · Data systems · Web applications
Connecting users, stakeholders, and technical teams to move a product from problem → prototype → solution.