Inspiration
During major storms and natural disasters, people rely on the internet and cell networks to reach emergency services, check on loved ones, and receive updates. However, not every neighborhood has the same level of infrastructure resilience. Wealthier areas often have redundant power and backup fiber connections, while underserved communities experience long outages and slow recoveries. We built this simulation to visualize that inequality in real time and to show how infrastructure investment directly impacts network stability during crises.
What it does
The simulation models how a moving storm affects the telecommunications infrastructure across different neighborhoods, each with unique traits like elevation, redundancy, and repair crew access. It tracks the storm’s physics (wind, surge, and pressure), calculates damage to towers and fiber nodes, and then estimates how that damage impacts latency, download speeds, packet loss, and call quality. The system updates dynamically over time as the storm evolves, showing which areas stay connected and which are at risk of complete network failure.
How we built it
We built a backend physics engine that simulates how storms impact digital infrastructure using real civil and telecom engineering formulas. Wind pressure is calculated using the standard engineering formula: 0.00256 times the square of wind speed. The storm is categorized using the NOAA Saffir-Simpson scale, and its impact on each zone decreases with distance using exponential decay. We also factor in elevation, flood risk, and terrain to simulate how storm surge and physical stress vary across neighborhoods.
Each zone has its own infrastructure profile, including cell towers, fiber nodes, power backup, and redundancy. The model calculates failures for each component based on storm pressure and local vulnerabilities. For example, towers exposed to high wind in flat terrain are more likely to fail, and zones without backup power lose service faster.
From there, we compute network performance metrics like latency, download speed, packet loss, and voice quality. These metrics are based on real-world telecom standards, including formulas for jitter and mean opinion score (MOS). All of this data is exposed via a backend API.
We built a frontend dashboard in React that polls this API every few seconds. It displays real-time metrics for each zone, shows live storm data, and flags areas at risk. The result is a fully dynamic system that helps visualize how infrastructure inequality affects network resilience in real time.
Challenges we ran into
One of the main challenges was connecting physical storm behavior to realistic network degradation in a way that stayed technically sound across different conditions. Modeling wind pressure, surge, and terrain in a dynamic system meant balancing many interacting variables, each of which influenced different infrastructure components. Translating those physical impacts into meaningful network metrics like latency, throughput, and packet loss required careful calibration.
Performance was another challenge. Since the simulation updates in real time and processes multiple zones with unique attributes, we had to keep calculations efficient without cutting corners on the physics or telecom logic. We also spent time tuning the visual outputs and thresholds so that both technical and non-technical users could understand what was happening and why, without losing the underlying accuracy of the model.
Accomplishments that we're proud of
We created a fully functioning, interpretable storm simulation that produces believable results. The model can demonstrate why infrastructure inequality matters by showing how underserved areas lose connection faster and take longer to recover. We’re proud of how much depth we achieved while keeping the math transparent and explainable. The simulation is also modular, meaning new cities or weather conditions can be added easily. It’s both a technical and educational tool.
What we learned
We learned how to merge physical modeling, network reliability analysis, and software engineering into a single system. We also realized that visualization and storytelling matter as much as accuracy,as people grasp the issue of infrastructure inequality much faster when they can see it simulated. Finally, we learned that even a slightly simplified mathematical model can yield powerful insights if it’s grounded in real-world logic and data.
What's next for Networking Simulation for Underserved Communities
We plan to add more neighborhoods and real historical storm data to validate the model. We also want to let users adjust parameters like infrastructure investment, crew allocation, or backup power to explore how different strategies change outcomes. Another goal is to integrate the tool with emergency response planning so cities can use it to prioritize resource deployment. Over time, we hope to turn this into a public platform that helps visualize the “digital divide” in disaster resilience.
Log in or sign up for Devpost to join the conversation.