Inspiration
Disasters like earthquakes, floods, and building collapses create chaotic situations where rescue teams must make critical decisions quickly. Modern rescue operations generate huge amounts of data from drones, sensors, and cameras, which can overwhelm responders during the most important moments.
Our team wondered how AI could help process this information without removing human judgment. Instead of replacing rescuers, we wanted to build a system that works alongside them. This idea led to HSARN (Human-Supervised Autonomous Rescue Navigator) — an AI assistant designed to support rescue teams while keeping humans in control of critical decisions.
What it does
HSARN is an AI-assisted platform that helps rescue teams analyze disaster zones more efficiently. It collects and processes data from sources like drones, cameras, and environmental sensors, and converts it into a real-time map of the affected area.
The system analyzes risks, highlights potentially dangerous zones, and suggests actions such as drone deployment or route adjustments. Instead of acting autonomously in critical situations, HSARN requires human approval before executing risky operations, ensuring safer and more controlled decision-making.
How we built it
HSARN combines data processing, AI analysis, and a visual command interface. Data from different sources is integrated into a unified model of the disaster environment.
AI models analyze this data to estimate risk levels and confidence scores for different areas. We built a mission-control style interface that displays a live map and system recommendations, allowing human operators to review and approve actions before they are executed.
Challenges we ran into
Handling multiple data sources and converting them into meaningful insights was one of our biggest challenges. Disaster environments involve different sensor types and data formats, making integration complex.
Another challenge was balancing automation with human control. We had to design the system so AI assists with analysis while still allowing humans to make the final decisions.
Accomplishments that we're proud of
We successfully designed a concept that focuses on human-supervised AI, prioritizing safety rather than full automation.
We also built a working prototype interface that demonstrates how AI recommendations and human decisions can interact in a disaster response scenario.
What we learned
This project taught us that AI is most powerful when it supports human expertise rather than replacing it. We also learned how to integrate multiple technologies into a single system and design tools that present complex data in a clear and useful way for real-world situations.
Log in or sign up for Devpost to join the conversation.