Moved by firsthand accounts from relatives in post-earthquake Turkey, we were was inspired to create the Expand-a-Conda for our hackathon project. In response to stories of friends lost and agonizing waits for rescue, we developed an innovation born from real-life experiences, designed to swiftly detect and locate individuals buried in debris. This initiative combines technology with empathy, offering a lifeline in the aftermath of disasters.
Expand-a-Conda is ingeniously designed to detect people buried in debris using image segmentation and AI-powered pathfinding. It incorporates soft robotics by building a cheap and indestructible vine robot, capable of navigating through inaccessible terrains.
We leveraged a finetuned version of the NVIDIA MIT-B5 model to detect humans trapped in rubble. Utilizing 3D printing technologies, we mounted a camera on our indestructible vine robot. The video feed from the camera is reverse-proxied through Cloudflare and processed by Google Cloud.
Our sole challenge was computing power. With limited resources, we endeavored to maintain accuracy while achieving acceptable performance. Currently, we face a low frame rate, which we aim to improve by integrating local GPUs for enhanced computing capabilities.
We're proud to have designed a solid proof of concept and developed a fully-functional prototype that demonstrates its feasibility and potential impact in disaster scenarios.
The constraint of limited computing power pushed us to optimize our algorithms and processes rigorously. This endeavor sharpened our problem-solving skills and taught us to innovate within real-world constraints.
Our roadmap includes scaling up the computing power to achieve a higher frame rate, resulting in faster and more accurate detection capabilities. Additionally, we plan to incorporate a user identification system using the OpenCV library, which will aid in the swift identification of victims.
🚧 Under Construction 🚧
This project is no longer being developed, and new features are not being constructed!