What it does
ORCAS is an intelligent robot designed to clean up oil spills autonomously. It uses AI to locate spills, then collects and filters the oil using eco-friendly coconut-mica filters—protecting both the ocean and people from toxic exposure.
Your inspiration
We were deeply moved by the impact of oil spills in countries like the Philippines and Japan—places with rich marine biodiversity and communities that rely on the sea. Traditional cleanup methods are often slow, wasteful, and risky. We asked: what if we could clean oil spills without endangering people or harming the environment? That’s when we envisioned a self-operating robot that could do it all—navigate, detect, and clean—using safe, natural materials. We combined the oil-absorbing power of coconut husk and mica with AI to make a smarter, safer, and more sustainable solution.
How it works
ORCAS works like a robotic ocean vacuum that maps, navigates, and cleans oil spills automatically. First, it uses machine learning and sensor technology to scan the water and locate oil slicks. Then, guided by AI, it moves toward the spill and activates a suction mechanism to collect the oil. Inside, it filters the oil using layers of coconut husk and mica powder. The coconut absorbs oil while the mica repels water, allowing efficient and eco-friendly separation. A central control system handles real-time decisions, adjusting operations based on wave conditions, oil density, and surrounding currents. This allows ORCAS to operate reliably in both tropical and temperate marine environments with little to no human input.
Design process
Our process started with a clear goal: create a safer, smarter way to handle oil spills. We began by studying current technologies and consulting with marine engineers and AI specialists from the Philippines and Japan. Then we built prototypes to test different filter materials. Lab tests showed that coconut husk combined with mica had superior oil-absorbing properties. We added water sensors, AI-driven mapping systems, and an initial pathfinding algorithm called “Predicted First Left.” Using Arduino and vacuum prototypes, we simulated ocean conditions in controlled tanks. Field testing in a simulated sea setup followed—complete with waves, wind, and various oil types. Throughout, we used expert feedback to improve efficiency, speed, and responsiveness. Each version became more capable, and today’s ORCAS prototype can recover over 57% of oil in real-world-like tests. We continue to refine AI calibration, wave resistance, and environmental adaptability for broader deployment.
How it is different
ORCAS is unique because it doesn't just improve oil cleanup—it completely reimagines it. While traditional methods rely on risky human labor or chemical dispersants that can harm the environment, ORCAS works autonomously and safely. It uses smart, AI-powered mapping to detect oil, navigates on its own, and filters oil using biodegradable coconut-mica filters that outperform cotton and hair. Its modular design means it can adapt to different climates and oil types, working effectively in both Philippine and Japanese waters. It’s scalable, efficient, and environmentally responsible. Unlike other systems, ORCAS doesn’t just respond—it learns, adjusts, and performs smarter with every deployment, making it a standout solution in marine disaster response.
Future plans
We’re moving forward with open-sea testing in the Philippines and Japan to see how ORCAS performs in real marine conditions. We also plan to upgrade the navigation algorithm to respond more dynamically to oil movement. Our team is exploring partnerships with environmental agencies and port authorities for official adoption. Eventually, we hope to scale ORCAS for global use, offering an affordable, eco-friendly oil spill response solution that supports marine conservation and public safety in vulnerable coastal areas.
Awards
₱50K Grant from 3 Minute Research Challenge, DLSL; Presenter at Wakasa Forum, Japan; Delegate at Global Science Forum, Fujishima HS; National Delegate for DICT Startup Challenge; IP filed with IPOPHL; Incubatee at Nexus, DLSL; Awardee at Harvard Social Visionary and Pitch; Finalist at SEA-CICSIC 2025, Xiamen Univ MY
Share this page on