AQUILA

💎 Idea

The Problem: Water scarcity is a critical global crisis, heavily exacerbated by aging infrastructure. In vast water distribution networks, underground leaks can go undetected for months, wasting millions of liters of clean water before they ever surface. Traditional leak detection methods (like acoustic sensors or manual patrols) are reactive, slow, and prohibitively expensive to scale across thousands of kilometers.

The Solution: Enter AQUILA. We have built a highly automated geospatial orchestration engine designed for early, proactive water leak detection. By fusing multispectral satellite imagery, high-resolution topographic models, and cadastral databases, AQUILA filters out noise (like forests or physical shadows) and computes advanced terrain indices (NDVI, NDWI, PSRI, NIRRR). This allows us to identify abnormal soil moisture and unexpected vegetation growth near pipelines, spotting invisible leaks from space before they become catastrophic surface ruptures.

🛰️ EU space technologies

Data Used: AQUILA relies heavily on Earth Observation data, specifically from the Copernicus Sentinel-2 mission, complemented by high-resolution digital elevation models (DEM).

How it brings value: Sentinel-2 provides high-frequency, high-resolution (10m) multispectral bands (including Near-Infrared and Short-Wave Infrared). These specific bands are the core of our mathematical pipeline, enabling the precise calculation of water and vegetation indices over massive geographical areas. Copernicus allows AQUILA to shift leak detection from a localized, ground-based effort to a highly scalable, continuous, and cost-effective monitoring system capable of protecting entire regions.

🌊 EU Space for Water

Challenge Addressed: We are addressing the challenge of resilient water infrastructure and management.

Contribution: By pinpointing underground leaks at their earliest stages, AQUILA fundamentally changes how utility companies manage their distribution networks. Our platform helps prevent massive non-revenue water loss, reduces the immense energy footprint required to pump wasted water, and allows maintenance teams to prioritize interventions based on predictive geospatial insights rather than waiting for structural failures.

🤼 Team

- Pau Campillo (Project Lead & MLOps Engineer) Drives the core software architecture, Apache Airflow orchestration, and the integration of our complex data processing pipelines.

- Maksym Plashkin (Geospatial Data Scientist) Specializes in GIS processing, multispectral satellite imagery analysis, and the mathematical optimization of terrain-use indices.

- Marc Cassanmagnago (Infrastructure & Software Engineer) Focuses on backend development, Docker containerization, and ensuring our models run efficiently in scalable production environments.