Technology alone won’t save the planet—but in the right hands, it can completely change how communities protect their environment. And that’s exactly what a team of Saint Louis University BSIT students set out to prove on the national stage. Their work doesn’t just talk about sustainability; it turns data and code into real-world tools for climate resilience. But here’s where it gets even more interesting… this solution comes from students, not a big tech company or government agency.
A team of 4th-year Bachelor of Science in Information Technology (BSIT) students from the School of Accountancy, Management, Computing and Information Studies (SAMCIS) at Saint Louis University has secured a coveted Top 10 finalist spot in the BPI-DOST Innovation Awards 2025. The event took place on 4 December 2025 at the Edsa Shangri-La in Manila, Philippines. Their recognized project is titled “Cordimap: Hybrid Machine and Deep Learning-Based Web Application for Interactive Mapping of Land Cover and Water Balance in the Cordillera Administrative Region (CAR).”
Guided by their adviser, Dr. Beverly Estephany P. Ferrer, the team developed a cutting-edge web-based platform that combines machine learning (ML) and deep learning (DL) to confront pressing environmental issues in the Cordillera Administrative Region. CAR, an ecologically vital region in Northern Luzon, is under mounting stress from deforestation, rapid urban growth, and the far-reaching impacts of climate change. These pressures manifest in changing weather patterns, increasing water scarcity, soil and land degradation, and the loss of natural habitats—problems that are often difficult to monitor and manage using traditional tools.
What Cordimap Actually Does
Cordimap brings together advanced AI models to make environmental data more understandable, accessible, and actionable. It uses the U-Net deep learning architecture to automatically classify different types of land cover—such as forests, built-up areas, and agricultural land—based on Sentinel-2 satellite imagery. This means it can turn raw satellite data into meaningful, color-coded maps that show how land is being used and how it is changing over time.
On top of that, the system integrates an XGBoost-based time series forecasting model to predict water balance trends using ESRI-derived water balance data. In simpler terms, water balance refers to the relationship between water coming into an area (like rainfall) and water going out (through processes like evaporation, plant use, and runoff). By estimating how this balance changes over time, Cordimap offers an early look at potential water shortages, drought risks, or stress on local water resources.
The result is a real-time, interactive mapping platform where users—especially local governments, planners, and community stakeholders—can:
- View updated land cover maps across the Cordillera Administrative Region.
- Track how landscapes are changing, such as forest loss or urban expansion.
- Examine projected water balance trends in different areas.
- Use reliable, AI-powered insights to plan land use, manage water resources, and prepare for environmental risks.
And this is the part most people miss: tools like Cordimap don’t just visualize data; they help shift decisions away from guesswork and toward evidence-based governance.
Why This Matters for Climate and Communities
The Cordillera region is not just scenic—it is an important ecological zone that supports communities, agriculture, biodiversity, and water systems downstream. However, with urbanization and deforestation accelerating, green spaces are shrinking and more surfaces are becoming impermeable (like concrete and asphalt). This can worsen flooding, reduce groundwater recharge, and increase the likelihood of landslides and other natural hazards.
At the same time, climate change is intensifying evapotranspiration (the combined process of water evaporating and plants releasing moisture), making rainfall more unpredictable, and raising the risk of droughts and water shortages. When these factors interact, they can push communities into a cycle of resource depletion and disaster vulnerability.
Cordimap directly responds to these realities by providing:
- Detailed, location-specific views of current land use.
- Forecasts of water balance that help anticipate shortages and stress.
- Accurate overlays using official political boundaries, enabling province- and municipality-level analysis.
With this information, government agencies and planners can draft better policies related to zoning, watershed protection, reforestation, urban development controls, and disaster preparedness. For example, local leaders could identify which areas are losing forests fastest, where reforestation would be most impactful, or which communities might be at higher risk of water scarcity in the coming years.
A Tool for Sustainability and Global Goals
Cordimap is not just a local mapping tool; it is aligned with global sustainability frameworks. The system supports the achievement of the United Nations Sustainable Development Goals (SDGs), particularly:
- SDG 11: Sustainable Cities and Communities – by helping urban areas grow more responsibly through informed land-use and resource planning.
- SDG 15: Life on Land – by enhancing the monitoring and protection of terrestrial ecosystems, forests, and biodiversity.
By integrating environmental data, predictive modeling, and user-friendly visualization, Cordimap helps bridge the gap between global sustainability targets and local, on-the-ground decision-making. Some might argue that relying on AI models could create overconfidence in projections—but could it actually be riskier to keep making environmental decisions without these kinds of tools?
Scalable Beyond the Cordillera
While Cordimap is designed for the Cordillera Administrative Region, its architecture makes it inherently scalable. Many other regions in the Philippines—and even in other countries—face similar issues: forest loss, rapid development, water stress, and climate-related hazards. Because Cordimap is based on accessible satellite imagery and a flexible modeling framework, it can be adapted to new areas with comparable environmental data.
This raises an important and potentially controversial question: Should governments start standardizing AI-powered environmental monitoring systems like Cordimap as part of their official planning process? Some decision-makers may feel uneasy about depending on algorithms, but ignoring these tools could mean missing out on critical insights that save lives, livelihoods, and ecosystems.
Student Innovation on a National Stage
The Cordimap team’s success at the BPI-DOST Innovation Awards 2025 showcases how young innovators can meaningfully contribute to both technological advancement and environmental stewardship. Their project demonstrates how data science, machine learning, and web technologies can move beyond academic exercises and become practical, impactful solutions.
This achievement also underscores Saint Louis University’s strong commitment to forming leaders capable of addressing complex, real-world challenges. By supporting projects like Cordimap, the university helps nurture professionals who are not just technically skilled, but also socially and environmentally engaged.
Congratulations to the SLU BSIT students and their adviser, Dr. Beverly Estephany P. Ferrer, for this remarkable accomplishment. Their work stands as a powerful example of how innovation in information technology can be harnessed to confront environmental crises, offering hope and practical tools for communities in the Cordillera and beyond.
(Article adapted from content by Saint Louis University | Original article by Kasima Rose M. Mendoza | Photos by SAMCIS)
Project Snapshot
- Project Title: Cordimap: Hybrid Machine and Deep Learning-Based Web Application for Interactive Mapping of Land Cover and Water Balance in the Cordillera Administrative Region
- Project Adviser: Dr. Beverly Estephany P. Ferrer
- Institution: Saint Louis University
Cordimap addresses the escalating environmental pressures in the Cordillera Administrative Region—urbanization, climate change, and deforestation—that reduce green cover and increase impermeable surfaces. These changes heighten the risk of natural disasters while climate change drives higher evapotranspiration, unstable rainfall patterns, and more frequent droughts and water shortages. In response, Cordimap provides a hybrid ML–DL web solution for land cover classification and water balance forecasting, strengthening data-driven environmental planning and resource management.
By combining U-Net for land cover mapping with XGBoost for time-series prediction of water balance, and presenting results through an interactive interface with accurate political boundaries, Cordimap supports government agencies in drafting and refining policies for resource management, climate adaptation, and disaster risk reduction. In doing so, it actively contributes to the realization of SDG 11 and SDG 15, and promotes a more informed, holistic response to environmental challenges linked to urban growth, forest loss, and a changing climate.
Now it’s your turn: Do you think governments and local planners should trust AI-driven tools like Cordimap enough to base major land-use and water policies on them, or should they remain only as ‘supporting’ references? Where do you personally draw the line between human judgment and algorithmic insight in environmental decision-making? Share your thoughts—do you strongly agree with this direction, or do you see risks that others might be overlooking?
