Within the MMS2 program, Indonesia’s next-generation Maritime Meteorological System (MMS), BMKG (Indonesia’s Meteorological, Climatological and Geophysical Agency), and CLS are jointly developing a new generation of Artificial Intelligence models designed to strengthen weather and ocean forecasting, improve disaster preparedness, and ultimately better protect citizens.

MMS2 embodies a shared vision: combining international expertise and local knowledge to build operational services that will support Indonesia for decades to come.

We met Francisco Dos Santos, Oceanographer and AI Development Work Package Leader for the MMS2 program, to better understand how this ambitious co-development is shaping the future of environmental intelligence.

Two AI Models Are Currently Being Developed Within MMS2. What Are Their Objectives?

The two models address complementary challenges while pursuing the same ambition: providing authorities with more reliable, more complete, and more timely environmental information.

The first model focuses on ocean currents. It uses Deep Learning models to reconstruct missing observations from local coastal HF radar networks and generates short-term forecasts. By ensuring continuity in ocean observations, it strengthens services supporting maritime safety, navigation, and search and rescue operations.

The second model focuses on rainfall nowcasting. By combining weather radar observations, satellite imagery, and numerical weather models, it generates synthetic radar coverage over areas where no radar infrastructure exists. The objective is to detect heavy rainfall events earlier and more accurately across the Indonesian archipelago.

“We are not developing Artificial Intelligence for its own sake. We are developing better information for decision-makers.”

Why Use AI Instead of Relying Solely on Traditional Numerical Models?

Traditional numerical models are extremely powerful, but they require considerable computational resources, especially when operating at very high spatial resolution.

Artificial Intelligence offers another approach.

By learning from years of radar observations and complementary environmental datasets, the models are able to reconstruct missing information almost instantly. Once trained, they provide continuous synthetic observations and short-term forecasts much faster than conventional numerical simulations.

The objective is not to replace observed data, but to ensure that operational services always have the information they need.

Discover How CLS and BMKG Combine Satellite Data, Environmental Science, and Artificial Intelligence

Indonesia’s geography makes weather forecasting exceptionally complex.

Thousands of islands, mountainous terrain, and tropical weather dynamics generate highly localized rainfall events that can develop within minutes.

While weather radars provide invaluable observations, they do not cover the entire country and can only provide information up to the present.

The AI model addresses this challenge by combining radar observations with satellite imagery and numerical weather prediction data to create a synthetic radar coverage. The objective is to produce radar-quality information even in areas where no radar exists and generate short-term forecasts.

The result is faster and more detailed rainfall nowcasting, helping authorities identify potentially dangerous weather situations as they emerge.

“Every minute gained can help improve emergency response.”

What Impact Will These Models Have for Citizens?

Their ultimate purpose is simple: enabling faster and better-informed decisions.

By providing more accurate information on ocean conditions and heavy rainfall events, the models will support emergency services, improve disaster preparedness, and help authorities anticipate floods and other weather-related hazards, and improve search and rescue operations at sea.

Over the longer term, these new datasets will also contribute to better infrastructure planning and stronger climate resilience across Indonesia.

Artificial Intelligence therefore becomes a powerful decision-support tool serving public authorities and, ultimately, the citizens they protect.

“The real innovation is co-development.”

MMS2 Seems to Be Much More Than a Technology Transfer?

Absolutely.

One of the defining characteristics of the project is that it has been designed as a genuine partnership between BMKG and CLS.

CLS contributes expertise in Artificial Intelligence, environmental modeling, operational systems, and data science.

BMKG brings deep meteorological expertise, local operational knowledge, and a unique understanding of Indonesia’s climate and geography.

Together, both teams are building solutions specifically adapted to Indonesia’s needs while ensuring that BMKG will progressively gain the capability to operate, maintain, and further develop these models independently in the future.

Technology for Good: Building AI That Matters

Behind every AI model developed by CLS lies far more than an algorithm. It is the result of a unique ecosystem combining satellite technologies, Earth observation expertise, environmental sciences, and advanced data engineering.

At the heart of this ecosystem is the CLS DataLab, where data scientists, data engineers, and environmental experts work side by side to design operational AI solutions tailored to real-world challenges. Their expertise is supported by secure computing infrastructures, dedicated high-performance servers operating 24 hours a day, 365 days a year, and a unique environmental data lake bringing together decades of satellite observations, in situ measurements, and numerical models.

These capabilities enable CLS to develop high-value Artificial Intelligence models, and also to deploy, operate, and continuously improve them within mission-critical operational environments.

Together with partners such as BMKG, CLS transforms environmental data into trusted operational intelligence, providing authorities with faster, more comprehensive, and more actionable information to anticipate extreme weather events, strengthen climate resilience, and better protect citizens.

This is our vision of Technology for Good: developing Artificial Intelligence that serves a purpose. AI designed not simply to automate tasks, but to enhance scientific expertise, support public decision-making, and help build a safer and more sustainable future.