When Data Meets Earthquakes: Saving Lives with Analytics
Imagine living on an island where the ground never truly rests.
That’s life on Sulawesi, an island in Indonesia that is sitting at the crossroads of four tectonic plates. That means one thing: earthquakes are part of daily life.
But not all areas face the same risk. Some experience frequent mild tremors, while others face rare but devastating shocks. How do we identify which communities are most vulnerable?
That’s what the Probability Team from the IPB University in Indonesia decided to answer in the global student competition The Curiosity Cup 2025.
They brought together two critical data sources:
- Earthquake records from Indonesia’s meteorological agency (BMKG)
- Population density data from the national statistics office (BPS)
To make sense of it all, they used SAS® OnDemand for Academics. With it, they applied a method called clustering — grouping regions that share similar risk patterns. And to find the “sweet spot” for grouping, they tested different options before deciding on three clusters.
Here’s what they discovered:
- Cluster 1: Many quakes, but mostly mild → education and earthquake-ready buildings help.
- Cluster 2: Very high population density → even small quakes could be disastrous.
- Cluster 3: Stronger, deeper quakes → stronger infrastructure and better coordination needed.
Each community needs its own tailored strategy and thanks to this study, policymakers have a clearer map for building safer, more resilient futures.
Very well done, your work truly belonged to the top 15 teams in this global competition for 2025. Congratulations to Wawan Saputra, Muh. Sunan, Muh. Akbar Idris, Muhammad Rizal and their Faculty Advisor Bagus Sartono.
The full paper can be seen as pdf under the main headline of this article.
