Cali Government Uses AI and AWS to Predict and Prevent Landslides

Author
.
2025

Overview

The Cali city government, responsible for risk management and emergency response, launched an initiative to reduce the impact of landslides in critical areas. The goal was to implement a predictive model using advanced technologies to enable early detection of high-risk zones and support more proactive interventions.

Challenge

  • The administration needed to analyze large volumes of urban data to better understand city dynamics and improve response capabilities.
  • There was no centralized system in place to integrate and process geologic and meteorological data from multiple sources in near real-time.
  • A cost-benefit analysis was needed to determine whether a cloud-based solution would be more viable than building an on-premise system.

Solution

  • A Data Lake with an automated continuous ingestion system was implemented to centralize and structure urban data.
  • Artificial intelligence and machine learning models were developed to extract, analyze, and predict risk areas using descriptive, predictive, and prescriptive analytics.
  • Core AWS services used include:
    • Amazon S3 (secure data storage),
    • AWS Glue (data processing and transformation),
    • Amazon QuickSight (data visualization).
  • A Total Cost of Ownership (TCO) analysis confirmed that the AWS-based architecture was more cost-effective and scalable than an on-premise alternative.

Impact

  • Significantly improved the city’s ability to anticipate and respond to landslide risks, reducing community impact.
  • Enabled real-time decision-making based on predictive analytics.
  • Promoted greater transparency and improved responsiveness to citizens’ needs.
  • The TCO analysis validated the cost-efficiency and long-term scalability of AWS services for urban data management.

Key Data Points