event_available

ICA 2026 |Camera-ready deadline: 25th September 2026 | Registration closed

Springer CCIS
IIT Ropar
ICA 2026IIT ROPAR

ICA 2026 · Workshop 02

How GeoAI Understands the Earth: From Satellite Observations to Geospatial Intelligence

Dr. Shashank Tamaskar

Associate Professor, Plaksha University

GeoAIEarth ObservationGeospatial Foundation ModelsMultimodal Learning

About the Workshop

About the Workshop

Satellites, drones, and environmental sensors continuously generate vast amounts of data capturing changes in agricultural activities, vegetation health, water resources, land use, and climate. Transforming these observations into meaningful information and actionable intelligence remains a significant challenge.

This workshop introduces Geospatial Artificial Intelligence (GeoAI) and explores how advances in deep learning, self-supervised learning, and geospatial foundation models are transforming the interpretation of Earth observation data. Participants will gain an understanding of how AI can learn from diverse satellite and sensor observations to identify patterns, analyze environmental changes, and support informed decision-making.

Through intuitive explanations, real-world case studies, and a guided practical demonstration, the workshop will examine applications of GeoAI in agriculture, environmental monitoring, disaster assessment, and climate resilience. It will also introduce emerging developments in multimodal learning and geospatial reasoning, along with challenges related to trustworthy AI, geographic generalization, and data scarcity.

Workshop Outline

Workshop Outline

01

How Do We Observe and Understand the Earth?

Introduction to Earth observation technologies, including satellite and drone imagery, optical and radar sensing, multispectral observations, and spatial and temporal resolution.

02

From Earth Observation Data to Intelligent Maps

Fundamentals of machine learning and deep learning for geospatial applications, including land-cover classification, crop monitoring, vegetation assessment, and environmental change detection.

03

Geospatial Foundation Models – Teaching AI to Learn More Generally

Introduction to geospatial foundation models, self-supervised learning, transfer learning, multimodal learning, and spatial–temporal representations, with examples such as Prithvi-EO, SatMAE, and TerraMind.

04

Beyond Mapping – GeoAI for Real-World Intelligence

Applications of GeoAI in agricultural monitoring, drought and vegetation stress analysis, flood assessment, land-use change, and environmental decision support.

05

Hands-on Demonstration – From Satellite Images to AI Insights

A guided demonstration using publicly available Earth observation data and accessible computational tools to explore satellite imagery, visualize environmental indicators, apply AI-based analysis, and interpret results.

06

The Future of GeoAI – Toward Intelligent Earth Understanding

Emerging directions in multimodal geospatial foundation models, physics-informed learning, cross-region generalization, natural-language interaction with Earth observation data, and reasoning-enabled GeoAI systems.

Learning Outcomes

Learning Outcomes

By the end of the workshop, participants will be able to:

  • Understand how satellites, drones, and other sensing systems observe the Earth.
  • Explain how AI transforms geospatial observations into meaningful information and insights.
  • Distinguish between traditional machine learning, deep learning, and geospatial foundation models.
  • Understand the opportunities and limitations of integrating observations from multiple sensors and time periods.
  • Identify practical applications of GeoAI in agriculture, environmental monitoring, land-use analysis, and climate resilience.
  • Explore open datasets, pretrained models, and accessible tools for developing GeoAI applications.
  • Recognize emerging research directions in geospatial foundation models and intelligent Earth understanding.

Target Audience

Target Audience

This workshop is intended for undergraduate and postgraduate students, doctoral researchers, faculty members, AI practitioners, remote sensing professionals, and researchers working in agriculture, environmental science, civil engineering, and related disciplines.

Prerequisites: Basic familiarity with AI concepts is beneficial but not mandatory. No prior experience with geospatial foundation models is required.

Workshop Format and Requirements

Workshop Format and Requirements

The workshop will combine conceptual explanations, visual illustrations, real-world case studies, and a guided practical demonstration.

The practical component will use publicly available Earth observation data and accessible computational tools, with a prepared Google Colab notebook or equivalent browser-based environment to minimize installation requirements.

Internet access and a projector will be beneficial. A GPU is not required for the planned demonstration.