🌟 Welcome to AS26

AI-Powered Earth Observation:
From Idea to Pitch

Today we form teams, choose projects, and start building. By the end of this session you will have a team, a concept, and a plan. The doing starts now.

📅 Monday, June 8, 2026
🕒 14:00 - 16:00
📍 ISU, Strasbourg
🎯 Ideation Workshop
🎯

Today's Goals: Doing, Not Lecturing

This course is about building real things. Today we skip the long lectures and get straight to work. By 16:00, every student will leave with:

👥
A Team (3-4 people)
Diverse skills: coding, design, domain expertise, presentation. Every team needs a mix.
💡
A Project Concept
Pick one of our suggested tracks or bring your own idea. The key: it must combine satellite data with AI.
📄
A One-Page Concept Sketch
Problem, solution, data sources, target user. Napkin-quality is fine. Clarity of thought matters more than polish.
🎧
A Factory 2026 Pitch Plan
On Thursday you pitch to real investors and industry professionals. Today we start preparing for that moment.
🔥
The clock is ticking. You have 3 days to go from idea to pitch-ready prototype. Every hour counts. We start NOW.
🌎

What is Advanced Studies 26?

AS26 is a 2-week intensive course where you build AI-powered web applications that use Earth Observation (satellite) data to solve real problems. This is not a traditional class. It is a startup accelerator for space technology.

What You Will Build

  • A working web dashboard that uses satellite data (Sentinel, Landsat, NDVI)
  • An AI layer (Google Gemini) that interprets the data and provides insights
  • An interactive map (Leaflet.js) that visualizes results
  • A pitch presentation for the Factory 2026 conference

The Timeline

Mon-Wed Foundations + ideation + first prototype
Thursday Factory 2026: Pitch to investors!
Week 2 Hackathon sprints, polish, final presentations
🏆

Factory 2026: Your Stage

On Thursday, June 11, the ISU campus hosts Factory 2026: a full-day conference with keynotes, startup showcases, and investor panels. You will pitch your project on the main stage.

🎤
3-Minute Pitch
Problem, solution, live demo, and your ask. Practiced and polished. This is your moment.
💼
Real Audience
Industry professionals, investors, ISU faculty, startup founders. Not a classroom exercise.
🌏
Exhibition Table
Every team gets a table in the Startup Village to demo their app and network.
🏆
Laureate Awards
The strongest pitches are selected for the main stage at the Amphitheatre Galaxy.
🚀
This is not hypothetical. Everything you build this week will be presented to a real audience on Thursday. That is 3 days from now. Let's make them count.
🚀

Project Tracks: Your Launchpad

We have prepared three suggested tracks as starting points. Each one solves a real problem using satellite data + AI. You can also bring your own idea. The only requirement: it must combine Earth Observation data with artificial intelligence.

🛰
Track 1: Parametric Insurance
Automated crop insurance payouts driven by satellite data. No adjustors needed.
🍇
Track 2: Alsace CubeSat
AI dashboard for a student-built CubeSat monitoring Alsace's vineyards and environment.
🏭
Track 3: EU Green Deal ESG
Compliance monitoring dashboard for the EU Green New Deal using satellite emissions data.
💡
Track 4: Your Own Idea
Bring your passion. Water resources, disaster response, urban heat islands, biodiversity: the sky is the limit.
🛰

Track 1: Parametric Satellite Insurance

A lean, automated insurance platform for farmers. No human adjustors required: if satellite data (NDVI, soil moisture) detects crop damage, the payout triggers automatically.

The Problem

Traditional crop insurance is slow, expensive, and requires physical inspections. Many smallholder farmers cannot afford it or cannot access it in time.

Your Solution

  • Define NDVI thresholds that indicate crop stress
  • Monitor satellite imagery over insured farmland
  • Use AI to analyze trends and detect anomalies
  • Trigger automated payouts when thresholds are breached

Key Data Sources

📡
Sentinel-2 NDVI
Vegetation health every 5 days, free from Copernicus
💧
SMAP Soil Moisture
NASA soil moisture data for drought detection
🌡
ERA5 Climate
Historical temperature and precipitation records
🍇

Track 2: The Alsace CubeSat Dashboard

Design an AI dashboard for a student-built CubeSat that monitors the Alsace region. This project imagines a small satellite launched and operated by ISU students, providing specialized data for local use cases.

The Vision

A CubeSat designed in Alsace, built by Alsace students, for the benefit of Alsace. Your dashboard simulates the data this satellite would produce and demonstrates its value.

Potential Use Cases

  • Vineyard health monitoring (NDVI time-series)
  • Air quality tracking over Strasbourg
  • Forest fire risk assessment in the Vosges mountains
  • Urban sprawl detection along the Rhine corridor

Inspiration

🛰
Explore the CubeSat concept brochure: sounny.github.io/cubesat/brochure-alsace.html

Your AI dashboard would serve as the ground station interface: receiving simulated telemetry, processing imagery with AI, and presenting actionable insights to local stakeholders.

🏭

Track 3: EU Green New Deal Compliance

Build an ESG (Environmental, Social, Governance) monitoring dashboard that helps industrial facilities track their compliance with the EU Green New Deal using satellite data.

The Opportunity

The EU Green Deal requires companies to report environmental metrics. Satellite data can independently verify emissions, land use, and environmental impact at scale.

Your Dashboard Could

  • Monitor NO2 and CO2 emissions near industrial sites
  • Track land use changes and green space preservation
  • Generate AI-powered compliance reports
  • Compare facilities against regulatory benchmarks

Key Data Sources

🏭
Sentinel-5P
Atmospheric pollution (NO2, CH4, CO) at 7km resolution
🗺
Corine Land Cover
European land use classification, updated regularly
📈
EU ETS Data
Emissions Trading System records for verification
💡

Track 4: Bring Your Own Idea

Do you have a passion project? A problem you care about? Bring it. The only constraint is that your project must combine satellite/Earth Observation data with AI to solve a real problem.

Ideas to Spark Your Thinking

🌊
Flood Risk
Real-time flood prediction using SAR imagery and weather forecasts
🌳
Deforestation
Track illegal logging using Sentinel-2 change detection
🏙
Urban Heat
Map heat islands and suggest cooling interventions
🚢
Marine Pollution
Detect oil spills and plastic accumulation from space
🌾
Food Security
Predict crop yields from NDVI trends and climate data
💨
Air Quality
Correlate satellite pollution data with health outcomes
💫
Pro tip: The best projects come from personal experience. What problem have you seen in your home country that satellite data could help solve?
🌎

Example: TerraX (Built by ISU Students)

TerraX is a working example of what you will build. It helps smallholder farmers in Sub-Saharan Africa choose the right crops using AI + satellite climate data. See it live.

🧠
AI Layer
Google Gemini: receives climate data, generates crop recommendations as JSON
📡
Data Layer
Open-Meteo API, NASA POWER: temperature, rainfall, solar radiation for any lat/lng
🗺
Map Layer
Leaflet.js with OpenStreetMap: interactive map, click events, marker popups
🖥
UI Layer
HTML5, CSS Grid, chat sidebar: responsive split-panel layout
💡
Your project will follow this same architecture. A map, a data source, an AI layer, and a user interface. The difference is what problem you solve and what data you use.
👥

Team Formation Workshop (15 min)

Stand up. Find your teammates. You have 15 minutes to form teams of 3-4 people. Here are the rules:

👥
3-4 People Per Team
No solo projects. No teams larger than 5. You need diverse skills to succeed.
🎨
Mix of Skills
Every team needs: someone who can code, someone who can design, someone who can present. Ideally all three.
🎯
Pick a Track (or Propose)
Agree on which track to pursue. You can pivot later, but start with a direction.
💬
Talk to Strangers
The best teams form across nationalities and disciplines. This is ISU: use the diversity.
⏰
Timer starts now. Stand up, introduce yourself, find teammates who share your interest or complement your skills. Report back with your team name, members, and chosen track.
📝

Concept Sketch: Your First Deliverable (20 min)

With your team, create a one-page concept sketch. Use paper, a whiteboard, or a shared doc. This is your first deliverable. Keep it rough: clarity matters more than polish.

What to Include

1. The Problem
Who has this problem? Why does it matter? One paragraph.
2. Your Solution
What does your dashboard do? How does it use satellite data + AI?
3. Data Sources
Which satellites/APIs will you use? (Sentinel-2, ERA5, NDVI, etc.)
4. Target User
Who will use this? A farmer? A city planner? An insurance company? A student?
📝
Quick sketch, not a thesis. Draw a rough wireframe of what the dashboard looks like. Label the data flow. This will evolve every day. Start messy, iterate fast.
✅

Wrap-Up: What Happens Next

Great work today. You now have a team, a track, and a concept. Here is what comes next:

Tomorrow AM AI Fundamentals: learn how LLMs work and how to call AI APIs from JavaScript
Tomorrow PM Dev Setup + First Build: VS Code, Git, API keys, and your first working AI chat interface
Wednesday PM Internal Pitch: present your concept and mockup to the faculty for feedback
Thursday Factory 2026: Pitch on the main stage to real investors and industry leaders

Before You Leave Today

  • Submit your team registration: Team name, member names, chosen track
  • Submit your concept sketch: Photo of your whiteboard sketch or a shared Google Doc
  • Install VS Code on your laptop (you will need it tomorrow): code.visualstudio.com
  • Create a GitHub account if you do not have one: github.com
🚀
You are already further along than you think. You have a team, a concept, and a deadline. That is more than most startups have on Day 1. See you tomorrow.