Storm Chasers: The Impact AI Mission for Jamaica’s Recovery
When Hurricane Melissa struck Jamaica, it left a trail of destruction that stretched from Black River to Mandeville, Savanna-la-Mar to Lucea. Winds above 180 miles per hour tore through homes, farmlands, and hospitals. Roads were flooded, bridges collapsed, and thousands of families were displaced.
But while Melissa tested our physical resilience, it also revealed something deeper: how data and technology can guide recovery, rebuild trust, and create smarter systems for the future.
At Impact26 AI Lab, we believe that Jamaica’s recovery can be powered by intelligence, human and artificial.
The Mission: Building Solutions for Relief
Your task is simple but important:
Design an AI or data-driven solution that supports Jamaica’s and Haiti’s disaster relief and recovery efforts.
You’ll use the lessons from our latest session, “AI for Relief and Recovery: Technology, Data, and Resilience after Hurricane Melissa,” and apply them to one real-world problem that emerged from the storm.
Think of this as your chance to contribute directly to the country’s rebuilding, through science, empathy, and innovation.
Choose a Focus Area
Each participant or group will select one domain that they believe technology can strengthen:
Shelter & Housing – Predict where temporary shelters will face overcrowding or damage risk.
Health & Medical Response – Use IoT or AI to monitor conditions in emergency clinics.
Food & Water Distribution – Build models to map where supply chains are breaking down.
Infrastructure & Roads – Apply satellite or drone data to identify inaccessible areas.
Economic Recovery – Estimate community-level losses to help plan support.
Your Tools and Approach
Here’s what we’re encouraging you to use in your project:
1. Data Sources
Rainfall and wind data from the Meteorological Services
Satellite images (Sentinel, Landsat) for damage detection.
Demographic data for regions for Statistical Entities, such as STATIN.
Social media or survey data for citizen reports.
2. Technologies
Python for data cleaning, visualisation, and model building.
Geospatial tools like QGIS or Google Earth Engine for mapping.
Machine learning models such as Random Forest, LSTM, or K-Means for forecasting and clustering.
IoT prototypes (Arduino, Raspberry Pi) for live monitoring of rainfall, air quality, or medical metrics in shelters.
3. AI Techniques
Computer Vision: Flood detection using satellite image segmentation.
Natural Language Processing: Analysing WhatsApp or Twitter messages for distress signals.
Predictive Modelling: Estimating hospital demand or crop loss.
Optimization Models: Designing efficient routes for relief trucks.
Technical Guidance
You don’t need to solve everything—focus on clarity and impact.
We’re looking for ideas that show:
A clear problem definition (What data do you need and why?)
An applied model or tool (How does your idea work?)
Expected impact (How will it help real people in Jamaica?)
If you’re using code, visualisations, or device designs, include them in your submission. You may also include short video demos, notebooks, or mock dashboards.
Evaluation Criteria
Your submissions will be evaluated on:
Relevance How directly your idea helps in Jamaica’s relief and recovery.
Technical Quality Sound use of data science, AI models, or IoT technology.
Feasibility Can this idea be realistically implemented or tested?
Creativity Does it bring a fresh, insightful approach to a known challenge?
Impact Potential Could it genuinely help communities recover faster?
Submission Details
You may submit individually or as a group (maximum 5 members).
All submissions must include:
A 1–2 page concept paper or report (PDF format).
Optional Jupyter Notebook, diagram, or video demonstration.
Team member names
Deadline: November 19 2025.
Why your innovation matters
After every storm comes a choice, rebuild the same way, or rebuild smarter.
By participating in this mission, you’re not just showing what AI can do, but what Jamaica can lead.
The work you start here could become the foundation for real national resilience systems. A chatbot that connects families during floods. A model that predicts road washouts. A dashboard that helps the government prioritise shelters.
Every line of code and every data point matters.
