AedesAhead

Inspiration

Dengue prevention has a timing problem.

Existing dashboards are excellent at showing where dengue activity is happening now, but public-health teams still face a harder question: where is risk likely to rise next?

In Singapore, dengue remains a persistent public-health challenge, with more than 13,600 reported cases in 2024. At the same time, inspection teams and community organisations have limited resources. Every inspection, outreach campaign and preventive intervention therefore needs to be prioritised carefully.

That inspired us to build AedesAhead — an AI-powered Dengue Command Centre designed to transform dengue and environmental data into an explainable 14-day early-warning forecast.

Instead of only asking:

"Where are the dengue clusters today?"

AedesAhead asks:

"Where should we act next?"

What it does

AedesAhead combines historical dengue cases, active dengue clusters, rainfall and temperature data to identify areas where dengue risk may increase over the next 14 days.

Our system is designed to provide three things at once:

1. Predictive risk mapping

Areas are classified into Low, Medium or High forecasted dengue risk so users can quickly identify locations requiring attention.

2. Explainable AI

A prediction alone is not enough for a public-health decision. AedesAhead uses SHAP explanations to show which factors contributed most strongly to a forecast, such as recent dengue activity, rainfall patterns, temperature or seasonal trends.

3. Actionable prioritisation

Rather than stopping at a prediction, AedesAhead converts forecasts into an emerging-risk priority queue, helping teams identify areas that may deserve earlier inspection or community outreach.

A particularly valuable scenario is an area with relatively few reported cases today but signals indicating elevated future risk. Detecting these areas could provide additional time for preventive action before case numbers rise.