dynamic fields ai driven vegetation monitoring || ActInSpace Challenges

Monitoring Earth's vegetation is critical for food security and ecosystem health. While open-source data from satellites like Sentinel (radar/optical) is abundant, it is complex to translate into simple, actionable products. Using this data, both by themselves and combining in a multi-modal fashion, can lead to reliable and valuable results.

FROM SPACE

Design a holistic, end-to-end service for vegetation monitoring that is both technically novel and commercially viable. This means designing a smart pipeline (using radar, optical, and other data, either by themselves or combining them) and a robust business case for a real-world application, such as in agriculture or forestry.

TO THE CHALLENGE

A complete solution must use geospatial data for a viable product (e.g. using Sentinel-1, Sentinel-2, but also potentially thermal data or LiDAR). This data could be fed into an AI model (that can vary from machine learning to complex deep learning models) to create reliable, uninterrupted insights, like yield forecasts, pest alerts, or biomass estimates.
This advanced pipeline enables the commercial strategy: a SaaS or API business model selling actionable answers, not just data. This could target. Some starting point can be:
Precision Agriculture: Per-hectare subscriptions for farmers (e.g., viticulture, broad-acre crops).
Forestry: Monitoring for deforestation, fire risk, or logging compliance.
Any other idea from an application point of view, technical solution or data is more than welcomed. Participants are warmly encouraged to consider solutions that can even solve more than one task at once! Thinking outside the box is an added value!