LandDrop

Inspiration

According to the United Nations, 40% of all the land on Earth is considered degraded. This number could rise to 90% by 2050 if no intervention is made. To tackle this problem, the United Nations seeks to analyze relationships between land degradation, drought, and human populations. That’s why we created LandDrop—an analytical tool to visualize and correlate these relationships, providing insights that can guide impactful, world-changing solutions.

What It Does

LandDrop allows users to analyze soil data from any location on Earth. With a single click, users can select a location and receive a summary of both soil and weather data. We also provide graphical correlations, including land degradation and drought risks.

Additionally, users can ask questions directly to our AI agents, who interpret and explain the data in more detail. For example, users can ask questions like, “Is it possible to plant avocados here?” and our agents will provide an answer based on soil and environmental data, along with an explanation of the underlying analysis.

How We Built It

Our team utilized Fetch.ai's uAgents, Microsoft Azure's LLM, Next.js, and FastAPI.

LandDrop leverages Fetch.ai’s uAgent framework and GPT-4 as our language model, hosted on Microsoft Azure. Here’s how it works:

When a user selects a location, the Environmentalist Agent retrieves external datasets with details on land degradation and drought indicators such as precipitation and soil moisture.
The Environmentalist Agent then communicates with the Socioeconomist Agent, prompting it to gather social and economic data relevant to the region, including migration patterns and economic diversity.
Finally, the Predictor Agent analyzes historical data and generates forecasts for future trends.

All of this data is then integrated into visual graphs and summary analyses for the user. This agent-based architecture enables efficient retrieval and correlation of both environmental and socio-economic data for each unique location.

Challenges We Ran Into

One of the biggest challenges we faced was managing the sheer volume and variety of data needed to provide meaningful insights. Our project requires a wide range of data—environmental, social, and economic—which made sourcing, parsing, and accurately analyzing the information complex. Each dataset has its own structure and formatting, so integrating them into a cohesive framework required extensive work in data handling and processing.

Accomplishments That We're Proud Of

Handled Complex Data: Successfully managed and combined a wide range of data for seamless analysis.
Interactive Map Feature: Created a user-friendly map interface for data exploration.
Fetch.ai Multi-Agent Integration: Implemented an agent-based system to retrieve and analyze data efficiently.

What We Learned

Data Parsing and Integration: Developed skills in managing and standardizing diverse datasets.
Agent-Based Architecture with Fetch.ai: Gained hands-on experience with multi-agent workflows.
Scalability and Global Data Challenges: Tackled issues related to working with global data sources.
Real-World Applications of Predictive Modeling: Built predictive models to forecast trends based on historical data.

What's Next for LandDrop

With limited time, we couldn’t include all the features we envisioned. In the future, we aim to:

Expand Data and Forecasting Range: Extend both the historical data range and the prediction horizon to allow for longer-term forecasts.
Introduce a Simulation Tool: Enable users to adjust variables (e.g., precipitation levels) and observe impacts on related factors like soil moisture, temperature, and vegetation health. This feature would allow experts to assess hypothetical scenarios, offering a powerful way to understand environmental dynamics.

With LandDrop, we aim to make complex environmental and socioeconomic data accessible and actionable, supporting better decision-making in the face of global challenges.