Mapping of Agroforestry Systems based on Artificial Intelligence

Challenge Background

The development of the use of Information Technologies (IT) and artificial intelligence to be implemented in a new proposal in the implementation of restoration mechanisms, results in one of the pedagogical opportunities to take into account IT in higher education, mainly seeking a structural and cognitive change, through technological innovation and teaching/learning processes, that allow flexibility and promote the training of human talent that can optimise and detect the needs and parameters to execute forest restoration mechanisms in the Colombian tropics.

This project aims to create an algorithm that allows automated mapping of the location of agroforestry settlements through the analysis of satellite images and thus facilitate its implementation through the use of scripts that allow training a detection model and apply it to new images in a few steps.

The Problem

Agroforestry Systems (AS) are widely known for their contribution to the generation of environmental goods and services, such as small plots to hydrographic basins. From an ecological point of view, the use of trees can contribute to improving the productivity and sustainability of existing systems through an increase in the yield of pastures and associated crops, as well as the increase in animals that eat fruits or foliage and from the sale of wood or carbon stored by trees. From the economic point of view, the Agroforestry System can favour the increase and diversification of production.

However, the contribution of this land use in the structure and composition of agricultural landscapes on a national scale has been little valued, since it is known that 4 million hectares are suitable for agroforestry land use, but only 216 thousand hectares are occupied. for this land use, which denotes a low adoption of these systems, which leads to a need for innovation with emerging technologies, as well as the need to carry out large-scale studies that allow identifying and predicting the distribution of agroforestry promoting this use. sustainable development of the soil at both local and regional planning scales.

Goal of the Project

• Automated mapping of the location of Agroforestry Systems.
• Generation of an analysis of satellite images to facilitate the implementation of Agroforestry Systems.
• Develop algorithms with an emphasis on their ease of implementation.
• Lower access barriers to artificial intelligence (AI) tools in the forestry field.
• Allow government agencies, researchers, and other stakeholders to apply it to their own use cases.

Project Timeline

What you'll learn

1. Collection of Data.
2. Data Cleaning.
3. Data Analysis.
4. Data Visualization.
5. Machine Learning.
6. Algorithmic Trading.
7. Docker.
8. Geographic Information Systems.
9. Python.