Back to Experience

Food Desert Detection System

IBM Case Competition

Participant · Fall 2024 · DFW Area

Watson AIPythonData Visualization

Overview

Developed an AI-powered solution to identify and address food deserts in the Dallas-Fort Worth metropolitan area. The project combined machine learning, geospatial analysis, and data visualization to provide actionable insights for improving food accessibility in underserved communities.

Challenge

Food deserts are areas with limited access to affordable and nutritious food, particularly affecting low-income communities. Traditional methods of identifying these areas were time-consuming and often inaccurate, making it difficult for local governments and organizations to effectively allocate resources and implement solutions.

Solution

Leveraged IBM Watson AI to analyze geospatial data, demographic information, and food access metrics. Created a predictive model that could identify food deserts with high accuracy and provide recommendations for optimal placement of food resources and community programs.

Key Features

AI-powered food desert identification

Interactive data visualization dashboard

Predictive modeling for resource allocation

Geospatial analysis and mapping

Community impact assessment tools

Government reporting and recommendations

Results & Impact

1

Increased food desert detection accuracy by 25% compared to traditional methods

2

Successfully presented data-driven strategies to IBM judges and local government officials

3

Provided actionable recommendations for improving food access in 15+ DFW communities

4

Integrated multiple API data streams for comprehensive model training

5

Ensured high data quality through automated validation processes

Technology Stack

Frontend

PythonJupyter NotebooksMatplotlibPlotly

Backend

IBM Watson AIPythonPandasNumPy

Database

CSVJSONGeospatial Data

Tools

IBM CloudWatson StudioGitTableau