🌿 Advancing Sustainable Mobility: A Data Visualization Analysis of U.S. EPA Automotive Trends

                  __
                .----.
     __       __||  ||__
    |  |_____|  ||__||  |
    |__|_____|__||  ||__|
     (o)             (o)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Team Tres Sigmas  |  Course: 4CSE (Computer Science)  |  Subject: DataViz-E 1

---

👥 Team Members Inventado, Charles Fredric G. Rodelas, John Vincent B. Valles, James Vincent V.

📖 Project Overview

The project explores historical trends in the U.S. automotive industry (1975–2024) using data from the U.S. Environmental Protection Agency (EPA). By utilizing advanced data visualization techniques, we analyze the progression of fuel economy, CO2 emissions, and the adoption of cleaner vehicle technologies.

🎯 Objectives & SDG Alignment

The analysis aims to address specific policy questions aligned with the United Nations Sustainable Development Goals:

⚡ SDG 7 (Affordable and Clean Energy): Examining real-world fuel economy improvements and the shift toward alternative fuel vehicles.

🌍 SDG 13 (Climate Action): Analyzing carbon dioxide CO2 emission reductions and the impact of vehicle weight/horsepower trade-offs on environmental goals.

---

📂 Repository Structure

Note

The project follows a linear data pipeline structure. The comprehensive analysis and final conclusions can be found in the Final Report.

📦 Advancing-Sustainable-Mobility
 ┣ 📂 Datasets
 ┃ ┣ 📜 A-Detailed-Real-World-Fuel-Economy-Raw-Dataset.csv       # [SOURCE] Original Raw EPA Data
 ┃ ┣ 📜 Output - TresSigmas - DataCleaning.csv                   # [OUTPUT 1] Cleaned Data (0 Imputation)
 ┃ ┣ 📜 Output_TresSigmas_Rev2_CSV.csv                           # [OUTPUT 2] Cleaned Data (Revised)
 ┃ ┗ 📜 TresSigmas_DataClean_HuliNa.csv                          # [OUTPUT 3] Final Cleaned Data (Avg Imputation)
 ┃
 ┣ 📂 Visualizations
 ┃ ┣ 📜 TresSigmas_Proposal_Visualization_v1.ipynb               # Python Analysis (Baseline Code)
 ┃ ┗ 📜 TresSigmas_Proposal_Visualization_v2.ipynb               # Python Analysis (Refined Code)
 ┃
 ┣ 📜 DataCleaning - TresSigmas.tfl                              # [PIPELINE] Tableau Prep Builder Flow
 ┣ 📜 TresSigmas_Final Course Project_FinalReport.pdf            # [FINAL] Full Project Report (Methodology, Visuals, Conclusion)
 ┗ 📜 TresSigmas_4CSE_DataViz-E 1_CourseProjectProposal_Final.pdf # Project Proposal & Literature Review

---

📊 Methodology & Execution Workflow

Our methodology consists of two distinct phases: Data Cleaning and Imputation Strategy. The setup instructions below mirror this workflow.

🧹 Phase 1: Data Cleaning (ETL)

• Tool Used: Tableau Prep Builder (DataCleaning - TresSigmas.tfl)
• Process:
  1. Ingestion of A-Detailed-Real-World-Fuel-Economy-Raw-Dataset.csv.
  2. Standardization of string formats (e.g., manufacturer names).
  3. Handling of null values (see Phase 2).

📈 Phase 2: Imputation & Visualization Strategy

We utilized two different analytical approaches to handle missing performance data (e.g., missing specific metrics for older car models).

Version	Methodology	Dataset Used	Purpose
V1 (Baseline)	Zero (0) Imputation: Missing numerical values replaced with 0.	Output - TresSigmas - DataCleaning.csv	Initial Exploratory Data Analysis (EDA) and distribution checks.
V2 (Refined)	Average Imputation: Missing values replaced with column/category averages.	TresSigmas_DataClean_HuliNa.csv	Statistical analysis minimizing skew; used for Heatmaps, Treemaps, and Dual Axis charts.

---

🚀 Key Findings (From Final Report)

Based on the visualizations generated in the TresSigmas_Final Course Project_FinalReport.pdf, the study concluded:

🎭 The "Masking Effect": While internal combustion engine technology has improved, gains are largely negated by a consistent increase in vehicle weight and horsepower since 1975.

🚘 Segment Disparity: Sedans/Wagons remain the most efficient vehicle types, while Truck SUVs and Pickups consistently rank lowest in fuel economy.

🏭 Production vs. Pollution: A Treemap analysis revealed that manufacturers with the highest production volumes (e.g., GM, Stellantis) also exhibit high CO2 emission intensities due to fleet compositions dominated by heavier vehicles.

⚖️ SDG Implications: The market shift toward heavier vehicles creates a critical barrier to achieving SDG 7 (Energy Efficiency) and SDG 13 (Climate Action).

---

🛠️ Setup Instructions

Follow these steps to reproduce the analysis in the correct methodological order.

Prerequisites

• Tableau Prep Builder (for the ETL pipeline)
• Tableau Desktop (for Dashboards)
• Python 3.x with Jupyter Notebook (Optional, for .ipynb analysis)
  • Libraries: pandas, numpy, matplotlib, seaborn, plotly

---

Step-by-Step Execution

1. Run the Data Cleaning Pipeline

To see how the raw data was processed:

1. Open Tableau Prep Builder.
2. Load the file DataCleaning - TresSigmas.tflx using Tableau Prep Builder.
3. Note: If the file shows a connection error, right-click the "Input" node and locate Datasets/A-Detailed-Real-World-Fuel-Economy-Raw-Dataset.csv on your local machine.
4. Run the flow to generate the output CSVs (optional, as they are already provided in the Datasets folder).

2. Launch the Visualizations

To view the final dashboards:

1. Navigate to the Visualizations folder.
2. Open TresSigmas_Proposal_Visualization_v1.ipynb or TresSigmas_Proposal_Visualization_v2.ipynb in Jupyter Notebook or Google Colab.
3. Run all cells to generate the correlation heatmaps and statistical summaries.