🔍 Flagged Vehicles Pattern Analysis

📅 Date: December 24, 2024
📊 Dataset: OBFCM 2021-2023 PHEV data (995,511 records)
🎯 Analysis Focus: Identifying patterns, correlations, and interesting connections in flagged vehicles

📑 Table of Contents

1. Executive Summary
2. Detailed Findings by Filter Step
   • Step 1: CS Invalid
   • Step 2: Missing RW_EC
   • Step 3: Missing OEM/Model
   • Step 4: RW_EC Zero
   • Step 5: VFN Issue
   • Step 6: Physics CO2/FC
   • Step 7: Mileage/FC Inconsistency
   • Step 8: EDS/Energy Violation
3. Cross-Step Patterns
4. References

📋 Executive Summary

Analysis of flagged vehicles reveals several significant patterns:

Key Insights

1. Manufacturer Patterns - Stellantis group shows consistent data quality issues across multiple steps
2. Vehicle Characteristics - Flagged vehicles show systematic differences in mass, engine power, and electric range
3. Model-Specific Issues - Certain models (Porsche Panamera, Hyundai Tucson, Mitsubishi Eclipse Cross) show extreme overrepresentation
4. Data Quality Concerns - Patterns suggest both genuine vehicle characteristics and potential data reporting issues

📊 Detailed Findings by Filter Step

🔴 Step 1: CS Invalid (33,348 vehicles flagged)

What it means: Vehicles with invalid Charge-Sustaining (CS) mode data

Top Manufacturers

Top Models (by Overrepresentation)

Vehicle Characteristics Comparison

💡 Key Insight

Vehicles flagged in Step 1 tend to be lighter and have smaller engines, but paradoxically show higher CO₂ emissions and longer electric range. This suggests potential issues with: - Charge-sustaining mode operation - Data reporting in these specific models - Possible calibration or sensor issues

🌐 Online Research Context

• Jeep/Chrysler: Research shows these brands have high complaint rates (1,488 complaints per car for Jeep) and recall rates (1.06 for Jeep/Chrysler), suggesting broader quality control issues that may extend to data reporting.
  • Source: Carscoops - Safety Ratings Study
• Mitsubishi Eclipse Cross PHEV: Limited specific information found, but the extreme overrepresentation suggests model-specific issues with CS mode data collection or reporting.

🟠 Step 2: Missing RW_EC (12,554 vehicles flagged)

What it means: Vehicles missing Real-World Electric Consumption (RW_EC) data

Top Manufacturers

Top Models (by Overrepresentation)

Vehicle Characteristics Comparison

💡 Key Insight

Missing RW_EC is strongly associated with Hyundai models (Tucson, Santa Fe) and Ford vehicles. These vehicles tend to be: - Lighter than average - Have lower emissions - Have lower total fuel consumption - May be newer models or have different monitoring systems

🌐 Online Research Context

• Hyundai/Kia: Recent recalls affecting millions of vehicles, though not specifically related to PHEV data reporting. The extreme overrepresentation of Hyundai models suggests potential issues with their OBFCM implementation or data transmission.
  • Source: AP News - Hyundai/Kia Recalls

🟡 Step 3: Missing OEM/Model (5,753 vehicles flagged)

What it means: Vehicles with missing manufacturer or model information

Top Manufacturers

💡 Key Insight

Step 3 flags vehicles with missing OEM or Model information. This is a data quality issue that affects vehicle identification and analysis. Volvo vehicles show the highest number of missing information cases, followed by Peugeot.

🔵 Step 4: RW_EC Zero (18,779 vehicles flagged)

What it means: Vehicles reporting zero Real-World Electric Consumption (indicating no electric mode usage)

Top Manufacturers (by Overrepresentation)

Top Models (by Overrepresentation)

Vehicle Characteristics Comparison

💡 Key Insight

This is the most striking pattern - Porsche luxury PHEVs (Panamera, Cayenne) are reporting zero electric consumption. These are: - High-performance vehicles - Heavy vehicles (2,350 kg average) - Large engines (2,513 cc average) - High power (213 kW average)

Possible explanations for zero RW_EC: 1. 🔋 Battery issues preventing electric mode operation 2. 🚗 Driver behavior (not charging vehicles) 3. 📊 Data reporting problems specific to Porsche’s OBFCM implementation 4. ⚙️ Design issues where electric mode is rarely engaged

🟣 Step 5: VFN Issue (32,363 vehicles flagged)

What it means: Vehicle Family Name (VFN) validation failures

Top Manufacturers (by Overrepresentation)

Top Models (by Overrepres

🟢 Step 6: Physics CO2/FC (420 vehicles flagged)

What it means: Vehicles with physically implausible CO₂ or fuel consumption values

Criteria: FCgap_perc > 1800 OR TA_CO2 >= 190 OR RW_CO2 > 800

Top Manufacturers

💡 Key Insight

Step 6 flags vehicles with physically impossible values, suggesting data reporting errors or sensor malfunctions. Mercedes-Benz vehicles, particularly GLC models, show the highest incidence of physics violations.

🟠 Step 7: Mileage/FC Inconsistency (565 vehicles flagged)

What it means: Vehicles with inconsistent mileage and fuel consumption relationships

Criteria: Logical impossibilities such as fuel consumption without distance, or large distance with zero fuel

Breakdown by Inconsistency Type

💡 Key Insight

Step 7 identifies logical impossibilities in the data, such as reporting fuel consumption without corresponding distance traveled. These inconsistencies suggest data collection or reporting errors.

🔵 Step 8: EDS/Energy Violation (3,506 vehicles flagged)

What it means: Vehicles with EDS or energy values outside acceptable ranges

Criteria: EDS outside 0-100%, negative energy values, or energy accounting inconsistencies

Breakdown by Violation Type

Top Manufacturers

💡 Key Insight

Step 8 flags vehicles with EDS values outside the physically possible range of 0-100%. Most violations occur in vehicles with missing OEM information, suggesting data quality issues in identification and reporting.

💡 Key Insight

VFN issues are primarily with: - Volvo/Polestar models (owned by Geely) - Other Geely-owned brands

This suggests: - VFN whitelist may need updating - Naming inconsistencies in Geely’s reporting - Corporate structure changes affecting data standardization

🌐 Online Research Context

• Geely-Volvo Relationship: Geely acquired Volvo Cars in 2010, explaining VFN naming inconsistencies between different reporting systems.

🔗 Cross-Step Patterns

🏢 Stellantis Group (Fiat, Opel, Peugeot, Chrysler, Jeep)

Consistent Issues Across Multiple Steps:

Analysis

The consistent data quality issues across Stellantis brands suggest:

1. Common OBFCM implementation across brands
2. Shared data reporting infrastructure
3. Potential systemic issues in their PHEV monitoring systems

🌐 Online Research Context

• Stellantis Formation: Stellantis was formed in 2021 from the merger of FCA (Fiat Chrysler Automobiles) and PSA (Peugeot Société Anonyme)
• The merger may have created integration challenges in data reporting systems

📚 References

Online Research Sources

1. Jeep/Chrysler Quality Issues:
   • Carscoops: Volvo, Subaru, Tesla Lead Study with Most 5-Star Safety Ratings
   • Jeep: 1,488 complaints per car, recall rate of 1.06
2. Hyundai/Kia Recalls:
   • AP News: Millions of recalled Hyundai and Kia vehicles with dangerous defect remain on road
3. Automotive Recall Statistics:
   • AutoInsurance.com: Car Recall Facts & Statistics
   • Ford: 94 recalls affecting 5.6M vehicles (2024-2025)
   • Tesla: 20 recalls affecting 5.8M vehicles
4. Stellantis Formation:
   • Stellantis formed in 2021 from merger of FCA (Fiat Chrysler Automobiles) and PSA (Peugeot Société Anonyme)
5. Geely-Volvo Relationship:
   • Geely acquired Volvo Cars in 2010, explaining VFN naming inconsistencies

Data Sources

• OBFCM dataset analysis (2021-2023)
• Pattern analysis using statistical tests (t-tests, overrepresentation analysis)
• Internal flagging statistics from cleaning pipeline

📌 Important Notes

⚠️ Disclaimer: This analysis is based on OBFCM data from 2021-2023. Patterns may reflect both genuine vehicle characteristics and data quality issues. Further investigation with manufacturers is recommended for flagged models.

Recommendations

1. Contact manufacturers for flagged models to verify data reporting accuracy
2. Update VFN whitelist to include Geely/Volvo naming variations
3. Investigate Porsche OBFCM implementation for zero RW_EC reporting
4. Review Stellantis data pipeline for systemic issues
5. Monitor Hyundai models for missing RW_EC data in future datasets

Document Version: 1.0
Last Updated: December 24, 2024
Author: Data Analysis Team