Contents
EXECUTIVE SUMMARY 7
1.0. INTRODUCTION 10
2.0. BACKGROUND DISCUSSION 10
3.0. PRE INSPECTION DATA REVIEW 12
3.1. Data Selection 12
3.2. NHTSA Data Preparation 12
3.3. Data Preparation 17
4.0. VEHICLE INSPECTION AND DATA COLLECTION 18
4.1. Three Dimensional Laser Scanning 18
4.2. Pedal Force and Displacement Measurements 20
5.0. CONTROL LOCATION MEASUREMENTS 21
6.0. DATA ANALYSIS 23
6.1. LMR/HMR Groups 23
6.2. Correlation of Measured Variables 24
6.3. Correlation of Variables to Pedal Misapplication Rate 26
6.4. Regression Analysis 29
6.5. Force Versus Displacement Analysis 34
6.6. Analysis Including Driver Characteristics 38
7.0. SUMMARY AND OBSERVATIONS 41
Appendix A 43
Appendix B 47
Appendix C. List of test vehicles measurements 49
Table 1. Lowest Misapplication Rates (LMR) 14
Table 2. Highest Misapplication Rates (HMR) 15
Table 3. List of Special Interest Vehicles 16
Table 4. List of Dimensions Analyzed 21
Table 5. Pedal Dimensions Measured, Variables Shown in Bold Eliminated From Analyses 25
Table 6. Correlation of Measured Variables to Each Other 25
Table 7. Correlation of Measured Variables to Pedal Misapplication Rate 26
Table 8. Correlation of (Measured Variables) Squared to Pedal Misapplication Rate 27
Table 9. Correlation of Products of Measured Variables to Pedal Misapplication Rate 28
Table 10. Correlation of Ratios of Measured Variables to Pedal Misapplication Rate 28
Table 11. Regression Analysis: Pedal Misapplication Rate Versus Measured Dimensions 29
Table 12. Regression Analysis: Pedal Misapplication Rate Versus Measured Dimensions and Dimensions Squared 30
Table 13. Regression Analysis: Pedal Misapplication Rate Versus Measured Dimensions and Two-way Products of Variables 30
Table 14. Regression Analysis: Pedal Misapplication Rate Versus Measured Dimensions and Two-Way Ratios of Variables 30
Table 15. Regression Analysis: Pedal Misapplication Rate Versus Measured Dimensions, Dimensions Squared, Two-Way Interactions, and Ratios of Variables 31
Table 16. Stepwise Regression Model Including Driver Characteristics and Measured Vehicle Dimensions 39
Figure 1. Typical Scanning Session 19
Figure 2. Typical Screenshot of Computer Rendering 19
Figure 3. Pedal Force and Displacement Measuring Device 20
Figure 4. Typical View the 3D Rendering Used for Variables A, B, C, and D Showing the Accelerator Travel Path 22
Figure 5. ANOVA for Difference Between LMR, HMR and Special Interest Rates of Pedal Misapplication 24
Figure 6. Predicted Value Versus Measured Rate of Pedal Misapplication for Vehicles: L = LMR, S = Special Interest, H = HMR Groups 32
Figure 7. Distribution of Error Term From Regression 33
Figure 8. ANOVA of Predicted Value for LMR and HMR Groups 34
Figure 9. Force Versus Deflection Example for Vehicle With Similar Brake and Accelerator Responses 35
Figure 10. Force Versus Deflection Example for Vehicle With Different Brake and Accelerator Responses 36
Figure 11. 95 Percent Confidence Levels for Difference Between Brake-Force and Accelerator-Force Versus Deflection for LMR and HMR Groups, With 95 Percent Confidence Intervals 37
Figure 12. Predicted Rate of Pedal Misapplication Versus the Measured Rate, With 95 Percent Confidence Levels Shown 39
Figure 13. Effects of the Interaction of Driver Age and Stepover. Predicted Rate Increases Significantly When Stepover Is Very High for Older Drivers. Calculated From Stepwise Regression Model With Driver Characteristics and Measured Pedal Dimensions 40
Figure 14. Effects of the Interaction of Driver Height and Pedal Lateral Separation Shown With 95 Percent Confidence Levels (Dashed Lines). Large Pedal Separation Is Predicted to Be Associated\With Higher Rates of Pedal Misapplication for Taller... 41