Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Evaluating the Level-of-Service of Protected Bike Lanes

627 views

Published on

Nick Foster, Senior Planner, Kittelson and Associates

Evaluating the Level-of-Service of Protected Bike Lanes

  1. 1. Evaluating the Level-of-Service of Protected Bike Lanes Nick Foster, AICP Senior Planner, Kittelson & Associates 11/21/14
  2. 2. Outline • Introduction • Previous Research • Project Methods • Results • Implications/Limitations Evaluating the Level-of-Service of Protected Bike Lanes 2
  3. 3. Objectives • LOS Model for Protected Bike Lanes – Segments Only – Readily Available Inputs • Comparison to Other Facilities Evaluating the Level-of-Service of Protected Bike Lanes
  4. 4. Protected Bike Lanes Evaluating the Level-of-Service of Protected Bike Lanes
  5. 5. Measuring User Perception • Quality of Service (QOS) – Level of Service (LOS) • Comfort/Stress/Safety • Typically ‘A-F’ scale Evaluating the Level-of-Service of Protected Bike Lanes
  6. 6. Motivation • Increased Interest in Non-Capacity Performance • No North American-based Model Evaluating the Level-of-Service of Protected Bike Lanes
  7. 7. Previous Research Evaluating the Level-of-Service of Protected Bike Lanes
  8. 8. Select Previous Efforts Method Form # of Participants # of Study Sites Considers Protected Bike Lanes? HCM 2010 BLOS OLS Regression 145 30* No Danish BLOS Logistic Regression 407 56* Yes FDOT BLOS OLS Regression 60-150 21-30* No FHWA BCI OLS Regression 202 78* No Level of Traffic Stress Index Not Based on Empirical Data Yes San Francisco BEQI Index Not Based on Empirical Data Partially Evaluating the Level-of-Service of Protected Bike Lanes *All Sites Not Shown at Each Viewing
  9. 9. Typical Factors Considered • Motor Vehicle Speeds/Volumes • Facility Type • Space Evaluating the Level-of-Service of Protected Bike Lanes
  10. 10. Methods Evaluating the Level-of-Service of Protected Bike Lanes
  11. 11. Video Collection Evaluating the Level-of-Service of Protected Bike Lanes Image Source: Tara Goddard, Portland State University
  12. 12. Site Selection • Protected Bike Lanes & Reference Videos • 20-30 Seconds Length • 23 clips • Criteria – Buffer & Facility Type – 1-way vs. 2-way – Traffic Volumes Evaluating the Level-of-Service of Protected Bike Lanes
  13. 13. Selected Clip Examples Evaluating the Level-of-Service of Protected Bike Lanes #1, 11, 17a – NE Multnomah St #2, 20a – Dearborn St #5 – SW Broadway St #8 – Fell St #19 – Cully Blvd #20b – Milwaukee Ave
  14. 14. Selected Clip Examples Evaluating the Level-of-Service of Protected Bike Lanes #3a, 7, 10 – SW Barbur Blvd #4 – NE Knott St #9 – Springwater Trail #13 – SW Barbur Blvd #3b – SE Ankeny St #17b – NE Multnomah St
  15. 15. Survey Administration • Online & In-Person Evaluating the Level-of-Service of Protected Bike Lanes
  16. 16. Video Clip Example • https://www.youtube.com/watch?v= F7gXQX54-HE Evaluating the Level-of-Service of Protected Bike Lanes
  17. 17. Results – In Person Survey Evaluating the Level-of-Service of Protected Bike Lanes
  18. 18. Demographics – Age & Gender • 53% Female Evaluating the Level-of-Service of Protected Bike Lanes Age (Years) # of Participants 20 30 40 50 60 70 80 90 0 20 40 60 80 100 n=219
  19. 19. Demographics – Riding Habits Evaluating the Level-of-Service of Protected Bike Lanes Commuting 6+/wk 3-5/wk 1-2x/wk 1-2x/mo Never Recreation Other Purposes # of Participants 0 20 40 60 80 100 120
  20. 20. Demographic Impacts Factor Correlation with Ratings Age 0.06* Gender (0=Male) 0.03 Riding Habits 0.10** Evaluating the Level-of-Service of Protected Bike Lanes *=Significant at the 95% confidence level **=Significant at the 99% confidence level
  21. 21. Score by Infrastructure Type1 Evaluating the Level-of-Service of Protected Bike Lanes Off-Street Path 1-way PBL 2-way PBL (w/ Traffic) Bike Lane w/ Parking 2-way PBL (against Traffic) Bike Boulevard Buffered Bike Lane Bike Lane Shared Street Better <-- Mean Score --> Worse A B C D n=220 n=221 n=514 n=146 n=513 n=148 n=441 n=1,984 n=221 1Some infrastructure types have only one location. Chart is shown for informational purposes and should not be considered an absolute preference rating hierarchy.
  22. 22. Score by Buffer Type Evaluating the Level-of-Service of Protected Bike Lanes Planters Parked Cars Raised/Parking Posts Mean Score A B C D n=589 n=514 n=1,615 n=220
  23. 23. Other Factors Factor Correlation MV Volume (Adjacent Lane) 0.06 MV Volume (Total in Video) 0.06 MV Volume (ADT) 0.09 MV Speed 0.03 Unsignalized Conflicts/Mile 0.03 # of Travel Lanes 0.18 Buffer Width -0.002 Evaluating the Level-of-Service of Protected Bike Lanes Significant at 99% confidence level – all protected bike lanes Significant at 99% confidence level – after controlling for one-way vs. two-way travel
  24. 24. Models • Index Table • Regression Models (OLS & Logistic) • Variables Considered – Buffer Type – 1-Way vs. 2-Way – MV Speed – # of Travel Lanes – MV Volume (ADT) Evaluating the Level-of-Service of Protected Bike Lanes
  25. 25. Regression Models Model 1 Model 2 Model 3 Buffer Type One-way vs. Two-Way MV Speed MV Volume (Adjacent Lane) # of MV Lanes MV Volume (ADT) Buffer Width Log Likelihood= -3,676 -3,671 -3,657 Evaluating the Level-of-Service of Protected Bike Lanes
  26. 26. Model Distribution Comparison Evaluating the Level-of-Service of Protected Bike Lanes # 1 # 11 # 12 # 14 # 15 # 16 # 17a # 18 # 19 # 2 # 20a # 20b # 5 # 6 # 8 clm1 clm2 clm3 obs clm1 clm2 clm3 obs clm1 clm2 clm3 obs clm1 clm2 clm3 obs clm1 clm2 clm3 obs 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 Distribution of Responses Model grade A B C D E F
  27. 27. Recommended Model Odds Ratios Evaluating the Level-of-Service of Protected Bike Lanes -200% 0% 200% 400% 600% 800% ADT (1,000 vehicles/ day) * MV Speed Two-Way Facility Raised/Parking Buffer Parked Car Buffer Planter Buffer Better <-- Change in Odds -->
  28. 28. Comparison to Intercept Surveys Evaluating the Level-of-Service of Protected Bike Lanes Barton Springs Dearborn Fell L Street Milwaukee (parked cars) MIlwaukee (posts) Oak clm2 Danish Obs clm2 Danish Obs clm2 Danish Obs clm2 Danish Obs clm2 Danish Obs clm2 Danish Obs clm2 Danish Obs 0.00 0.25 0.50 0.75 1.00 Distribution of Responses Model grade A B C D E F
  29. 29. Example Application Evaluating the Level-of-Service of Protected Bike Lanes
  30. 30. Existing Conditions • 11,000 ADT • HCM Link LOS ‘D’ Evaluating the Level-of-Service of Protected Bike Lanes
  31. 31. Build Conditions Evaluating the Level-of-Service of Protected Bike Lanes
  32. 32. Build Conditions LOS Calcs • Probability of an ‘A’ rating = 1/(1+e(-(-1.60)-1.38-0.001*(30*11,000/1,000))) = 0.53 • p(B) = 1/(1+e(-(0.05)-1.38- 0.001*(30*11,000/1,000))) - 0.53 = 0.32 Evaluating the Level-of-Service of Protected Bike Lanes
  33. 33. Build Conditions LOS Evaluating the Level-of-Service of Protected Bike Lanes 0% 20% 40% 60% 80% 100% Predicted Response Distribution A B C D E F
  34. 34. Index Table Evaluating the Level-of-Service of Protected Bike Lanes
  35. 35. Index Table Evaluating the Level-of-Service of Protected Bike Lanes Variable Comfort Score A B Buffer Type Planters Parked Cars Posts Motor Vehicle Speed (MPH) <=30 35 ADT (vehicles/day) <15,000 >=15,000 # of MV Travel Lanes 2 3
  36. 36. Index Table Performance Evaluating the Level-of-Service of Protected Bike Lanes Clip #1 Predicted Median Score Observed Median Score Difference 1 (1-way - P) A A None 2 (2-way - PC) A B Better 5 (1-way – PC) A A None 6 (2-way – PC) B B None 8 (1-way – PO) B B None 11 (1-way – P) A A None 12 (1-way – PO) A A None 15 (1-way – PC) B B None 16 (2-way – PC) B B None 17a (1-way – P) A A None 18 (1-way – PC) A A None 19 (1-way – R) B B None 20a (2-way – PC) A B Better 20b (1-way – PO) B B None 1Directionality and buffer type indicated in parentheses. P = Planters; PC = Parked Cars; PO = Posts; R = Raised/Parking (mostly unoccupied)
  37. 37. Results – Online Survey Evaluating the Level-of-Service of Protected Bike Lanes
  38. 38. Comparison to In-Person Survey • Older – Mean age 43 vs. 36 years • More Male – 65% vs. 47% • Bicycle More Often • Administration Method Effect – 0.28 points less comfortable Evaluating the Level-of-Service of Protected Bike Lanes
  39. 39. Conclusions Evaluating the Level-of-Service of Protected Bike Lanes
  40. 40. Recommended Model • Regression Model #2 – Readily Available Data • Model Valid Ranges – ADT: 9,000-30,000 vehicles/day – MV Speed: 25-35 MPH – Buffers: Planters, Parked Cars, Posts, Raised w/ Unoccupied Parking Evaluating the Level-of-Service of Protected Bike Lanes
  41. 41. Secondary Conclusions • Protected Bike Lanes > Other On- Street Infrastructure • Buffer Type Significant • One-way vs. Two-way Matters • MV Volumes Significant • Online Surveys Produce Different Results – Advertising Method Matters Evaluating the Level-of-Service of Protected Bike Lanes
  42. 42. Limitations • Variety of Protected Bike Lanes • Range of Traffic Conditions • No Intersections • Video Production Methods • Sample Demographics Evaluating the Level-of-Service of Protected Bike Lanes
  43. 43. Implications/Future Work • Model is Ready for Use – within identified ranges only • Future Work: – Intersection Research – Overall Method for All On-Street Infrastructure Evaluating the Level-of-Service of Protected Bike Lanes
  44. 44. •Thesis Committee –Chris Monsere (Chair) –Jennifer Dill –Kelly Clifton •Questions? –nfoster@kittelson.com •Thesis available at: –http://www.its.pdx.edu/publications.php Acknowledgments Evaluating the Level-of-Service of Protected Bike Lanes

×