Evaluating Driver and Pedestrian Behaviors at Enhanced Multi-lane Midblock Pedestrian Crossings

Nick Foster, AICP

A Case Study in Portland, OR

Introduction

• Problem

– Over 4,000 pedestrian deaths in 2010

• Potential solution

– Enhanced crossings

• Effectiveness?

• Use?

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings 2

Project Overview

• Two marked midblock crossings

– Rectangular rapid flash beacons (RRFB)

– Raised median refuge islands

– Z-crossing (Danish offset)

• One site only

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings 3

Background

• RRFB research focused on drivers

– Driver yielding rates: 54%-88%

• Crossing decisions based on distance to crosswalk

– Limited research on attraction

• No literature on Z crossing use

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings 4

Site 1 – Barbur Boulevard

• 30,700 ADT

• 35 MPH

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings 5

Site 2 – B-H Highway

• 26,400 ADT

• 40 MPH

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings 6

Methodology

• 62 hours of video

– Weekdays in February 2013

• Driver and pedestrian behavior MOEs

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings 7

RESULTS

Presentation Title

Driver Yielding Rates

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings

0% 20% 40% 60% 80% 100%

Overall

Stage 2

Stage 1

Overall

Stage 2

Stage 1

B-H

Hig

hw

ayB

arb

ur

Bo

ule

vard

Driver Yielding Rate

Loca

tio

n

RRFB Not Activated Crossings RRFB Activated Crossings

n=240

n=16

n=20

n=135

n=13

n=0

n=16

n=468

n=33 n=297

n=228

n=162

9

Comparison to Other Studies

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings

0% 20% 40% 60% 80% 100%

This study

Shurbutt, et al. (4 Beacons)

Shurbutt, et al. (2 Beacons)

ODOT (Bend)

ODOT

Western Michigan

Hunter, et al.

Average Driver Yielding Rate

Stu

dy

10

Pedestrian Actuation Rates

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings

0% 20% 40% 60% 80% 100%

Overall

Cars Present

No Cars Present

Overall

Cars Present

No Cars Present

Bar

bu

r B

ou

leva

rdB

-H H

igh

way

RRFB Actuation Rate

Loca

tio

n

11

Crossing Locations

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings

Bus Stops

SW 6

2n

d A

ve

B-H Highway

15% 70%

15%

12

Diverted Crossings

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings

Bus Stops

SW 6

2n

d A

ve

B-H Highway

• 52% of crossings at crosswalk are out-of-direction

13

Diverted Crossings – SB Only

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings

Bus Stops

SW 6

2n

d A

ve

B-H Highway

56%

44%

14

Z-Crossing Use

• Path use = 52%

– High yielding rates

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings 15

Other Findings

• Avoidance maneuvers

– Hard braking (2)

• Stranded pedestrians

– RRFB activated (1 – 0.3%)

– RRFB not-activated (6 – 15%)

• Minimal pedestrian delay

– 20 sec max (RRFB not activated)

– All but one <15 sec (RRFB activated)

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings 16

Conclusions

• 91-92% overall driver yielding rate

• Marked midblock crossing with RRFB may encourage diversion

• Z-crossing effectiveness limited

– Adequate sight distance

– No physical barrier

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings 17

Future Research

• Pedestrian diversion

– More sites

– Before/after

– Wider field of view

– Automated analysis

– Survey

• Z-crossing

– More sites

• Driver understanding

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings 18

Acknowledgments

• Co-authors

– Chris Monsere

– Katherine Carlos

• Data reduction/earlier iteration

– Deanna George

– Fahad Alhajri

Evaluating Driver and Pedestrian Behavior at Enhanced Multi-lane Midblock Pedestrian Crossings 19