Pedestrian Crash Patterns and Visibility

1
Pedestrian Crash Patterns and Visibility

• Michael Flannagan
• John Sullivan
• Michael Sivak
• University of Michigan
• Transportation Research Institute
• March 14, 2007

2
Relevant reports

• UMTRI-93-33
• UMTRI-95-44
• UMTRI-99-21
• UMTRI-2001-33
• UMTRI-2004-14
• UMTRI-2005-22
• UMTRI-2006-1

3
Vehicle occupant deaths, FARS, 1987-2003
4
Pedestrian deaths, FARS, 1987-2003
5
Pedestrians as a proportion of all road traffic
fatalities
6
Pedestrians as a proportion of all road traffic
fatalities
7
Fatalities per 100,000,000 vehicle miles, U.S.
2003 (National Safety Council)
8
Analysis of nighttime risks
9
Isolating the effects of light in crash data

• Comparing night and day is not sufficient
• Differences between night and day
• Ambient light
• Alcohol
• Fatigue
• etc.
• Isolate light via seasonal and DST changes
  (assume exposure is linked to clock, not sun)

10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
Crash counts around fall PM return to standard
time
15
Crash counts around fall PM return to standard
time
16
Dark/light ratios in DST data from UMTRI-2001-33
17
Significant effects of light by first harmful
event, FARS 1987-1997 (UMTRI-2001-33)
18
Estimated effects of improved (perfect) lighting
UMTRI-2001-33
19
The effects of natural light/darkness

• When isolated from other factors that differ
  between night and day, the effects of light/dark
  are still very strong, but very specific (2,300
  pedestrian fatalities per year).
• For crashes that might be addressed by improved
  headlighting, virtually the whole story is
  pedestrian crashes.

20
People overdrive low beams

• Perel, Olson, Sivak, Medlin (1983) Safe speed
  with low beams is 70 km/h 45 mph.
• Burgett, Matteson, Ulman, Van Iderstine (1989)
  Maximum speed for which adequate light is
  achievable is 40 mph 64 km/h.

21
Pedestrian fatalities by posted speed
(UMTRI-2006-1)
22
Perfect lighting - potential safety benefits by
road class
UMTRI-2001-33
23
Possible countermeasures

• Pedestrian
• Stay out of the way
• Wear light clothes or retroreflectors
• Vehicle
• Better headlamps (HID, AFS, LED)
• Night vision systems
• Infrastructure
• Separation of pedestrians (sidewalks, crosswalks)
• Lighting

24
Fatal pedestrian crashes in the dark, or dark
with light, by alignment (FARS 1999)
25
The effect of darkness by roadway alignment (DST
analysis)
26
Fatal pedestrian crashes in the dark, or dark
with light, by profile (FARS 1999)
27
The effect of darkness by roadway profile (DST
analysis)
28
Stopping distance by initial speed (RT 1.5 s,
Braking 0.5 g)
29
Percent of encounters with pedestrian visible,
simulated (Bhise et al., 1977)
30
Recent trends in headlighting

• Reports for more information
• UMTRI-2004-23
• UMTRI-2004-25
• Method
• 2004 model year
• Photometer headlamps of 20 vehicles
• Include HID
• 25, 50, and 75 percentile light levels

31
Median Illuminance Values
32
Median Luminous Intensities
33
Halogen versus HID Headlamps
34
2004 Halogen minus 1997 Halogen
35
2004 Halogen minus 2004 HID
36
Two key low-beam test points
37
Halogen versus HID photometry