Extracted text (via pymupdf)
25052 characters
A Comparative Study of Speed Humps, Speed Slots and Speed
Cushions
LaToya Johnson and A.J. Nedzesky
Abstract. The primary objective of this study was to compare speed humps with two
newer traffic calming devices that are gaining popularity in the US, the speed slot and
speed cushion. Crossing speed and driver behavior were measured at selected traffic
calming devices on roadways in the Washington DC metropolitan during the summer of
2003. The subject devices include:
•
12-ft and 22-ft asphalt speed humps;
•
14-ft prefabricated speed humps;
•
22-ft speed slots; and
•
10-ft speed cushions.
All ranged from 2.5 to 4.0 inches in height. Video surveillance technology was used to
collect data, including vehicle crossing speed, lateral placement and braking frequency.
Preliminary results revealed that speed slots allowed the highest average and 85th
percentile crossing speeds. Speed cushions, 12-ft speed humps and 14-ft prefabricated
speed humps recorded the lowest crossing speed and relatively high frequency of braking
maneuvers.
The designs of the speed hump and speed cushion encouraged drivers to travel centrally
within their lane. Lateral positioning while traversing the speed slot was varied; a large
percentage of drivers attempted to place the vehicle’s left tires in the slot.
INTRODUCTION
Statement of the Problem
As the adoption of various traffic calming practices continues throughout the U.S., use of
the speed hump as a standard traffic calming device steadily increases. However, speed
humps have also become the center of a traffic engineering controversy. Emergency
response agencies and community groups have been cited in the belief that speed humps
increase the amount of time for an emergency vehicle to respond to calls.(1,2,3) This has
resulted in hesitation and resistance regarding installation of speed humps. In reply to
these concerns, two variations of the speed hump design are beginning to gain popularity
in the U.S., the speed slot and speed cushion. Although the use of the speed slot and
cushion is fairly common in European countries, its effectiveness as a traffic calming
device in the U.S. is yet to be seen. Differences in driver behavior and vehicle
characteristics between European countries and the U.S. make research in this area vital
to the progress of traffic calming in the United States.
Research Goals
The goals of this effort was to perform a comparative analysis of the three traffic calming
devices by examining crossing speed, driver behavior and brake pedal use. Specific
questions to be addressed are:
•
How do speed humps, slots, and cushions affect driver’s speed at the device?
•
When the devices are placed in series, is the crossing speed at a second or third
device different than at the initial device?
•
How do speed slots and cushions affect driver’s selection of lateral crossing location
behavior differently than speed humps?
BACKGROUND
Speed Humps
In 1997 the Institute of Transportation Engineers (ITE) approved the Guidelines for the
Design and Application of Speed Humps, RP-023A, which provided recommended
practice based on national and international research and experience.(4) ITE reported that
speed humps should be installed on roadway facilities classified as local streets by the
American Association of State Highway and Transportation Officials (AASHTO). The
roadway should not be more than two travel lanes or traveled significantly by long wheel-
based vehicles. Additionally, it should have a horizontal curve of 300 feet radius or more
and a grade of eight percent or less. The posted or prima facie speed limit should be 30
mph or less; ITE warned that installation on roadways with a higher speed limit
warranted careful consideration.
Design
The design of a speed hump can be defined by specifying the length of its base, the height
of its crown and the shape of its surface profile, as shown in Figure 1. ITE recommended
a height of 3-inches for speeds of 20 to 25 mph and 4-inches for speeds of 15 to 20 mph.
For length, ITE recommended 12 feet.
Width = 24 ft
Height = 3-4 in
Length = 12 ft
Length = 12 ft
Figure 1. Schematic of a Typical Circular Speed Hump.
For use on a typical residential street, ITE reported that the most common designs are the
circular or parabolic speed hump as shown in Figure 2. An alternative design, the flat-
topped design, is also shown in the Figure 2.(4)
\
Circular
Flat-Topped
Parabolic
Figure 2. Typical Design Profiles of Speed Humps.
Speed Slots and Speed Cushions
Due to concerns that speed humps influence response times and passenger comfort of
emergency response vehicles, modified designs of speed humps were created.(1,2,3) Like
speed humps, speed slots and speed cushions are both raised areas across the road with
the intent of reducing vehicle speed. However, speed slots and cushions were designed to
avoid excessive discomfort or damage to emergency vehicles by making separations in
the hump. Figure 3 compares the typical design of speed humps, slots, and cushions.
Speed slots are similar to speed humps in that they extend across the roadway but they
have “slots” or tire grooves along each side of the centerline in order to allow emergency
response vehicles to avoid of the device by driving through the slots along the middle of
the road. Unfortunately the emergency vehicle must straddle the centerline and travel in
both lanes of the roadway, increasing the risk to both the emergency vehicle as well as
other vehicles.
Speed cushions are smaller than lane width and are rectangular or square in shape.(5)
These characteristics allow for an emergency response vehicle to straddle the cushion
while remaining in its respective lane. Figure 3 shows the typical dimensions and layout
of speed humps, slot and cushions.
Figure 3. Schematic of Speed Hump, Speed Slot and Speed Cushion.
22 ft
6 ft
Minimum of 1 ft
24 ft
3 ft
24 ft
12 ft
22 ft
10 ft
6 ft
12 ft Speed Hump
Speed Slot
Speed Cushion
Direction
of Tra
vel
Direction
of Tra
vel
22 ft Speed Hump
The basic designs of both the speed slot and speed cushion are very much like the speed
hump. However, additional modifications have been made for the speed cushion to
accommodate for the wider vehicle width of cars in the US. Table 1 shows
recommendations made by the City of Austin Texas and the United Kingdom Department
of Transport. Figure 4 shows a diagram of the typical speed cushion.
Table 1. Recommended Speed Cushion Design Characteristics.
Design Characteristics
Austin, TX(6)
United Kingdom(7)
Base Length
10-ft or 12-ft
2 to 2.5 m (6.56 to 8.20 ft)
Base Width
6.5 ft or 3 ft
1.6 to 1.9 m (5.24 to 6.23 ft)
Maximum Height
3 ±¼ in
80mm (4.15 in)
On/Off ramp Gradient
1:8 at 18 in
1:8
Side Ramp Gradient
1:6 at 24 in
1:4
Transverse Gap
12 in wide
750 and 1000 mm (2.46 ft to
3.28 ft)
Side Ramp
Direction
On
of Travel Off Platform Base Height
Ramp
Ramp Width Width
Platform Length
Base Length
Figure 4. Aerial and Cross-Sectional View of a Speed Cushion.
Speed Cushion Studies
A 1998 study by Layfield and Parry that examined speed cushion schemes in the United
Kingdom concluded that although speed cushions are not as effective as speed humps in
reducing speeds, they are important because they decrease driver discomfort especially in
large buses. Speed cushions were reported to have two to seven mph higher 85th
percentile crossing speeds than speed humps and one to two mph higher 85th percentile
speeds between devices. (8)
The study by Layfield and Parry found that passenger discomfort was low at speed
cushions for large buses if the cushions were straddled centrally, but otherwise had
similar effects as speed humps if not straddled centrally. (9)
Driver Behavior at Speed Cushions
In observing driving behavior, Layfield and Parry found that 55 percent of all cars and 90
percent of all buses in the study attempted to centrally straddle the speed cushions. In the
three abreast configuration, 40 percent of all drivers drove with one tire between the
nearside and middle cushions. (8) A study by Pau on how speed bumps may induce
improper driver behavior in Italy, characterized improper movement as total or partial
avoidance in a park or bus lane. This study found that a significant percentage of drivers
attempted to totally avoid speed humps by traveling in the park or opposite lane.(5)
Emergency Vehicle Response Time at Speed Cushions
In a 2000 study by Bunte investigated the effects of the speed cushion on the response
times in Austin Texas. Results showed that speed cushions had very little impact, if any,
on increasing response times of emergency response vehicles. Average delay times were
less than a second, except for the vehicle that was transporting a critically ill/injured
patient which had an average delay of 4.84 seconds on total travel time. Overall, the
study found that speed cushions are less detrimental to negatively impacting emergency
response times than speed humps.(10)
DATA COLLECTION METHODOLOGY AND ANALYSIS
Site Selection
The study investigated speed humps, slots and cushions in the Washington, D.C.
metropolitan area. Beyond device length, which was a function of the device type, the
following criteria were used to select the ten sites used in the study:
•
Height: 2.5 – 4.0 inches;
•
Separation: 150 – 700 ft;
•
Street Width: 25 – 35 ft;
•
Number of Lanes: 2 lanes, one in each direction;
•
Street Classification: Residential, local; and
•
Parking: Unrestricted on one or both sides.
For each site, observations and photographs were taken at each location to record road
geometry, classification, posted and advisory speeds, traffic volume and speed hump,
slot, or cushion characteristics. Table 2 presents the ten sites and type of associated
traffic calming device. Additional information about each site is found in Appendix A.
During the site selection process, local transportation officials were contacted.
Table 2 Sites Selected for Study.
Site ID
Device Type
Road Classification
Segment
Length
Street Width and
Parking
Posted Speed
1
Hump-12-ft
Residential, school
1079 ft
24 ft wide, parking on
both sides
25 mph*
2
Hump-12-ft
Residential
1388 ft
24 ft wide, parking on
both sides
25 mph*
3
Hump-12-ft
Residential, school
1427 ft
32 ft wide, parking on
both sides
25 mph
4
Hump-22-ft
Residential, Collector for
local interstate, school
816 ft
27 ft wide, permit parking
on both sides
25 mph
5
Hump-22-ft
Residential, Local, school
1866 ft
25 ft wide, parking on
both sides
25 mph**
6
14-ft
Prefabricated
Hump
Residential, school, major
hotel
1372 ft
30 ft wide; 2 lanes
parking on one side
25 mph
7
Slot
Residential
2857 ft
34 ft wide, 2 lanes,
parking on both sides
25 mph
8
Slot
Residential, collector for
Route 50, school
2837 ft
36 ft wide, 2 lanes,
parking both sides
25 mph
9
Cushion
Residential
2743 ft
26 ft wide, 2 lanes,
parking on both sides
25 mph
10
Cushion
Residential; cut-through
for two local arterials
2456 ft
27 ft wide, 2 lanes,
parking on one side
25 mph
* 15 mph advisory speed placard at device
** 20 mph advisory speed placard at device
Vehicle Classification
Vehicles were classified as to belonging to one of seven different groups. These
classifications were primarily based on vehicle suspension, handling and ground
clearance. The seven classifications are as follows:
•
Passenger car;
•
Luxury / High performance car;
•
Pick-Up Truck;
•
SUV / Minivan;
•
Trucks;
•
Buses; and
•
Other (service vans, etc).
Data Collection Methodology
Video camera surveillance was used to collect speed data and to document driver
behavior. A digital video camcorder discretely set-up at the site recorded driver response
to the devices. The placement of the video camcorder permitted the observation of the
traffic calming device and at least a 50 ft approach to the device. Data for vehicles
traveling in both directions were collected simultaneously. Data were collected for two-
hour periods during weekdays between 10 am and 2 pm during good weather conditions.
Data Reduction
Videos from the video camera surveillance were viewed; speed and lateral position data
were extracted and put in an Excel spreadsheet. Devices that were in a series were
designated as first, middle, or last at a particular site and the same type of device was
used throughout the series. A series consisted of either two or three devices.
Video-frame analysis was used to collect the crossing speed data. Based on the
knowledge that the video camcorder recorded 30 frames per second, the number of
frames needed for the vehicle’s front tire to traverse the length of the device was used to
calculate the vehicle’s average crossing speed. Only vehicles traveling under free-flow
conditions were used for the analysis; following vehicles were of no interest due to the
influence of a lead vehicle. Data were extracted for vehicles traveling in each direction.
Approximately 100 data points were used for each site when possible.
Crossing vehicles were classified as passenger car, sports utility vehicle or pickup
truck/minivan. Also, braking and any erratic behavior such as evidence of loss of control
was documented.
Data relating the driver’s choice of lateral placement when crossing the device was
subjectively recorded from the video. Lateral placement was classified as the following:
•
Driving in the center of the lane;
•
Crossing over the centerline;
•
Driving with the left tires in the groove (only for slots and cushions); or
•
Driving towards the right side of the lane.
RESULTS
Speed Analysis
Descriptive statistical analysis was performed, in which the average and 85th percentile
speeds and standard deviation were calculated. Various results will be highlighted in the
following tables. Table 3 shows the average and 85th percentile speeds of devices that
were either stand-alone or were the first in a series. As can be seen, the 12-ft speed
humps, 14-ft speed prefabricated speed humps and the speed cushions all generated
average speeds that were approximately 10 mph and 85th percentile speed that were less
than 15 mph. The 22-ft speed humps and the speed cushions had higher average speeds.
The 85th percentile speed at the speed slots was over 25 mph.
Table 3. Average and 85th Percentile Speed (in mph), by Device Type.
Device Type
Average Speed
85th Percentile Speed
Speed Hump-12-ft
9.6
12.3
Speed Hump-22-ft
15.2
18.8
Prefabricated Speed Hump-14-ft
10.6
14.3
Speed Slot
20.5
26.5
Speed Cushion
10.1
12.8
Table 4 shows the average and 85th percentile speeds for devices that were installed in
series. For installations that consisted of only two devices, the middle device column
contains “n/a.” Site 4 consisted of only one speed hump. Recall that all roads were
posted at 25 mph.
From the table it can be seen that speeds tended to remain relatively constant at each of
the devices in the series. Two sites demonstrated a variation in their average and 85th
percentile speeds. At site ID 5 (22-ft humps) speeds decreased and then increase along
the series of humps, which were spaced approximately 500 feet apart. At site ID 9 (speed
cushion) there was an increase along the series of cushions, which were separated by 550
feet apart.
It was observed that most drivers depressed their brakes when crossing any of the
devices, independent of the position of the device in the series.
Table 4 Average and 85th Percentile Speeds for the Various Devices, (in mph).
First Device
Middle Device
Last Device
Site
ID
Type of
Device
Average
85
Percentile
Average
85
Percentile
Average
85th
Percentile
1
12-ft-hump
10.8
12.3
n/a
n/a
9.9
12.3
2
12-ft-hump
10.1
12.2
n/a
n/a
10.2
12.3
3
12-ft-hump
9.4
11.9
n/a
n/a
9.4
12.2
4
22-ft hump
14.3
17.3
n/a
n/a
n/a
n/a
5
22-ft hump
16.3
19.6
14.6
17.4
19.2
23.7
6
Prefab
14-ft hump
10.6
14.3
n/a
n/a
10.5
13.0
7
Slot
19.5
24.7
18.3
23.7
18.6
23.7
8
Slot
21.2
26.5
17.7
21.4
19.4
22.5
9
Cushion
10.1
13.6
n/a
n/a
13.6
20.0
10
Cushion
10.1
12.0
9.7
11.4
10.5
13.3
Analysis of Traveling Speed by Vehicle Type
Crossing speeds were analyzed for each of the devices based on vehicle type. Speed
were classified as fitting into one of the following groups:
•
0.0-9.9 mph;
•
10.0 – 14.9 mph;
•
15.0 – 19.9 mph;
•
20.0 – 24.9 mph;
•
25.0 – 29.9 mph;
•
30.0 – 34.9 mph; and
•
35.0 mph and over.
No one vehicle group performed differently than any other vehicle group for a given type
of device. For the 12-ft humps, the most common traveling speed for each vehicle type
was in the 0.0-9.9 mph speed category. For the 22-ft humps, the most travel speed for
each type of vehicle was in the 15.0-19.9 mph speed category. For the 14-ft temp humps,
approximately half of the vehicles were classified in the 0.0-9.9 mph speed category and
half in the 10.0-14.9 mph speed category. For speed cushions, the majority of speeds
were in the 0.0-9.9 mph speed category.
Table 5 presents the percentage of vehicles traveling in each speed category at the speed
slots. For the most part, travel speeds were in the 15.0-19.9 mph speed category. The
only result of interest was the percentage of vehicles (shown in italics) that were
identified in the speed categories 30.0-34.9 and 35.0 mph and over.
Table 5. Percentage of Vehicles Traveling in each Speed Category, by Vehicle Type for
Speed Slots.
Traveling speed (mph)
Vehicle Type
Number of
Observations
0.0 -
9.9
10.0 –
14.9
15.0 –
19.9
20.0 –
24.9
25.0 –
29.9
30.0 –
34.9
35.0
+
Passenger Cars
238
2.1
18.5
48.7
22.3
3.8
2.1
2.5
Luxury and High
Performance
72
2.8
15.3
55.6
11.1
12.5
0.0
2.8
Pick-up trucks
34
2.9
23.5
35.3
17.6
11.8
5.9
2.9
SUVs and
Minivans
155
3.9
18.1
41.3
22.6
5.8
4.5
3.9
Trucks
0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Buses
10
20.0
30.0
10.0
10.0
20.0
10.0
0.0
Other, (e.g.,
service vans)
25
0.0
32.0
44.0
16.0
4.0
0.0
4.0
Lateral Placement Analysis
A similar analysis was performed looking at driver selection of the vehicle’s lateral
placement when crossing the device. Vehicles were classified as either driving in the
center of the lane, crossing the centerline, driving towards the left (or right) of the lane.
As expected, since speed humps do not offer the driver the opportunity of traversing with
a tire (or pair of tires) not contacting the hump, lateral placement of the vehicle tended to
be in the center of the travel lane. Lateral placement at speed slots was also consistent;
however, at speed slots most drivers tended to drive with their left tires along the grooves
of the slot.
Table 6 shows that when traversing a speed cushion, most drivers either chose to have
their vehicle centrally located over the cushion or traverse the cushion with their left tire
in the groove. The lateral placement selected by pick-up truck drivers is of possible
concern. Even with the small sample size of 27 it was noted that almost twice as many
pick-up trucks crossed the centerline, in an attempt to cross the smaller cushion located
under the centerline. The average speed of vehicles crossing the centerline was 10 mph,
which indicates that these drivers were not traveling at an unsafe speed, but this is an
erratic maneuver that may surprise oncoming drivers. No other erratic behavior (e.g.,
sudden braking, swerving, etc.) was observed at the speed cushions.
Table 6. Lateral Placement of Vehicles by Vehicle Type for Speed Cushions.
Lateral Placement
Vehicle Type
Number of
Observations
Center
Over
Centerline
Left Tire
in Groove
Towards Right,
Right Tires in Slot
Passenger Cars
246
39.0
12.2
39.0
9.8
Luxury and High
Performance
64
32.8
10.9
53.1
3.1
Pick-up trucks
27
44.4
22.2
25.9
7.4
SUVs and Minivans
103
40.8
8.7
42.7
7.8
Trucks
0
0.0
0.0
0.0
0.0
Buses
0
0.0
0.0
0.0
0.0
Other, (e.g., service
vans)
9
33.3
22.2
33.3
11.1
CONCLUSION
After collecting data for almost 2000 vehicles, it was found that speed slots followed by
22-ft speed humps allowed the highest average and 85th percentile crossing speeds.
Twelve-ft speed humps, speed cushions and prefabricated 14-ft speed humps recorded the
lowest crossing speeds.
The design of the speed hump encouraged drivers to travel centrally within their lane.
Lateral positioning while traversing the speed slot and cushion varied. At speed slots a
large percentage of drivers shifted to the left, in an attempt to place the vehicle’s left tires
in the slot. At speed cushions, drivers tended to drive either centrally down the lane or
shifted towards the left of the lane to place the left tires in the groove.
Speed slots, with many drivers shifting towards the left side of their lane, exhibited the
highest average and 85th percentile speeds in this study, and speed cushions, with a large
percentage of pick-up truck drivers crossing the centerline in order to traverse the
cushion, would appear to present a safety concern to the unsuspecting, oncoming driver.
Unfortunately, crash data was not collected as part of this exercise. Future research to
investigate these hypotheses may be justified.
It is recommended that further research investigate lateral acceleration generated by the
various devices for selected vehicle types as well as device spacing.
Special Thanks
The authors would like to thank the local traffic engineers working in the City of
Alexandria, Arlington and Fairfax Counties, Virginia, Washington D.C. and Montgomery
County, Maryland. Additional thanks are also due to Dr. Karen Dixon and Dr. Gabriel
Rousseau.
Contact Information:
LaToya Johnson
Georgia Institute of Technology
lnjohnson1300@hotmail.com
A.J. Nedzesky
FHWA
6300 Georgetown Pike
HRDS-05
McLean, VA 22101
(202) 493-3369
aj.nedzesky@fhwa.dot.gov
APPENDIX A: SELECTED SITE SPEED DEVICE PROFILE
Site
ID
Profile/
Configuration
Height
Length Width
Gap
Separation
Construction Markings
1
Parabolic
2.5 in
12-ft
n/a
n/a
Range of 130-
383 ft
Asphalt
Zebra
2
Parabolic
3.0 in
12-ft
n/a
n/a
437 & 419 ft
Asphalt
Zebra
3
Parabolic
3.0 in
12-ft
n/a
n/a
600 ft
Asphalt
Chevron
4
Parabolic
3.0 in
22-ft
n/a
n/a
460 ft
Asphalt
Chevron
5
Parabolic
3.5 in
22-ft
n/a
n/a
Range of 430-
530 ft
Asphalt
Zebra
6
Flat-top
4.0 in
14-ft
n/a
n/a
150 & 161 ft
rubber
arrow on
road prior to
hump
7
symmetrical
about centerline
3.0 in
22-ft
5 ft &
12-ft
18 in
490-535
Asphalt
Diagonal
Lines
8
symmetrical
about centerline
3.0 in
22-ft
5 ft &
14-ft
17.5 in
470-575 ft
Asphalt
Diagonal
Lines
9
three cushion
abreast;
symmetrical
about centerline
3.0 in
10-ft
7 ft
24 in
505 & 634 ft
Asphalt
Arrow
10
three cushion
abreast; middle
cushion off set
from centerline
3.5 in
10-ft
7 ft
18 in
285 & 470 ft
Asphalt
Arrow
REFERENCES
1. Stephens, Burton W. Road Humps for the Control of Vehicular Speeds and Traffic
Flow. Public Roads, Vol. 50, No. 3, December 1986, pp. 82-90.
2. Gorman, Michael, M. Moussavi, and P. McCoy. Evaluation of Speed Hump Program
in the City of Omaha. ITE Journal, Vol. 69, Issue 6, June 1989, pp.28-32.
3. Sumner, R. and C. Baguley. Speed Control Humps on Residential Roads. U.K.
Transport and Road Research Laboratory Report 878. TRRL, Crowthorne,
Berkshire, England, 1979.
4. Guidelines for the Design and Application of Speed Humps. Report RP-023A. ITE
Traffic Engineering Council Speed Humps Task Force, 1997.
5. Pau, Massimiliano. Speed Bumps May Induce Improper Drivers’ Behavior: Case
Study in Italy. Journal of Transportation Engineering. Sept. 2002, pp.472-478.
6. Austin, TX Speed Cushion Specifications. April 2002. Item No. SS 884 Speed
Cushions.
7. Speed Cushion Schemes. U.K. Department of Transport Advisory Leaflet. January
1998. http://www.roads.dft.gov.uk/roadnetwork/ditm/tal/traffic/01_98/. Accessed
June 5, 2003.
8. Layfield, R.E. and D.I. Parry. Traffic Calming: Speed Cushion Schemes. U.K.
Transport and Road Research Laboratory Report 312. TRRL, Crowthorne,
Berkshire, England, 1998.
9. Bunte, Les. An Analysis of Speed Cushion Effects on Response Times. August
2000.