Conference PaperPDF Available

Modeling Double Parking Impacts on Urban Street

Authors:

Abstract and Figures

Double parking (DP) is one of the key contributors to traffic congestion on urban streets. Double parking violations of commercial vehicles while they load and unload at delivery locations with insufficient curbside space can have significant negative impact on traffic. Motivated by the need to study such impact in urban cities, this paper utilizes parking violation records for New York City along with field data collected using video recording, and adopts a comprehensive modeling approach that combines available data with two types of models. The first is an M/M/∞ queueing model used to estimate double parking effect on the average travel time. The second is a micro-simulation model developed and calibrated to study individual and combined effects of various explanatory variables. Both models account for different effects of general vehicles and commercial trucks. Via case studies in Midtown Manhattan and Downtown Brooklyn (New York, US), double parking activities and driver behaviors are investigated and used for comparative analysis. The M/M/∞ queueing model has been empirically validated using field data collected as part of this study. Comparison results show a good fit for uncongested traffic conditions. Micro-simulation results indicate different impact levels for 21 scenarios in four categories namely, travel demand, double parking locations, frequency, and durations. This study can provide traffic agencies a potential approach to quantify the impact of double parking in a large-scale network and insights into the management and alleviation of on-street parking problems including incentives for encouraging off-hour deliveries and more effective enforcement during peak hours.
No caption available
… 
Content may be subject to copyright.
Gao, Ozbay 1
MODELING DOUBLE PARKING IMPACTS ON URBAN STREET
1
2
Jingqin Gao, M.Sc. (Corresponding author)
3
Graduate Research Assistant, Department of Civil and Urban Engineering;
4
New York University (NYU)
5
Six MetroTech Center, 4th Floor,
6
Brooklyn, NY 11201, USA
7
Tel: (646) 717-3652
8
E-mail: jingqin.gao@nyu.edu
9
10
Kaan Ozbay, Ph.D.
11
Professor, Department of Civil and Urban Engineering;
12
Center for Urban Science and Progress (CUSP);
13
New York University (NYU)
14
One MetroTech Center, 19th Floor,
15
Brooklyn, NY 11201, USA
16
Tel: (646) 997-0552, Fax: (646) 997-0560
17
E-mail: kaan.ozbay@nyu.edu
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Word count: 4696 Texts + 4 Table + 7 Figures = 7446
38
Abstract: 250
39
Reference: 486
40
Submission Date: August 1st, 2015
41
Revision Date: November 15th, 2015
42
43
44
Paper submitted for Presentation and Publication in the
45
Transportation Research Board’s 95th Annual Meeting, Washington, D.C., 2016
46
Gao, Ozbay 2
Abstract
1
Double parking (DP) is one of the key contributors to traffic congestion on urban streets.
2
Double parking violations of commercial vehicles while they load and unload at delivery locations
3
with insufficient curbside space can have significant negative impact on traffic. Motivated by the
4
need to study such impact in urban cities, this paper utilizes parking violation records for New
5
York City along with field data collected using video recording, and adopts a comprehensive
6
modeling approach that combines available data with two types of models. The first is an
7
M/M/ queueing model used to estimate double parking effect on the average travel time. The
8
second is a micro-simulation model developed and calibrated to study individual and combined
9
effects of various explanatory variables. Both models account for different effects of general
10
vehicles and commercial trucks. Via case studies in Midtown Manhattan and Downtown Brooklyn
11
(New York, US), double parking activities and driver behaviors are investigated and used for
12
comparative analysis. The M/M/ queueing model has been empirically validated using field data
13
collected as part of this study. Comparison results show a good fit for uncongested traffic
14
conditions. Micro-simulation results indicate different impact levels for 21 scenarios in four
15
categories namely, travel demand, double parking locations, frequency, and durations. This study
16
can provide traffic agencies a potential approach to quantify the impact of double parking in a
17
large-scale network and insights into the management and alleviation of on-street parking
18
problems including incentives for encouraging off-hour deliveries and more effective enforcement
19
during peak hours.
20
21
Gao, Ozbay 3
INTRODUCTION
1
In urban areas, obstructions of traffic such as double parking, commercial vehicle deliveries,
2
pedestrian jaywalking, taxi pick-ups and drop-offs, are potential impediments to road capacity and
3
vehicular speed, and causes traffic delay and safety risks. New York City Department of Finance
4
(NYCDOF) defines double parking” as standing or parking a vehicle on the roadway side of a
5
vehicle already stopped, standing or parked at the curb” (1). It is mainly due to the lack of available
6
on-street parking spaces and sometimes makes the street impassable, especially in one-way single-
7
lane situations. According to New York City Department of Transportation (NYCDOT) parking
8
regulations, double parking of passenger vehicles is illegal at all times in New York City (NYC),
9
regardless of location, purpose or duration (2). In most cases, a double parked vehicle will block
10
part of the street and a bus or a bicycle lane, if one is present. Double parking can become a serious
11
problem especially in congested urban areas since it is both obstructive and irritating.
12
Most cities rely on fines when dealing with double parking (3). In 2014, double parking
13
violation had 502,082 ticketed cases and was ranked seventh among all the parking violations in
14
New York City according to NYCDOF records. In reality, the total number of double parking
15
violations is highly underestimated as most of such activities may not even be recorded or ticketed.
16
Notably, it becomes more complex in mixed-used or commercial zones while double-parked
17
commercial vehicle can lead to further reduction in capacity and increase in travel time (TT) and
18
delay.
19
There are several other strategies to reduce double parking of commercial vehicles,
20
especially during the peak hours. A recent study that focused on Off-Hour Delivery (OHD)
21
strategies emphasizes the importance of promoting these strategies to reduce congestion caused by
22
double parking delivery trucks. On the one hand, driving on an OHD route saves about
23
$9,000/year/OHD-tour in parking fines since it is easier for truck drivers to find legal parking
24
spaces near their delivery destination during the off-hours (4). On the other hand, increased
25
enforcement of parking fines for double parking during the regular hours could encourage more
26
carriers to participate in OHD (5).
27
One of the challenges in quantifying the effects of double parking is the lack of reliable
28
analytical approaches dealing with this important problem. There’s a need to identify a robust
29
analytical methodology so that a large number of sites can be studied efficiently to quantify the
30
impact of double parking as well as various long and short term parking enforcement and
31
management strategies in a large urban network such as New York City. The use of a very detailed
32
microscopic simulation tool for this purpose might not be feasible due to time and budget
33
constraints.
34
As an alternative approach, a stochastic queuing model proposed in Gursoy-Baykal et al.
35
(6) to estimate incident induced delays is adopted to estimate average delayed travel time, given
36
the amount and duration of double parking activities. This kind of queuing model applicable to
37
transportation systems, can be applied at the macroscopic level, consists of multiple road segments
38
modeled as ‘individual servers’ that provide service to vehicles that join the queue. It has a set of
39
simplifying assumptions about vehicle arrival distribution, service distribution, and number of
40
servers to derive equations to predict the queue lengths and waiting times. It can be magnitude of
41
order inexpensive and faster to develop and implement such models compared with microscopic
42
simulation models. However, accuracy of such macroscopic models for various spatial-temporal
43
characteristics of a highly complex urban transportation network has to be validated carefully. This
44
paper describes such two models and discusses pros and cons of each approach along with
45
suggestions for improving both models.
46
Gao, Ozbay 4
LITERATURE REVIEW
1
Various studies have been taken into consideration for on-street parking and traffic congestion. In
2
2009, Portilla et al. (7) extended a model developed by Baykal-Gürsoy et al. (6) to quantify the
3
influence of parking maneuvers and badly parked cars on average link journey times. Baykal-
4
Gürsoy et al.s model (6) was a M/M/ queueing model subjects to random interruptions of
5
exponentially distributed durations that was originally developed to simulate the effects incidents
6
have on delays that she had proposed in a previous article (8). It describes a queuing system where
7
arrival and service processes of the vehicles are all Poisson processes with infinite number of
8
servers. In Portilla et al.’s study, badly parked cars are introduced as time events while parking
9
maneuvers are introduced as high-frequency short duration event. The results showed a 15%-199%
10
increase in average journey time and a 6%-55% reduction in capacity. Guo et al. (9) evaluated
11
influence factors of on-street parking by using a proportional hazard-based duration model.
12
Narrow lanes, frequent turnover rates, and parking occupancy show a negative effect on travel
13
time.
14
However, only a few studies focus on the occurrence of double parking. Recent efforts
15
have primarily concentrated on the price strategies or applying microsimulation models to explore
16
on-street parking impact on traffic. Millard-Ball et al. (10) evaluated the first two-year
17
performance of SFpark (11) a demand-responsive rate adjustment parking system embarked by
18
San Francisco Municipal Transportation Agency. Rate changes have helped achieve the city’s
19
occupancy goal (60%~80%) and reduced cruising by 50%. Another pilot called “PARK Smart
20
(12) was implemented in New York City by NYCDOT. Schaller et al. (13) summarized that
21
pricing can be effective in achieving the goals of commercial loading availability, increasing
22
turnover, and parking availability. Morillo and Campos (14) showed that the economic cost differs
23
depending on the DP location and the type of typology.
24
Besides pricing strategies and economic impact, Kladeftiras and Antoniou (15) conducted
25
a research project using microsimulation to estimate the impact of average speed, delay and
26
stopped time, as well as environmental impact caused by double parking. An auxiliary lane is
27
created in the model for one-lane roadways for the vehicles to overtake the double-parked vehicle.
28
Results showed that by eliminating double-parking activities, it can increase 10% - 15% in average
29
speed and reduce 15% and 20% in delay and stopped time. Researchers in Italy also studied illegal
30
parking impact in a high-density urban area in Palermo, Italy (16). Results showed that parking
31
durations over 5 minutes would have significant negative impacts.
32
To sum up, most studies in the literature only provide insights into policy and economic
33
aspects of double parking. Only a few micro-simulation models are used to simulate double
34
parking impacts, but these are mainly outside the US and usually shown to be time-consuming and
35
labor intensive for network-wide implementation. Therefore, there is a need to identify a more
36
efficient macroscopic approach such as the queueing based model described in Gursoy-Baykal et
37
al. (6) to study the impact of double parking activities. Moreover, such macroscopic methods have
38
to be validated using real-world data and/or microscopic simulations. The feasibility of using
39
microscopic simulation for modeling double parking under various street configurations has to be
40
established since none of the major simulation tools has a default functionality for modeling double
41
parking. This paper attempts to address both these needs by collecting and analyzing field data and
42
validating a queueing based macroscopic model, as well as demonstrating a Paramics based
43
microscopic simulation model extended through a special application program interface (API)
44
developed by the research team to realistically modeling of double parking on urban streets.
45
46
Gao, Ozbay 5
DATA DESCRIPTION
1
NYCDOF’s parking violation records in 2014 are summarized to investigate the frequency of
2
double parking activities in NYC. As shown in the following table (TABLE 1), double parking
3
ranked as the seventh highest parking violation in NYC, which represents 5.0% of the total
4
NYCDOF’s parking violation records in 2014. Since most of the double parking events are
5
difficult to record, this number can be assumed to be significantly less than the real number of such
6
violations.
7
8
TABLE 1 Top 10 NYC Parking Violations in 2014
9
Rank
Violation Code
Tickets
Percentage
Violation Description
1
21
1,337,993
13.4%
Stand or park during street cleaning
2
38
1,302,810
13.1%
Muni meter: failing to show a receipt
3
14
884,828
8.9%
General no standing
4
37
745,096
7.5%
Muni Meter: excess of allowed time
5
20
584,334
5.9%
General no parking
6
71
556,677
5.6%
No Inspection Sticker
7
46
502,082
5.0%
Double parking
8
7
482,051
4.8%
Not stop at red light
9
40
479,191
4.8%
Stand or park beside a fire hydrant
10
36
457,141
4.6%
Exceeding speed limit in school zone
10
An automatic geocoding program is developed by the research team based on Google
11
Geocoding API (17) to convert the original address information into geographic coordinates that
12
can be employed to geolocate each event on a map.
13
Violation records are also visualized onto streets and neighborhoods to identify DP
14
“hotspots”. The goal of the geocoding and visualization is to understand the spatial-temporal
15
nature of double parking violations in New York City.
16
FIGURE 1 shows the results of four-month data from July to October in 2014. On the
17
neighborhood level, the neighborhood of Upper East Side, Upper West Side, and Midtown
18
Manhattan have the highest double parking violation records for all type of vehicles. The
19
neighborhood of Upper East Side, Upper West Side, and Hudson Yards-Chelsea-Flatiron-Union
20
Square ranked top three of double parking violation records for commercial vehicles. Double
21
parking has more violation records in commercial districts or mixed commercial/residential
22
districts, while commercial DP has less records compare to total DP in boroughs other than
23
Manhattan that are mainly residential. It must also be noted that Police enforcement is a crucial
24
factor to the number of summons for passenger cars. However, this may not necessarily reduce the
25
occurrence of commercial vehicle double parking in large and highly congested urban areas such
26
as NYC, since their loading and unloading activities are often along a short distance at multiple
27
destinations with not many opportunities of legal parking during a given time frame.
28
On the street level, 58th Street and Livingston Street ranked the first in Manhattan and
29
Brooklyn. The highest number is up to 797 total violations and 420 commercial violations on a
30
single street block in July to October in 2014.
31
Gao, Ozbay 6
1
(a) Total DP by street (b) Total DP by neighborhood
2
3
(c) Commercial DP by street (d) Commercial DP by neighborhood
4
5
FIGURE 1 NYC double parking violation tickets, 2014 July to October.
6
Gao, Ozbay 7
Case Study Sites
1
After identifying the hotspots, two case study sites, where the occurrence of double parking is
2
found to be significant from the above analysis, were chosen specifically (FIGURE 2).
3
1. Manhattan: 58th Street between 8th Avenue and 9th Avenue (highest number of records)
4
2. Brooklyn: Jay Street between Myrtle Avenue and Metrotech Center (observation of high
5
DP occurrence)
6
7
8
FIGURE 2 Double parking activities in study sites.
9
10
Traffic movements were recorded in April, July and November in 2015 during AM (8-
11
9AM), Midday (12-1PM), and PM (5-6PM) time periods for both sites to obtain traffic flow, speed,
12
queue length, DP information, and driver behaviors. The following observations are found from
13
the recorded videos:
14
Double parked trucks have more impact than general cars (e.g. further reduction on the
15
effective lane width).
16
Manhattan site has a shorter duration of DP events compared to the Brooklyn site.
17
The passing speed is generally higher than the default Paramics incident passing speed
18
(10MPH) which indicates that double parked events should not be treated same as
19
incidents.
20
Speed reduction is subject to multiple factors such as the location of DP.
21
For both sites, AM and Midday have noticeable number of DP events (20-30/hour), while
22
PM has very few DP events (1-2/hour). The regular curbside parking is not fully
23
occupied and the turnover rate is high during PM period.
24
25
Double parking information includes location, arrival time, departure time, vehicle type
26
and duration. A sample count sheet is shown in TABLE 2. The location of the DP events is
27
according to the th (      regular curb-parked vehicle measured from the upstream end
28
of the link. So the “location” shown in TABLE 2 stands for the double parking location parallel to
29
the regular curb parking. AM period is used for both study sites to better distinguish the delay
30
caused by double parking from the delay caused by midblock jaywalking pedestrians, which
31
occurs a lot during Midday.
32
33
Gao, Ozbay 8
TABLE 2 Sample DP Events
1
Location
Arrival Time
Departure Time
Duration
Vehicle Type
5
8:41:18
8:41:52
0:00:34
Car
5
8:46:45
8:49:15
0:02:30
Car
6
8:50:11
8:52:08
0:01:57
Truck
2
8:51:15
8:51:55
0:00:40
Car
5
8:52:52
8:56:53
0:04:01
Car
2
8:57:09
9:04:17
0:07:08
Car
4
8:02:30
8:02:53
0:00:23
Taxi
1
8:04:40
8:06:16
0:01:36
Car
1
8:34:20
8:35:10
0:00:50
Truck
5
8:45:10
8:46:25
0:01:15
Car
2
Traffic information is collected and summarized for every 15 minutes (TABLE 3). One
3
important observation for the Brooklyn site is that the traffic was experiencing severe downstream
4
blocking during the last 15 minutes of the observed AM hour.
5
6
TABLE 3 Traffic Information for Case Studies
7
Location
Time
Volume
HV%
# DP
Status
Avg.TT(s)
Standard deviation(s)
Brooklyn
8:00-8:15 AM
66
12.1%
6
DP
5.52
1.19
w/o DP
5.26
0.93
8:15-8:30 AM
58
19.0%
3
DP
5.47
0.98
w/o DP
5.07
0.91
8:30-8:45 AM
59
18.6%
4
DP
6.01
1.83
w/o DP
6.00
0.82
8:45-9:00 AM
65
12.3%
8
DP
14.03
15.49
w/o DP
4.61
0.02
Manhattan
8:00-8:15 AM
114
17.5%
8
DP
7.86
2.54
w/o DP
6.74
1.43
8:15-8:30 AM
108
13.0%
6
DP
8.09
1.27
w/o DP
7.77
0.12
8:30-8:45 AM
107
15.9%
6
DP
8.15
1.40
w/o DP
6.93
0.98
8:45-9:00 AM
127
29.9%
7
DP
8.57
0.63
w/o DP
6.74
1.14
8
METHODOLOGY
9
M/M/ queueing model
10
A queueing model assumes that the space occupied by an vehicle on the roadway link represents
11
one queueing “server”, which starts its service once a vehicle joins the link and carries the
12
“service” until the end of the link is reached (18). The notation of queueing models is classified by
13
arrival distribution, service distribution and the number of servers. A queueing model that has a
14
Markovian arrival rate, a Markov modulated service rates, and an infinite number of servers in the
15
system is called M/M/ queueing model. A Markov process defines discrete states and the
16
Gao, Ozbay 9
transition probability from one state to another. One of the key features is that it solely depends on
1
the current state. The M/M/ model is a stochastic process that has the following features:
2
Arrivals process is a Poisson process at rate λ.
3
Exponential service times with parameter μ. There are always sufficient servers in the
4
system that every vehicle is served immediately after arriving.
5
When a vehicle arrives or departs, the system moves to an adjacent state
6
When double parking occurs, it indicates the high occupancy of the curbside parking lane.
7
This is also observed in the video with a turnover rate of only one or two vehicles per hour for
8
regular curbside parking. Thus, the regular curbside parking impact is negligible in this case study.
9
Adopted from (6, 7, 18), the system is designed with two server states “Failure (F)or “Normal
10
(N)” to represent the conditions with or without double parking events. Let denote the rate of DP
11
clearance time and the frequency of DP events. The system will be in state   if there are
12
vehicles in the system that are interrupted by double-parked vehicles, while the system will be in
13
state   if there are vehicles in the system that are travelling without interruption. The
14
depiction of the modified M/M/ queueing model on a two-lane link is shown in FIGURE 3. The
15
servers in the systems will work at a low service rate μ’ when a double parking event happens
16
(FIGURE 4).
17
18
FIGURE 3 Modified M/M/ queueing model on a two-lane link (based on (6)).
19
FIGURE 4 State transitions for M/M/ model with two server states (based on(18)).
20
21
One important result from Baykal-Gursoy and Xiao (8) showed that the expected number of
22
vehicles on the link can be represented as below when the M/M/∞ queueing system is experiencing
23
service interruptions:
24
2( ') ( )( ')
E(X) (1 )
( ) ( ' ')
ff
r f r f
   
 
 
 
 
(1)
25
26
Average travel time on the link is:
27
r
0, N
λ
1, N
µ
0, F
λ
1, F
µ’
r
f
f
λ
3, N
λ
3, F
3µ’
r
f
λ
2, N
λ
2, F
2µ’
r
f
4µ’
λ
λ
(i+1)µ
λ
λ
(i+1)µ
λ
i, N
λ
i, F
iµ’
r
f
Gao, Ozbay 10
2
( ) 1 ( ') ( )( ')
W = (1 )
( ) ( ' ')
E X f f
r f r f
   
 
 
 
 
(2)
1
To reflect road segment characteristics, Equation (2) is reformatted by Portilla et al. (7) to compute
2
average travel time t:
3
'
(1 ) ( )( ')
[1 (1 )]
1'
( ' )
vv
ff v v
LvL
tv v v
vf fv
d d L

 
 
(3)
4
Where L=length of the link (mile); D==traffic demand (veh/h); f=frequency of double parking
5
(events/hour), d=1/r=average duration time of double parking (h); v==average speed without
6
double parking (mph); and v’==average speed with double parking (mph).
7
8
With the intention of accommodating commercial vehicle violations, a double parked commercial
9
vehicle is treated as two double parking events in this study. Moreover, double parking impact can
10
vary among different street segments due to traffic volume, number of lanes, gross leasable area
11
(GLA) of commercial properties and so on. To capture these effects, equation (3) is modified by
12
adjusting a correction factor
as follows:
13
'
(1 ) ( )( ')
[1 (1 )]
1'
( ' )
f
vv
ff v v
LvL
tC v v v
vf fv
d d L

 
 
(4)
14
And,
15
1
1 Real TT
Model TT
N
fi
CN
(5)
16
Where Real TT = field travel time (s), Model TT = queueing model travel time (s), N = number of
17
observations.
18
19
We propose a correction factor of
= 1.07 for Manhattan site and
=1.02 for Brooklyn site
20
based on the observed data. However, since the observed data is limited, we are not capable of
21
generalizing
for all possible influencing factors. This will be the focus of future study where
22
automated data collection methods can be deployed.
23
24
Microsimulation Model
25
A Paramics (19) micro-simulation (FIGURE 5) is developed for the Brooklyn site to
26
simulate complex conditions especially when downstream blocking happens. The model is
27
bounded by Jay Street to the West, Flatbush Avenue Extension to the East, Tillary Street to the
28
North, and Willoughby Street to the South. The particular interest of this study focuses on the road
29
segment between Myrtle Avenue and Metrotech Center on Jay Street where DP data is available.
30
Its a two-way segment has one 10-ft travel lane, one 4-ft bike lane and one 10-ft curbside parking
31
lane in each direction and is referred as the reference link in this study.
32
Paramics is a high-performance software that model the movement and behavior of
33
individual vehicles (19). To simulate double parking activities, an “incident” file coded in C
34
programming language is used to specify where, when and how long the double parking event
35
should occur. Double parking is programmed as a special incident type that has the event duration,
36
Gao, Ozbay 11
lane specification and overspills. The occurrence can be defined as either random or fixed on each
1
link. In order to test the location impact of DP event, this case study uses fixed incident with
2
defined distance and time.
3
Since the incident feature does not work well for single-lane situations while the vehicle
4
behind the double-parked event will not move until the event is cleared, a narrow auxiliary lane is
5
used along with a customized API plug-in to allow vehicles to overtake the DP vehicle. The API
6
plug-in is developed to slow down vehicle on a certain distance of the road link to simulate driver
7
behaviors when passing a DP vehicle. The location, length, and speed of the slow-down area can
8
be customized. The micro-simulation model is calibrated based on field traffic counts, average link
9
travel time and observed behaviors collected in April, 2015. It is also validated with Novemeber’s
10
data in the same year to demonstrate reasonable prediction capability. Validation results shows a
11
GEH (20) of traffic volume less than 0.20 and a root mean square error (RMSE) of travel time less
12
than 1.20 for all time intervals.
13
14
15
FIGURE 5 Case study area in Downtown Brooklyn and Paramics network.
16
17
The Traffic Engineering Handbook (21), along with other publications (7, 22), have addressed the
18
effect of curb parking, mainly caused by the lack of a curb lane plus the parking activity next to a
19
moving traffic stream. To estimate the impact on the average travel time due to different factors
20
under DP condition, 21 different scenarios were designed into four categories:
21
Category 1: Location of Double Parking Events Scenarios
22
Scenario 1-3: Double parking events happen at the beginning of the block, midblock, and
23
the end of the block
24
Category 2: Demand Scenarios
25
Scenario 4-7: 110%, 120%, 130%, 140% of morning peak demand
26
Category 3: Frequency of Double Parking Events Scenarios
27
Scenario 8-11: Increase one, two, three, and four double parking event every 15 minutes
28
Category 4: Duration of Double Parking Events Scenarios
29
Scenario 12-21: Average duration of the events is increased by one minute in each scenario
30
(e.g. Scenario 12 has one-minute average duration, Scenario 21 has 10-minute average
31
duration).
32
Gao, Ozbay 12
RESULTS AND DISCUSSION
1
First, M/M/ queueing model is applied to estimate average travel time with a 15 minute
2
time interval. FIGURE 6 and TABLE 4 provide a summary of the results. The result implies that
3
the M/M/ queueing model has a percentage difference of less than 8% for both sites in terms of
4
average travel times compared with the actual field travel times when no downstream queueing
5
exists. While applying correction factors to the queueing model, the percentage difference is
6
further reduced to less than 4%. In this study, the correction factor is tied to site-specific conditions.
7
However, by introducing it, we intend to introduce the idea of a general factor that will help
8
modelers to capture the impact of factors on traffic as a result of double parking. Many parameters
9
such as traffic volume, number of lanes, GLA of commercial properties may have an effect on the
10
correction factor. Further study and data collections are needed to define it as a general function
11
in terms of all these different characteristics of city streets as part of ongoing efforts.
12
Average field travel time significantly increased due to the downstream blocking of the
13
Brooklyn site in the last 15 minutes of the observed hour (FIGURE 6(a)). Under this circumstance,
14
M/M/ queueing model is found to be incapable of capturing certain spillover conditions which
15
resulted in the difference between the field data and the queueing model output. The reason for
16
this, is that M/M/ queueing model assumes independence of different queues, as well as the
17
independence of the arrival of double parking events, and does not account for spatial queue or
18
queue capacity.
19
20
(a) Brooklyn Site (Congested) (b) Manhattan Site (Uncongested)
21
22
FIGURE 6 M/M/ queueing model results.
23
24
TABLE 4 Summary of Field, Queueing Model, and Microsimulation Result
25
Site
Time
Field TT
Model TT
Model TT with Cf
Microsimulation TT
Brooklyn
8:00-8:15 AM
5.41
5.27
5.39
5.54
8:15-8:30 AM
5.30
5.08
5.20
5.83
8:30-8:45 AM
6.01
6.00
6.14
6.21
8:45-9:00 AM
13.60
4.95
5.07
12.91
Manhattan
8:00-8:15 AM
7.32
6.77
7.21
-
8:15-8:30 AM
8.03
7.78
8.29
-
8:30-8:45 AM
7.38
6.98
7.43
-
8:45-9:00 AM
7.40
6.82
7.26
-
Gao, Ozbay 13
The micro-simulation model is then used to quantify complex characteristics of double
1
parking events such as DP location. Travel time results are shown in the following heat maps
2
(FIGURE 7) for the four scenarios described in the previous section.
3
The first category tested the impact due to different DP locations. DP events at the end of
4
the block and midblock have similar mild effects on the average TT. Events occurring at the
5
beginning of the block have a 5.15% higher impact in terms of the average TT increase over the
6
AM peak hour. This matches the field observation that once a vehicle is double-parked near the
7
beginning of the block (close to the upstream end of the road in the direction of traffic flow),
8
drivers are more likely to slow down. DP vehicle under this condition usually reduces the effective
9
lane width and causes potential conflicts with other activities at the intersection such as pedestrian
10
crossing.
11
The second category of simulation scenarios examined the impact on the travel time of
12
applying 110%, 120%, 130% and 140% of morning peak demand. Generally, higher demand
13
resulted in higher travel time, as expected. The increase of the average travel time is 7.3%, 18.5%
14
and 18.6% for 20%, 30% and 40% increase in the traffic demand, respectively, while 10% increase
15
in demand does not cause a significant change in average travel time.
16
The third category of simulation scenarios accounted for the travel time effect if the
17
frequency of DP event increased. The hourly average travel time increased by 3.1%, 13.6%,
18
20.5%, and 27.6% when there is an increase of DP event frequency by 1, 2, 3, and 4 vehicles per
19
15 minutes, respectively.
20
The fourth category of scenarios specifies various duration time for DP events from 1
21
minute to 10 minutes. As shown in FIGURE 7, the longer the event goes on, the greater the increase
22
in the average travel time. Average travel time can be more than two times higher in the worst-
23
case scenario where the average duration of DP event is 10 minutes.
24
25
26
FIGURE 7 Microsimulation results.
27
Gao, Ozbay 14
CONCLUSIONS AND FUTURE WORK
1
This study examined a macroscopic M/M/ queueing model and micro-simulation for
2
estimating average travel time in the presence of double parking activities. Under uncongested
3
traffic conditions without downstream blocking, applying the M/M/ queueing model produced
4
a good fit with the field data. The observations obtained from video recording were used to develop
5
and calibrate a microsimulation model for one of the two case study areas namely, Downtown
6
Brooklyn. Simulation output shows that the location, frequency, and duration of the double parking
7
event and overall traffic demand have a certain impact on the average travel time.
8
Overall, M/M/ queueing model is an effective approach to compute average travel time
9
for unsaturated traffic flows under double parking conditions. For a large urban network, it has the
10
potential to be implemented instead of micro-simulation models mainly due to its computational
11
efficiency, if it can be combined with an appropriate correction factor. Effect of commercial
12
vehicles is also estimated in the model where one double-parked truck is assumed to account for
13
two DP events. However, when downstream blocking happens, the traffic condition does not
14
completely depend on what is hapenning on the current road segment. Micro-simulation is a more
15
powerful tool than the M/M/ queueing model for evaluating such congested scenarios and can
16
be used to examine individual and combined effects of various explanatory variables.
17
Future research will focus on collecting more field data using automated image processing
18
techniques, investigating specific impacts of double-parked commercial vehicles, exploration of
19
the accurate estimation of the correction factor function, correction factor for various spatio-
20
temporal characteristics of the transportation network, and developing new APIs to improve the
21
micro-simulation model for more challenging geometric and traffic conditions.
22
23
ACKNOWLEDGEMENT
24
The authors appreciate the New York City Department of Finance for providing data
25
through NYC open data (nycopendata.socrata.com). The contents of this paper only reflect views
26
of the authors and do not necessarily reflect the official views or policies of any agencies. The
27
authors would also like to thank Krzysztof Lukasik, a former graduate student of Rutgers
28
intelligent transportation systems (RITS) laboratory for working in the development of an earlier
29
version of the Paramics API plug-in used in this study.
30
31
REFERENCES
32
1. NYCDOF, Official Compilation of Rules of the City of New York, Chapter 39,
33
www1.nyc.gov/site/finance/vehicles/services-violation-codes.page, Accessed Jul. 26, 2015.
34
2. NYCDOT, NYC Parking Regulations, www.nyc.gov/html/dot/html/motorist/parking-
35
regulations.shtml, Accessed Jul. 26, 2015.
36
3. De Cerreño, A., Dynamics of on-street parking in large central cities. In Transportation Research
37
Record: Journal of the Transportation Research Board, No. 1898, 2004. pp. 130-137.
38
4. Rensselaer Polytechnic Institute, New York City Department of Transportation, Rutgers University.
39
Integrative Freight Demand Management in the New York City Metropolitan Area: Implementation
40
Phase. Publication No. RITARS-11-H-RPI.USDOT, 2013.
41
5. Holguin-Veras, J., K. Ozbay, A. Kornhauser, A. Shorris, and S. Ukkusuri. Integrative Freight
42
Demand Management in the New York City Metropolitan Area. Publication No: DTOS59-07-H-
43
0002. USDOT, 2010.
44
6. Baykal-Gursoy, M., W. Xiao, Z. Duan, and K. Ozbay, Delay estimation for traffic flow interrupted
45
by incidents. In: Proceedings of the 86th Annual Transportation Research Conf, 2006.
46
Gao, Ozbay 15
7. Portilla, A., B. Orena, J. Berodia, and F. Diaz, Using M/M/infinity queueing model in on-street
1
parking maneuvres. Journal of Transportation Engineering 135(8), 2009.
2
8. Baykal-Gursoy, M. and W. Xiao, Stochastic decomposition in M/M/∞ queues with Markov
3
modulated service rates. Queueing Systems 48(1-2), 2004, pp. 75-88.
4
9. Guo, H., Z. Gao, X. Yang, X. Zhao, and W. Wang, Modeling travel time under the influence of on-
5
street parking. Journal of Transportation Engineering 138(2), 2011, pp. 229-235.
6
10. Millard-Ball, A., R.R. Weinberger, and R.C. Hampshire, Is the curb 80% full or 20% empty?
7
Assessing the impacts of San Francisco’s parking pricing experiment. Transportation Research Part
8
A: Policy and Practice 63, 2014, pp. 76-92.
9
11. SFMTA, SFPark, sfpark.org, Accessed Jul. 26, 2015.
10
12. NYCDOT, Park Smart, www.nyc.gov/html/dot/html/motorist/parksmart.shtml, Accessed July. 26,
11
2015.
12
13. Schaller, B., D. Stein, and M. Blakeley, Parking Pricing and Curbside Management in New York
13
City. In: Proceedings of the Transportation Research Board 90th Annual Meeting 2011.
14
14. Morillo Carbonell, C. and J.M. Campos Cacheda, On-street illegal parking costs in urban areas.
15
Procedia-Social and Behavioral Sciences 160, 2014, pp. 342-351.
16
15. Kladeftiras, M. and C. Antoniou, Simulation-based assessment of double-parking impacts on traffic
17
and environmental conditions. In Transportation Research Record: Journal of the Transportation
18
Research Board, No. 2390, 2013. pp. 121-130.
19
16. Galatioto, F. and M. Bell, Simulation of Illegal Double Parking: Quantifying the Traffic and
20
Pollutant Impacts. In: Proceedings of the IV International SIIV Congress, 2007. pp. 12.
21
17. Google, Geocoding API, developers.google.com/maps/documentation/geocoding, Accessed July
22
31st, 2015.
23
18. Baykal-Gursoy, M. and Z. Duan, M/M/C queues with Markov modulated service processes. In:
24
Proceedings of the Proceedings of the 1st international conference on Performance evaluation
25
methodolgies and tools, 2006. pp. 38.
26
19. Quadstone, The paramics manuals, version 6.6. 1. Quastone Paramics LTD, Edinburgh, Scotland,
27
UK, 2009.
28
20. Chu, L., H.X. Liu, J.-S. Oh, and W. Recker, A calibration procedure for microscopic traffic
29
simulation. In: Proceedings of the Intelligent Transportation Systems, 2003. pp. 1574-1579.
30
21. Pline, J.L., Traffic engineering handbook, Prentice Hall, 1992.
31
22. Box, P.C., Curb-parking problems: Overview. Journal of Transportation Engineering 130(1),
32
2004, pp. 1-5.
33
34
... Thus, it is almost impossible to track vehicles and to extract travel time or delay information from the videos. To overcome this limitation, we applied an M/M/∞ queueing model as an extension of previous research (Gao and Ozbay 2016) to estimate average link travel time under the appearance of illegal on-street parking vehicles. The advantage of using macroscopic models such as traffic flow models (Amer and Chow 2017) is that they can be an order of magnitude less expensive and faster to develop and implement, compared to microsimulation models. ...
... The advantage of using macroscopic models such as traffic flow models (Amer and Chow 2017) is that they can be an order of magnitude less expensive and faster to develop and implement, compared to microsimulation models. For a comparison of these two types of models, please see (Gao and Ozbay 2016;Gao, Ozbay, and Marsico 2017). ...
... Similarly, if the system has i vehicles without any illegal parking disruptions, it is assumed to be in state ( , ). A transition-rate diagram for M/M/∞ Model with two server states are shown in Figure 4. (Gao and Ozbay 2016) From the transition-rate diagram, it is easy to derive the steady-state balance equations for the state probabilities P i,F and P i,N yielding: ...
Article
Full-text available
The rapid development of the internet of things (IoT), sensing technologies, machine learning and deep learning techniques, along with the growing variety and volume of data, have yielded new perspectives on how novel technologies can be applied to obtain new sources of curb data to achieve cost-effective curb management. This study presents a new computer vision based data acquisition and analytics approach for curb lane monitoring and illegal parking impact assessment. The proposed "rank, detect, and quantify impact" system consists of three main modules, 1) hotspot identification based on rankings generated by local spatial autocorrelation analysis, 2) curb lane occupancy estimation leveraging traffic cameras and computer vision techniques, and 3) illegal parking traffic impact quantification using a M/M/∞ queueing model. To demonstrate the feasibility and validity of the proposed approach, it was tested and empirically validated using field data collected from three case study sites in Midtown Manhattan, New York City (NYC)-one of the most complex urban networks in the world. Different types of curb lane occupancy, including parking and bus lanes, and different frequencies of illegal parking (high, moderate, low frequency) were investigated. Specifically, the proposed method was proven to be effective even for low resolution and discontinuous video feed obtained from publicly available traffic cameras. All three case study sites achieved good detection accuracy (86% to 96%) for parking and bus lane occupancy, and acceptable precision and recall on detecting illegal parking events. The queueing model was also proven to effectively quantify link travel time with the appearance of illegal parking events with different frequencies. The proposed "rank, detect, and quantify impact" system is friendly for large-scale implementation and real-time application. It is also highly scalable and can be easily adopted by other cities to provide transportation agencies with effective data collection and innovative curb space management strategies.
... The striking growth of e-commerce leads to significant freight traffic growth in urban neighborhoods. As freight activities move closer to urban areas, there will be increased competition for space among freight vehicles, passenger vehicles, transit vehicles, pedestrians, and bicyclists (Gao & Ozbay, 2016;Giuliano et al., 2018). Such competition can increase conflicts between VRUs and freight vehicles. ...
Article
Increasing freight traffic has posed greater road safety threats to Vulnerable Road Users (VRUs), including pedestrians and bicyclists, in urban communities. Although there are some disaggregated studies on truck-related crashes, the literature offers limited knowledge of how crash-level factors influence crashes related to urban freight. Our study uses crash data from North Carolina (2007–2019) and Tennessee (2009–2019) to explore the relationship between the crash-level factors and the occurrence and injury severity of crashes involving freight vehicles and VRUs. Crash-level factors include socio-demographics of VRUs and freight vehicle drivers, driving behaviors, temporal and weather effects, and environmental factors. The results indicate that freight-related VRU crashes, compared to nonfreight-related VRU crashes, are more likely to occur at private properties and parking areas and have a higher probability of causing severe injuries or deaths. The results also show that the involvement of larger vehicles and the occurrence at midblock segments are positively associated with more severe injury outcomes. The research results could help community and transport designers to increase attention to the safety impacts of growing freight traffic in urban communities.
... A critical challenge to be overcome when designing and applying simulation models, highlighted by Waraich and Axhausen (2012), is to achieve an adequate level of abstraction, being representative of the drivers' behavior while computationally practical. We assume as relevant behaviors: (a) parking search, also known as cruising, (b) parking choice (Waraich & Axhausen, 2012in Horni et al., 2016Nourinejad et al. 2017), (c) dwell time representation (Gardrat & Serouge, 2016 and (d) impact of obstructions in passing traffic, such as double-parked vehicles (McLeod & Cherrett, 2011) and associated impacts (Alho et al., 2018;Gao & Ozbay, 2016;Kladeftiras & Antoniou, 2013) queue spillovers. The last three are the most relevant as, in LTGs, it is assumed that suitable infrastructures will be concentrated around the building. ...
Article
A growing body of research looks specifically at freight vehicle parking choices for purposes of deliveries to street retail, and choice impacts on travel time/uncertainty, congestion, and emissions. However, little attention was given to large urban freight traffic generators, e.g., shopping malls and commercial buildings with offices and retail. These pose different challenges to manage freight vehicle parking demand, due to the limited parking options. To study these, we propose an agent-based simulation approach which integrates data-driven parking-choice models and a demand/supply simulation model. A case study compares demand management strategies (DMS), influencing parking choices, and their impact in reducing freight vehicle parking externalities, such as traffic congestion. DMS include changes to parking capacity, availability, and pricing as well as services (centralized receiving) and technology-based solutions (directed parking). The case study for a commercial region in Singapore shows DMS can improve travel time, parking costs, emission levels and reducing the queuing. This study contributes with a generalizable method, and to local understanding of technology and policy potential. The latter can be of value for managers of large traffic generators and public authorities as a way to understand to select suitable DMS.
... The data collection was done through field data and video recording. The results showed that the duration of the double parking event and the overall traffic demand have a significant effect on the average travel time [5]. ...
... In the first group of works related to the analysis of the impact of on-street parking on the travel time of a road section, is the article [36]. The authors, A.I. Portilla et al., present the model of vehicle entry to and exit from the parking space as a Poisson process (queue model). ...
Article
Full-text available
The goal of smart cities and sustainable transport is to ensure the efficient movement of people while minimizing a negative impact on the environment. Many cities around the world conduct a policy aimed at limiting parking spaces; however, the complete elimination of parking spaces in cities currently does not seem possible. Parking vehicles cause disturbances in road traffic by searching for a parking space and performing the parking maneuver. This article analyzes the impact of the parking maneuver on the capacity of the inlets of intersections with traffic lights, and the significance of the time it takes to enter and exit a parking space on road traffic disturbance areas under Polish conditions. The research is carried out in on-street parking, characterized by different positions of the parking space to the road, the different surfaces of parking spaces, and the geometry of the road next to which the parking spaces are located. Differences in the time of entry to and exit from the parking space between the research areas indicate that different characteristics of the parking spaces affect the time of the parking maneuver. Drivers wait for the acceptable distance between vehicles on the road into which the vehicle can exit from the parking space. Drivers exiting from perpendicular parking spaces more often cause traffic disruptions than in the case of parallel parking spaces, due to limited visibility. The occupancy of parking spaces directly next to the analyzed ones also affects the time of entry to and exit from the parking space. However, between the time of entry to or exit from the parking space, and the use of the parking space, there is no relationship. This finding indicates that more factors determine the time of entry to and exit from the parking space. The results presented in the article show the need to conduct further research on the impact of parking maneuvers on the capacity of intersections with traffic lights for road traffic conditions in Poland. The results of the research will allow for the design and construction of an optimal parking infrastructure that will meet the needs of road users, while minimizing the negative impact on road conditions and the natural environment following sustainable development.
... In parcel delivery (mostly due to online commerce), vehicles in Amsterdam have been reported to be parked for about 70% of total time spent on their route. Because it is very difficult to find parking space, vehicles are searching for parking space extensively (Cassady and Kobza 1998;Shoup 2006;Dowling, Ratliff, and Zhang 2019;Dalla Chiara and Goodchild 2020) and may be forced to park illegally causing difficulties to other traffic (Kladeftiras andAntoniou 2013, Gao andOzbay 2016). In a field experiment, Fransoo, Cedillo-Campos, and Gamez-Perez (2021) show that by providing more delivery bay space, efficiency improvements in deliveries of up to 40% could be reached. ...
Article
Full-text available
As freight deliveries in cities increase due to retail fragmentation and e-commerce, parking is becoming a more and more relevant part of transportation. In fact, many freight vehicles in cities spend more time parked than they are moving. Moreover, part of the public parking space is shared with passenger vehicles, especially cars. Both arrival processes and parking and delivery processes are stochastic in nature. In order to develop a framework for analysis, we propose a queueing model for an urban parking system consisting of delivery bays and general on-street parking spaces. Freight vehicles may park both in the dedicated bays and in general on-street parking, while passenger vehicles only make use of general on-street parking. Our model allows us to create parsimonious insights into the behavior of a delivery bay parking stretch as part of a limited length of curbside. We are able to find explicit expressions for the relevant performance measures, and formally prove a number of monotonicity results. We further conduct a series of numerical experiments to show more intricate properties that cannot be shown analytically. The model helps us shed light onto the effects of allocating scarce urban curb space to dedicated unloading bays at the expense of general on-street parking. In particular, we show that allocating more space to dedicated delivery bays can also make passenger cars better off.
... Peer research has either developed mathematical models or implemented microscopic simulations to solve the on-street parking problem. A majority of the mathematical models are queuing models including the M/M/∞ model [24], M/M/C model [25], cellular automata models [8], [17], agent-based models [26] and network traffic flow models [27]. The advantage of the mathematical models is the explicit form for repeatability and comparison, without restrictions on software or environmental settings. ...
... A critical challenge to be overcome when designing and applying simulation models, highlighted by Waraich and Axhausen (2012), is to achieve an adequate level of abstraction, being representative of the drivers' behavior while computationally practical. We assume as relevant behaviors: (a) parking search, also known as cruising, (b) parking choice (Waraich & Axhausen, 2012in Horni et al., 2016Nourinejad et al. 2017), (c) dwell time representation (Gardrat & Serouge, 2016 and (d) impact of obstructions in passing traffic, such as double-parked vehicles (McLeod & Cherrett, 2011) and associated impacts (Alho et al., 2018;Gao & Ozbay, 2016;Kladeftiras & Antoniou, 2013) queue spillovers. The last three are the most relevant as, in LTGs, it is assumed that suitable infrastructures will be concentrated around the building. ...
Preprint
The emergence of ride-sourcing platforms has brought an innovative alternative in transportation, radically changed travel behaviors, and suggested new directions for transportation planners and operators. This paper provides an exploratory analysis on the operations of a ride-sourcing service using large-scale data on service performance. Observations over multiple days in Singapore suggest reproducible demand patterns and provide empirical estimates of fleet operations over time and space. During peak periods, we observe significant increases in the service rate along with surge price multipliers. We perform an in-depth analysis of fleet utilization rates and are able to explain daily patterns based on drivers' behavior by involving the number of shifts, shift duration, and shift start and end time choices. We also evaluate metrics of user experience, namely waiting and travel time distribution, and explain our empirical findings with distance metrics from driver trajectory analysis and congestion patterns. Our results of empirical observations on actual service in Singapore can help to understand the spatiotemporal characteristics of ride-sourcing services and provide important insights for transportation planning and operations.
Article
Full-text available
Either a result of inadequate supply or perceived high costs, illegal parking is an often-encountered phenomenon in urban areas worldwide and affects the performance and safety conditions of streets. In Greek cities in particular, illegal parking seems to be mostly observed when legal parking supply is limited, while at the same time, there is available space for such a practice. This paper investigates causes leading to illegal parking, with a focus on space availability, road geometry and the balance between parking demand and supply. In this context, Google StreetView (GSV) images are used as the main data source and multiple linear regression models are developed for investigating factors explaining illegal parking density. Results show that illegal parking is higher in cases where on-street parking is prohibited; intersection and breaking points of median islands are the next favorite places for illegal parking phenomena to occur. Furthermore, major differences in illegal parking behavior are found between two-wheeler and cars or heavy vehicles drivers. Based on findings, a framework consisting of different measures, namely: parking regulations, road design and public awareness, is proposed for mitigating illegal parking in Greek cities.
Article
Travel time reliability is the key aspect that indicates the quality of urban public transit service. The studies on travel time reliability of the public transit system in Indian traffic conditions are few. Also, the impact of side friction elements on travel time reliability has not been considered in the previous studies. Hence, the present study aims to quantify the different types of side friction elements and analyse their impact on the travel time reliability of the public bus transit system. The field data consisting of side friction elements, traffic volume, and travel time of public bus transit have been collected and extracted at two different road sections (divided and undivided) in the Mysore city (Karnataka, India) during weekdays and weekends. The data are grouped into static and dynamic side frictions. An approach has been proposed to represent different types of side friction elements with a single index called the Side Friction Index (SFI) using relative weight analysis. Travel time reliability is represented using measures such as Buffer Time Index (BTI), Planning Time Index (PTI), Travel Time Index (TTI) and Reliable Buffer Index (RBI). The impact of side friction on travel time reliability was found to be sensitive to traffic volume, and hence the thresholds for different traffic volume levels have been determined using K-means clustering method. It was observed from relative weight analysis that the static side friction has a higher weightage (0.509 and 0.327 for the undivided road and divided road respectively) than the dynamic side friction elements in describing the variation of travel time. The impact of side friction on reliability measures at different traffic volume levels has been studied and found to have a non-linear (exponential) relationship. The impact of SFI has been observed to be higher on TTI and PTI in comparison with BTI. The study outcomes show that the impact of side friction on TTI and PTI is sensitive to traffic volume, especially at higher traffic volume level and impact of side friction on BTI is less, especially at medium traffic volume level. The inference from the study shows that the impact of side friction elements varies with respect to the type of road (divided and undivided), traffic volume levels, different days of week (weekday and weekend), and different time periods of day.
Article
Full-text available
Private transport represents more than a third of all the journeys carried out in large cities urban areas. The high amount of traffic in metropolitan zones implies the appearance of congestion that is one of the most important complications transport engineers must fight against. Bottle necks are one of the most common reasons of congestion effect and on-street illegal parking (double parked or driving lane parked) implies the creation of a (or successive) bottle neck. This paper is focused on to evaluate the costs that appear when on-street illegal parking is detected. To achieve the paper's goal, first of all, the different costs of the journey have been defined and formulated; next, the effect of illegal parking has been considered and the cost related to that has been formulated for each of the different types of cost. In order to use the formulation obtained two real scenarios have been evaluated. The results obtained in real scenarios have demonstrated that the economic cost due to car parking indiscipline to the whole city is significant; also it is worth to remark that unit cost (cost provoked for a single vehicle) differs depending on the location of the violation and the affected vehicle. On-street illegal parking cost is significantly outstanding. According to data, on street illegal parking reduction has to be one basic pillar of mobility policies. This reduction would allow a higher road capacity and, consequently, greater traffic fluidity
Article
Full-text available
On-street parking is an important component of the parking system. Because of its occupancy of roadway resources, it can significantly impact traffic performance and safety. The aim of this paper is to give a quantitative analysis of the influence of on-street parking on travel time. The travel time data of the motor vehicles moving in the road sections with on-street parking are gathered by observers. A proportional hazard-based duration model is proposed to analyze the influential factors related to on-street parking, including effective lane width, the number of parking maneuvers, and occupancy. The results show that on-street parking has a significant impact on the travel time of vehicles. In addition, various factors can modify travel time in different degrees, and the model can be used to estimate the travel time under assumed conditions. It is hoped that this paper will help to improve the planning and management of on-street parking.
Article
Full-text available
Motivated by the need to study transportation systems in which incidents cause traffic to slow down, we consider an M/M/8 queueing system subject to random interruptions of exponentially distributed durations. System breakdowns, where none of the servers work, as well as partial failures, where all servers work with lower efficiency, are investigated. In both cases, it is shown that the number of customers present in the system in equilibrium is the sum of two independent random variables. One of these is the number of customers present in an ordinary M/M/8 queue without interruptions.
Conference Paper
Full-text available
Motivated by the need to study trac o w aected by incidents we consider M/M/C queue- ing system where servers operate in a Markovian environment. When a trac incident happens, either all lanes or part of a lane is closed to the trac. As such, we model these interruptions either as complete service disruptions where none of the servers work or partial failures where all servers work at some reduced service rate. We analyze the system with multiple failure states in steady state and present a scheme to obtain the stationary number of vehicles on a link. The special case of single breakdown case is further analyzed and performance measures in closed form are obtained.
Article
This paper quantifies the influence badly parked vehicles and on-street parking maneuvers have on average link journey times as a function of the duration of the events and the number of designated maneuvers and flow, by applying an M/M/infinity queueing model in which arrival and departure are all Poisson processes. The method has been validated using microsimulations calibrated by in-situ measurements taken in the streets of the city of Santander (Cantabria, Spain). The analysis on the delays shows a good fit for the M/M/infinity model for flows of 60-70% of capacity where the error is always lower than 5%. This demonstrates the efficiency of the M/M/infinity model for studying how on-street parking maneuvers and badly parked vehicles influence traffic flow and avoids the need to use generally more laborious microsimulation models. Microsimulations are used to calculate the reduction in link capacity for each case in the study and the increases in average journey times for the rest of the road users. This shows the effect that allowing on-street parking on arterial or main roads has on the rest of the traffic.
Article
The objective of this research was to estimate the effects that a specific pattern of illegal parking (double-parking) had on traffic conditions and the environment by using microscopic simulation. Through a sensitivity analysis, the effects of illegal double-parking on average speed, delay, and stopped time were estimated. Results showed that the existence of the phenomenon entailed a severe decrease in average speed and an important increase in delay and stopped time. Through a case study, the effects that a reduction or an elimination of the phenomenon would have in a real network were evaluated: all traffic indicators would be improved if double-parking were suppressed partially (e.g., through an intensification of enforcement) or, even better, completely. Results showed that limiting double-parking could result in an increase in speeds of about 10% to 15% and a decrease of about 15% and 20% in delay and stopped time, respectively. However, even greater improvements may be achieved if double-parking is eliminated: average speed can increase by up to 44%, while delay and stopped time can decrease by up to 33% and 47%, respectively. On the basis of the results extracted from the case study, the effects for the whole district of the municipality of Athens, as well as for the entire Athens region, were assessed by using several indexes and performance measures. Savings from decreased lost time and reduced carbon dioxide, carbon monoxide, and hydrocarbon emissions were calculated. Directions for future research are also proposed.
Article
Funded by the Federal Highway Administration, the purpose of this report is three-fold: (1) to determine, to the degree possible, the impact that on-street parking has on transportation, development, and land-use; (2) to identify and review comprehensively "on-street" parking policies and management practices in large cities; and, (3) to recommend best practice strategies for on-street parking in large cities. The report is the culmination of a year-long study, which included an extensive literature review, one-on-one discussions with city parking officials, a peer-to-peer exchange session in Boston, and a detailed questionnaire to which nine U.S. cities responded.
Official Compilation of Rules of the City of New York, Chapter 39, 33 www1.nyc.gov/site/finance/vehicles/services-violation-codes.page
NYCDOF, Official Compilation of Rules of the City of New York, Chapter 39, 33 www1.nyc.gov/site/finance/vehicles/services-violation-codes.page, Accessed Jul. 26, 2015. 34
Publication No. RITARS-11-H-RPI
  • Phase
Phase. Publication No. RITARS-11-H-RPI.USDOT, 2013. 41
Integrative Freight 42 Demand Management in the New York City Metropolitan Area
  • J Holguin-Veras
  • K Ozbay
  • A Kornhauser
  • A Shorris
  • S Ukkusuri
Holguin-Veras, J., K. Ozbay, A. Kornhauser, A. Shorris, and S. Ukkusuri. Integrative Freight 42 Demand Management in the New York City Metropolitan Area. Publication No: DTOS59-07-H43 0002. USDOT, 2010. 44