safety

Article
Road Safety Analysis of Urban Roads: Case Study of
an Italian Municipality
Francesca Demasi, Giuseppe Loprencipe and Laura Moretti *
Department of Civil, Construction and Environmental Engineering, Sapienza University of Rome, 00184 Rome,
Italy; [email protected] (F.D.); [email protected] (G.L.)
* Correspondence: [email protected]; Tel.: +39-06-4458-5114




Received: 4 October 2018; Accepted: 14 November 2018; Published: 1 December 2018


Abstract: Attention to the most vulnerable road users has grown rapidly over recent decades.
The experience gained reveals an important number of fatalities due to accidents in urban branch
roads. In this study, an analytical methodology for the calculation of urban branch road safety is
proposed. The proposal relies on data collected during road safety inspections; therefore, it can be
implemented even when historical data about traffic volume or accidents are not available. It permits
us to identify geometric, physical, functional, and transport-related defects, and elements which
are causal factors of road accidents, in order to assess the risk of death or serious injuries for users.
Traffic volume, average speed, and expected consequences on vulnerable road users in case of an
accident allow us to calculate both the level of danger of each homogeneous section which composes
the road, and the hazard index of the overall branch. A case study is presented to implement the
proposed methodology. The strategy proposed by the authors could have a significant impact on the
risk management of urban roads, and could be used in decision-making processes to design safer
roads and improve the safety of existing roads.

Keywords: road safety; road defects; urban branch road; index risk; road safety inspection

1. Introduction
User safety is one of the most important issues in road design. The Organization for Economic
Co-operation and Development (OECD) demonstrated that in the last two decades, the trend of
road accidents is not positive all over the world [1]. The problem is serious, as confirmed by the
attention given to it at international level [2]. The United Nations General Assembly declared the years
2011–2020 as a Decade of Action for Road Safety [3]. At the European level, the trend of road fatalities
since 2001 appears not to be compliant with the Road Safety Program pursued by the European
Commission, which aims at halving road casualties by 2020 [4].
A non-negligible rate of road accidents (73% in Italy according to [5]) occurs in urban areas, where
externalities of motorized traffic (e.g., congestion, noise, and pollution) encourage the use of “soft
mobility” [6,7]. In such conditions, different users and different vehicles share urban road spaces [8,9].
In the literature, several studies demonstrated that pedestrians, cyclists, and motorcyclists are the most
vulnerable users in urban areas [10–12].
In order to prevent accidents and reduce their consequences on people, the system composed of
drivers, vehicles, the environment, and roads should be investigated [8]. According to several studies
available in the literature, most accidents are caused by bad behavior of drivers, i.e., the only ones
capable of adapting their behavior towards non-living components [13,14]. However, it has been
observed that the incorrect design or management of road infrastructure can induce bad behavior
or failure to appreciate risk [15–18]. Therefore, in order to maximize on-going safety efforts, it is
necessary to analyze each component of the road system and evaluate its effect on road safety.

Safety 2018, 4, 58; doi:10.3390/safety4040058 www.mdpi.com/journal/safety

Safety 2018, 4, 58 2 of 16

To this end, the European Directive 2008/96/EC [19] on the safety management of road infrastructure
establishes management procedures ensuring the safety of road networks. It defines four basic tools
for road safety management: Road Safety Impact Assessment, Road Safety Audit, Safety Ranking and
Management of the Road Network in Operation, and Road Safety Inspections. These activities work
on planning, design, and management of both constructed and to-be-constructed roads included in
the Trans-European Transport Network (TEN-T). Some Member States transposed the Directive as
a code of good practice also for their national roads which are not included in TEN-T. For example,
in Italy, the transposition acts [20,21] advised authorities to implement a road infrastructure safety
management network composed of four phases: analysis, inspection, classification, and intervention.
Particularly, Road Safety Inspections (RSI) should be carried out on existing roads in order to detect
infrastructure elements/defects which could increase the rates of traffic conflicts and exacerbate their
negative consequences.
Moreover, interesting and effective actions have been developed to promote tools and procedures
in EU member states and in other countries [22–25].
The European Road Assessment Program (EuroRAP) has been developed by European road
authorities to assess the risk to car occupants and highlight sections with the highest risk of death and
serious injury [26]. It is based on objective variables: crash types and seriousness, distance and type of
roadside obstacles, speed, and traffic level of the road.
In the USA, the AASHTO Highway Safety Manual (HSM) provides standardized and accurate
methods for the estimation of both crash frequency or severity for urban roads and the effects of road
safety measures [27,28].
The New Zealand Transport Agency developed the procedure Road Infrastructure Safety
Assessment to monitor road safety performances. Cross section, alignment, surface conditions, access
ways, and statistical data on crashes are the examined variables [29]. Moreover, in New Zealand,
procedures for Safety Audit of Existing Roads (SAER) allowed for the development of an empirical
measure of safety according to a subjective scale depending on the relative hazard frequency and the
crash severity probability [30].
In recent years, much research has analyzed the results from statistical data and RSI to evaluate
safety condition towards physical components of road-system [31–33]. The proposed methods permits
us to assess, at section and branch level, the current level of safety for the most vulnerable users,
to identify the most critical conditions, and to implement correct strategic decisions [34]. The authors
propose an original analytical model which is useful to assess the risk of road accidents in existing
urban roads. The model relies on data collected during RSIs (e.g., geometric characteristics, road
signs, and urban furniture, users’ categories); therefore, it can be implemented even when historical
data about traffic volume or accidents are not available. The value of applying this technique in
the risk assessment of a specific urban branch road lies in the possibility of identifying the critical
defects related to potential accidents which cause deaths or serious injury. This, in turn, provides the
information required by the road management body to adopt a strategy in order to adopt mitigation
strategies and protect citizens’ well-being. The proposed method allows for an objective assessment of
the risk level, and the numerical results of the analysis are comparable to a target level of risk to be
defined during the analysis. Moreover, the method permits authorities to define guidelines for road
design with the aim of creating safer roads for all types of users.

2. Methods
In this study, the authors proposed an innovative quantitative approach to assess the Branch Index
Risk (BIR) and the Section Index Risk (SIR) of existing urban roads considering their geometry, layout,
users, and traffic. BIR assesses the overall risk of road accidents in urban branch roads: the higher
BIR value, the lower the safety along the examined infrastructure stretch. Therefore, the value of BIR
depends on the Section Factor Risk (SFRj,r ) of each homogeneous section j which composes the branch r.
Each homogeneous section is 100 m ± 20%-long: road branches are considered homogeneous if they
Safety 2018, 4, 58 3 of 16

have uniform/homogeneous attributes related to physical and operating conditions (i.e., accident
rate, geometric layout, composition of cross section, traffic spectrum, average operating speed) [35].
This approach allows for the comparison of “density” of hazardous elements/defects between different
(short) sections whose length is suited to the considered urban context. Moreover, it gives good results
in terms of fitting the safety performance of road sections, as confirmed by Cafiso et al. [36].
SFRj,r takes into account both the general characteristics of r (i.e., type and frequency of hazardous
elements/defects, expected damage on vulnerable road users, and traffic level), and all n elements
which affect the dangerousness of j. Equation (1), which complies with an approach recently proposed
to analyze and plan maintenance of road safety barriers [32], allows us to calculate SFRj,r :

SFRj,r = Σ n i=1 Bi × K1i × K2i × K3 × K4i × K5i (1)

Bi is the base value associated to defects i which are along j.

K1i is the priority factor of the category to which the element i belongs.
K2i is the vulnerability factor of users (i.e., pedestrians, cyclists and motorcyclists) along r; and it
depends on their volume.
K3 is the motorized traffic factor of r.
K4i is the hazardousness factor; it depends on the consequences of defect i on the most vulnerable
road users.
K5i is the extension factor; it depends on continuous or discrete elements/defects i along j.

Given SFRj,r , the corresponding value of SIR can be calculated according to Equation (2)

SIRj,r = SFRj,r /SFRmax,r × 100 (2)

where SIRj,r is the Section Index Risk of the section j belonging to r, and SFRmax,r is the highest value of
SFRr by attributing the maximum values to defects found on r.
Therefore, SFRj,r depends on the detected road elements/defects and the road users of the section
j, and SIRj,r depends on the comparison between the attributed and the maximum values of K1i , K2i ,
K3 , K4i , and K5i .
Similarly, the Branch Index Risk BIRr can be calculated according to Equation (3):

BIRr = Rr /Rmax,r × 100 (3)

where Rr is the sum of the SIRj,r of m sections which compose the road branch r given by Equation (4)

Rr = Σ m j=1 SIRj,r (4)

where Rmax,r is the reference value of the Risk Factor given by Equation (5)

Rmax,r = m × SFRmax,r (5)

Similar to SIRj,r , BIRr depends on the attributed values of K1i , K2i , K3 , K4i , and K5i , and on the
defects found along r. Both SIRj,r and BIRr range between 0 and 1.
The implementation of the proposed method requires road inspections to identify and categorize
elements and/or defects of infrastructure which could cause accidents. The authors identified
9 categories of elements/defects: geometry (G), cross-section (C); private access (A); pavement (P);
lighting (L); road signs (S); intersection (J); urban furniture (F); and stopping (ST). The attribution
of possible values of K1i , K2i , K4i , and K5i , and Bi required interviewing technicians from different
backgrounds, as well as academic experts in the fields of roads (both geometry and safety), urban
planning, transport management, and human health. Eight road engineers, eight urbanists, seven
traffic managers, and six traumatologists were interviewed. The authors defined, for each variable,
Safety 2018, 4, 58 4 of 16

the maximum and minimum value, then each technician attributed the values within the established
ranges. Finally, the geometric mean has been used to aggregate individual judgements.
Table 1 lists the priority factors K1 for each category.

Table 1. Priority factor values of considered elements/defects.

Category Code K1
Geometry G 0.9
Cross-section C 1.0
Private access A 0.9
Pavement P 0.8
Lighting L 0.5
Road signs S 0.7
Intersection J 1.0
Urban furniture F 0.6
Stopping ST 0.4

Taken together, the authors identified 55 road elements/defects which can cause accidents. Their
base values Bi satisfy Equation (6) and are listed in Table 2. Road layout and geometry were evaluated
against the Italian standards and recommendations about the geometric and functional characteristics
of roads [37] and intersections [38], road infrastructure safety management [21], and road lighting [39].
These standards were considered as references to identify “not compliant” conditions, and to assign
the Bi values.
1 ≤ Bi ≤ 4 (6)

K2i depends on the observed traffic condition of vulnerable users (i.e., pedestrians, cyclists, and
motorcyclists) which are exposed to the risk induced by defect i. The Highway Capacity Manual [40]
suggests performing surveys during a 15 min count period in order to obtain the traffic volume data.
The authors observed the traffic flow when weather conditions were not an obstacle to traffic, and
when all work- and school-related activities were ongoing. Finally, they calculated K2i according to
Equation (7).
K2i = KPi × KCi × KMi (7)

where KPi , KCi , and KMi depend on the pedestrian, cyclist, and motorcyclist flow, respectively. Their
values satisfy Equations (8)–(10), and depend on the traffic volume observed along the overall examined
branches, if the analysis involves a road network.

1 ≤ KPi ≤ 2.5 (8)

1 ≤ KCi ≤ 2.5 (9)

1 ≤ KMi ≤ 2.5 (10)

Particularly, having considered all the examined branches, the average hourly flow of cyclists and
motorcyclists (ACF and AMF, respectively), and their standard deviation (DCF and DMF, respectively)
are calculated. Finally, these values are compared to the hourly flow of cyclists and motorcyclists (ACFr
and AMFr , respectively) and the standard deviations (DCFr and DMFr , respectively) of the examined
branch r in order to calculate KCi and KMi , respectively (Table 3).
Safety 2018, 4, 58 5 of 16

Table 2. Base values of elements/defects.

Description Code Bi
Geometry: low planimetric radius; not enough stopping sight distance G1 3
Geometry: steep longitudinal slope G2 3
Carriageway: narrow lane width C1 1
Carriageway: missing shoulder C2 2
Carriageway: narrow shoulder width C3 1
Carriageway: median not appropriate C4 2
Carriageway: missing sidewalk C5 4
Carriageway: narrow sidewalk (less than 0.5 m-wide) C6 2
Carriageway: irregular sidewalk pavement C7 1
Carriageway: distressed sidewalk pavement induces pedestrians to use the carriageway C8 4
Carriageway: missing waiting area at bus stops C9 4
Carriageway: narrow waiting area at bus stops C10 2
Carriageway: missing pedestrian connections to and from the bus stop C11 3
Carriageway: missing accessible connections to and from the bus stop C12 2
Carriageway: pedestrian crossing without ramps C13 2
Carriageway: ramps not aligned to pedestrian crossing C14 2
Carriageway: missing pedestrian crossing C15 4
Carriageway: more than 12 m-long crossing C16 4
Carriageway: distressed pavement at pedestrian crossing C17 2
Carriageway: lack of visibility (both at pedestrian crossing or not) C18 4
Carriageway: missing tactile paths C19 2
Carriageway: not appropriate tactile paths C20 1
Carriageway: missing cycle path C21 3
Carriageway: narrow cycle path C22 2
Private access: not compliant position A1 2
Private access: lack of visibility A2 3
Private access: lack of ramps A3 1
Pavement: irregular pavement P1 3
Pavement: joint faulting and drainage distress P2 2
Pavement: tram crossing P3 2
Pavement: inefficient drainage system P4 3
Pavement: missing drainage system P5 2
Lighting: missing lighting system L1 4
Lighting: inefficient drainage system L2 2
Road signs: missing horizontal road markings S1 4
Road signs: not complete horizontal road markings S2 3
Road signs: not compliant horizontal road markings S3 3
Road signs: missing roadside signs S4 4
Road signs: not visible roadside signs S5 3
Road signs: not complete roadside signs S6 2
Road signs: not compliant roadside signs S7 3
Road signs: missing traffic light S8 4
Road signs: wrong phases of traffic light S9 3
Road signs: wrong installation of traffic light S10 2
Road signs: not visible traffic light S11 2
Intersection: lack of visibility J1 4
Intersection: missing reserved lane J2 3
Intersection: hazardous maneuvers J3 3
Intersection: missing weaving section J4 3
Urban furniture: not compliant safety barriers F1 3
Urban furniture: missing safety barriers F2 4
Urban furniture: urban furniture forces pedestrians to use carriageway F3 2
Urban furniture: lack of visibility at intersection due to urban furniture F4 4
Urban furniture: urban furniture occupies shoulder and/or lane F5 3
Urban furniture: missing maintaining greenery at roadside F6 1
Stopping: illegal parking ST1 1
Safety 2018, 4, 58 6 of 16

Table 3. Values of the vulnerability factor K2i .

Users Observed Condition Traffic Flow Condition Code K2 Values

Pedestrians not exposed to defect i - 1.0
Not more than 5 commercial activities and/or any
Low KPi 1.5
bus stops are in the examined section.
Not more than 10 commercial activities and/or not
Pedestrians Medium 2.0
more than 2 bus stops are in the examined section
The examined section is very attractive: more than
10 commercial activities and 2 bus stops, schools, High 2.5
public offices, or collective spaces are.
Cyclists are not exposed to defect i. Cyclists do not
- 1.0
use or are not admitted
Cyclists ACFr < ACF − DCF Low KCi 1.5
ACF − DCF ≤ ACFr < ACF + DCF Medium 2.0
ACFr > ACF + DCF High 2.5
Motorcyclists are not exposed to defect i.
- 1.0
Motorcyclists do not use or are not admitted
Motorcyclists AMFr < AMF − DMF Low KMi 1.5
AMF − DMF ≤ AMFr < AMF + DMF Medium 2.0
AMFr > AMF + DMF High 2.5

The motorized traffic factor K3 depends on the observed volume of motorized vehicles along the
surveyed stretch. According to [41–44], the authors assumed that the higher their average speed, the
greater the danger of exposure for vulnerable users. Therefore, the values of K3 in Table 4 depend
on the typical observed traffic fluidity and the number of daylight hours with congested traffic flow.
The term “congested flow” complies with the methodology proposed by [45] to evaluate the level of
service of an urban arterial: this condition occurs when the percentage speed reduction from free flow
speed appears greater than 50%.

Table 4. Values of the motorized traffic factor K3 .

Daylight Hours with Congested Traffic Flow (h)

Typical Traffic Flow >5 3–5 <3
K3
Stable flow 2.0 2.5 2.5
Approaching unstable flow 1.5 2.0 2.5
Unstable flow/Stop&go 1.5 1.5 2

The hazardousness factor K4i depends on the expected fatality induced by the defect i on
vulnerable road users. It is calculated according to Equation (11)

K4i = K4Vi × K4Pi (11)

where K4Vi and K4Pi refer to expected consequences on motorized vehicle users (both drivers and
passengers), and non-motorized users (pedestrians and cyclists), respectively. Their values satisfy
Equations (12) and (13), and are listed in Table 5.

1 ≤ K4Vi ≤ 2.5 (12)

1 ≤ K4Pi ≤ 5.0 (13)

Safety 2018, 4, 58 7 of 16

Table 5. Values of the hazardousness factor K4i .

Road users Condition Hazardousness Code K4 Values

Motorized vehicle users are not exposed to defect i - 1.0
Motorized vehicle users are not damaged by defect i,
Motorized vehicle Low K4Vi 1.5
or crashes occur at not more than 40 km/h
users
Crashes occur at 40–60 km/h Medium 2.0
Crashes occur at more than 60 km/h High 2.5
Non-motorized vulnerable users not exposed to defect i - 1.0
Non-motorized vulnerable users are not damaged by
defect i, or crashes with motorized vehicles occur at not Low K4Pi 2.0
Non-motorized
more than 30 km/h
users
Cashes occur at 30–40 km/h Medium 3.0
Cashes occur at 40–50 km/h High 4.0
Cashes occur at more than 50 km/h Critical 5.0

The established ranges for K4Vi and K4Pi take into account that at the same speed, consequences
on vulnerable users are higher than on users of motorized vehicles. On the road, vulnerable users
conflict with vehicles with larger dimensions and masses; cyclists and pedestrians clearly constitute
the “unprotected” element, and they are more exposed in the event of an accident, as confirmed by [46].
Indeed, Wramborg defined probability curves to represent the fatality risk versus collision speed
when a motorized vehicle collides with a vulnerable user, or frontal or hard object collision occurs to
motorized vehicles.
K5i is the extension factor of each element i found along j. According to Table 6, it considers both
surveyed continuous and discrete elements/defects.

Table 6. Values of the extension factor K5i .

Type of
Condition Level of Extension K5 Values
Elements/Defects
Element/defect i is along less than one-third of section length Low 1.0
Continuous Element/defect i is along more than one-third and less than two-thirds
Medium 1.5
of section length
Element/defect i is along more than two-thirds of section length High 2.0
1 element/defect i is along the examined section Low 1.5
Discrete 2 elements/defects i are along the examined section Medium 2.0
More than 2 elements/defects i are along the examined section High 2.5

Given the above procedures, it is possible to simulate values of BIRs and define appropriate
classes of risk. To this end, the authors considered six probabilistic classes of risk level, as usually
done for transport infrastructure risk assessment [47–53]. The definition of ranges for each class
requires a significant number of monitored branches. Monte Carlo simulations were used to generate
a distribution of simulated BIRs obtained from randomly-assigned values of the proposed variables.
Therefore, the Monte Carlo technique gave a virtual sample useful to study the statistical variability
of BIR, assuming that the virtual sample was comparable to the real (surveyed) one. According
to the Central Limit Theorem [54], as the number of samples from any population increases, the
probability distribution of the means will approach a normal distribution. Therefore, it was possible to
compare the distribution of simulated BIRs to a Gaussian distribution. Figure 1 shows the results from
3000 simulations, and it compares the simulated and analytical frequency curves: the former derives
from the Monte Carlo simulation, the latter represents the Gaussian curve [55].
Figure 2 compares the cumulative frequency curves of the simulated and analytical distributions:
their trend confirms the appropriateness of the procedure.
Safety 2018, 4, x 8 of 16

Safety 2018, 4, 58 8 of 16
Safety 2018, 4, x 8 of 16
0.05

simulated results
0.04
Gaussian curve
Relative frequency 0.05
0.03
simulated results
0.04
Gaussian curve
Relative frequency

0.02
0.03
0.01
0.02
0
0.070
0.090
0.110
0.130
0.150
0.170
0.190
0.210
0.230
0.250
0.270
0.290
0.310
0.330
0.350
0.370
0.390
0.410
0.430
0.450
0.470
0.490
0.510
0.530
0.550
0.01

0 BIR
0.070
0.090
0.110
0.130
0.150
0.170
0.190
0.210
0.230
0.250
0.270
0.290
0.310
0.330
0.350
0.370
0.390
0.410
0.430
0.450
0.470
0.490
0.510
0.530
0.550
Figure 1. Comparison between simulated and analytical relative frequency curves.
BIR
Figure 2 compares the cumulative frequency curves of the simulated and analytical
distributions:Figure
their trend confirmsbetween
1. Comparison
Comparison the appropriateness of the procedure.
simulated and analytical relative frequency curves.

Figure 2 compares the cumulative frequency curves of the simulated and analytical
distributions: their trend confirms the appropriateness of the procedure.
1
0.9
0.8
1
frequencyfrequency

0.7
0.9
0.6
0.8 Simulated curve
0.5 Gaussian curve
0.7
Cumulative

0.4
0.6
0.3 Simulated curve
0.5 Gaussian curve
Cumulative

0.2
0.4
0.1
0.3
0
0.2
0.095 0.195 0.295 0.395 0.495 0.595
0.1
BIR
0
0.095
Figure Comparison0.195
2. Comparison 0.295and analytical
between simulated 0.395 0.495
cumulative frequency 0.595
curves.
BIR
Table 77 lists
liststhethe
six six classes
classes of BIR, BIR, minimum
of their their minimum and maximum
and maximum values,
values, and and
chromatic
chromatic categorization.
categorization.
Figure 2. Comparison between simulated and analytical cumulative frequency curves.

Table 7 lists the six classes of BIR, 7. Classes

their
Table of risk.and maximum values, and chromatic
minimum
categorization. BIR (%) BIR (%)
Class Class Risk Level
Risk Level Criterion
Criterion Chromatic Categorization
Chromatic Categorization
Minimum
Minimum Maximum
Maximum
I Not relevant I < μ − 2σ Table 7. Classes
0 of risk.
14.5 White
I Not relevant I < µ − 2σ 0 14.5 White
II II Low Low μ − 2σ < I <
µ − 2σ < I < µ − σμ − σ >14.5
>14.5 21.2 21.2 Green Green
BIR (%)
III Class Risk
IIIModerate Level
Moderate µμ−Criterion
−σσ<< II<< µμ >21.2
>21.2 28.0 28.0 ChromaticYellow
Categorization
Yellow
Minimum Maximum
IV IV High High µμ<<II<<µμ++2σ 2σ >28.0
>28.0 34.8 34.8 Orange Orange
V I Not relevant µ + 2σI << μI <− µ
2σ 0 >34.8 14.5 White Red
VVery high
Very high μ + 2σ < I < μ++2σ 2σ >34.8 41.5 41.5 Red
VI II CriticalLow μ −I 2σ> µ<+I 2σ
<μ−σ >14.5
>41.5 21.2 100 GreenBurgundy
VI Critical I > μ + 2σ >41.5 100 Burgundy
III Moderate μ−σ<I<μ >21.2 28.0 Yellow
IV High μ < I < μ + 2σ >28.0 34.8 Orange
V Very high μ + 2σ < I < μ + 2σ >34.8 41.5 Red
VI Critical I > μ + 2σ >41.5 100 Burgundy
Safety 2018, 4, x 9 of 16
Safety 2018, 4, 58 9 of 16

The proposed method could be applied by policy and decision makers when they have to
manage
The urban
proposedroad safety could
method because it provides
be applied resultsand
by policy that could be
decision used when
makers to critically approach
they have this
to manage
strategic
urban sector
road whose
safety impacts
because are economic,
it provides social,
results that couldand environmental.
be used to critically Indeed,
approach input data are
this strategic
available to a road management body, and the presented method is comprehensive
sector whose impacts are economic, social, and environmental. Indeed, input data are available to aand versatile;
therefore,
road it may be
management applied
body, to different
and the presentedurban
methodscenarios varying theand
is comprehensive examined variables
versatile; andit their
therefore, may
factors.
be applied to different urban scenarios varying the examined variables and their factors.

3. Case Study
The proposed methodology has been applied on multiple branches totaling 50 km together in
an Italian municipality in order to assess their BIR values. All the roads had the same classification:
parking spaces and sidewalks
two-lane urban roads with parking sidewalks on both
both sides
sides (Figure
(Figure 3).
3). Their
Their maximum
maximum
was 50
allowable speed was 50 km/h.
km/h.

Figure 3. Cross-section
Figure 3. Cross-section of
of the
the examined
examined urban
urban roads.
roads.

For
For the
the sake
sake of brevity, in
of brevity, in this
this study,
study, the
the authors
authors present
present the
the most
most interesting
interesting and critical roads,
and critical roads,
where users have to perform crucial and dangerous maneuvers (e.g., intersections
where users have to perform crucial and dangerous maneuvers (e.g., intersections without traffic without traffic
lights,
lights, poor
poorvisibility, many
visibility, conflict
many points).points).
conflict Table 8 summarizes geometric and
Table 8 summarizes technical and
geometric characteristics
technical
of the selected roads: they are 7 branches (i.e., ROAD1 to ROAD7) with similar
characteristics of the selected roads: they are 7 branches (i.e., ROAD1 to ROAD7) with similar geometric and
traffic characteristics.
geometric and traffic characteristics.
Table 8. Characteristics of the selected roads.
Table 8. Characteristics of the selected roads.
Most Frequently
Branch Total
Total Length (m) Number of Sections Minimum SIR (%)Maximum
Length Maximum SIRSIR (%)Most Frequently Detected
Branch Number of Sections Minimum SIR (%) Detected Base Values
(m) (%) Base Values
ROAD1 1700 17 32.3 59.8 2
ROAD1
ROAD2
1700 1400
17 14
32.3 8.1
59.8 61.7
23
ROAD2
ROAD3 1400 550 14 5 8.1 18.5 61.7 31.1 33
ROAD3
ROAD4 550 2000 5 20 18.5 15.0 31.1 30.0 32
ROAD5
ROAD4 2000 650 20 6 15.0 4.8 30.0 26.2 23
ROAD6
ROAD5 650 2450 6 25 4.8 6.1 26.2 30.6 33
ROAD7
ROAD6 2450 1400 25 14 6.1 0.0 30.6 31.21 33
ROAD7 1400 14 0.0 31.21 3

Figure 4a to f show some of the most frequently-detected defects: irregular pavement, narrow
Figureinefficient
sidewalk, 4a to f show somesystem,
lighting of the most
urbanfrequently-detected
furniture occupiesdefects: irregular
shoulder and/orpavement, narrow
lane, pedestrian
sidewalk,without
crossing inefficient lighting
ramps, system,
and not visibleurban
traffic furniture occupies shoulder and/or lane, pedestrian
light, respectively.
crossing without ramps, and not visible traffic light, respectively.
Safety 2018, 4, 58 10 of 16
Safety 2018, 4, x 10 of 16

(a) (b)

(c) (d)

(e) (f)
Figure
Figure 4. 4.
MostMost frequentlydetected
frequently detecteddistresses.
distresses.(a)
(a) Irregular
Irregular pavement
pavementP1 P1(B
(Bi i==3);3);(b) Narrow
(b) Narrow sidewalk
sidewalk
C6C6(Bi (B = 2);
= i2); (c) (c) Inefficient
Inefficient lighting
lighting system
system L2 L2
(Bi (B = 2);
= i2); (d) (d) Urban
Urban furniture
furniture occupies
occupies shoulder
shoulder and/or
and/or lane
lane
F5(B i = F5(B
3); i = Pedestrian
(e) 3); (e) Pedestrian crossing
crossing without
without ramps ramps
C13 C13
(B i =(B i =(f)
2); 2); (f)
Not Not visible
visible traffic
traffic light
light S11S11
(B i
(B
= =
i 2).

2).
Safety 2018, 4, 58 11 of 16
Safety 2018, 4, x 11 of 16
Safety 2018, 4, x 11 of 16
The
Thebar
bargraph in Figure
graph in 5 represents, in descending BIRs (i.e.,
order,order, blue bars) tobars)
bettertocompare
them The
to thebar
accident in Figure
graphdensity
Figure
(AD)
55 represents,
represents,
(i.e., orange
in
in descending
bars)descending
retrieved byorder,
the
BIR
BIRss (i.e.,
authors (i.e.,
from
blue
blue
the bars)
Italian
better
tostatistical
better
compare them to
compare themdatabase the accident
to the accident density
density (AD) (i.e.,
(AD) (i.e., orange
orange bars) retrieved by the authors from the
bars) retrieved by the authors from the
geo-referenced
Italian of occurred accidents [56].
Italian statistical
statistical geo-referenced
geo-referenced database
database ofof occurred
occurred accidents
accidents [56].
[56].

100 45
100 45
90 40
90 40
80 35
80 35
(accident/km/year)
AD(accident/km/year)

70
70 30
30
60
60 25

BIR%)%)
50 25
50 20

BIR
40 20
40
15
30 15
30
AD

20 10
20 10
10 5
10 5
0 0
0 0
ROAD 1 ROAD 2 ROAD 3 ROAD 4 ROAD 5 ROAD 6 ROAD 7
ROAD 1 ROAD 2 ROAD 3 ROAD 4 ROAD 5 ROAD 6 ROAD 7
BIR (%) AD (accident/km/year) Critical Very high High Moderate Low
BIR (%) AD (accident/km/year) Critical Very high High Moderate Low

Figure5.5.Comparison
Comparison of BIR
BIR values and
and accident
accident density
densityfor
fordifferent
differentroads.
Figure 5. Comparison of
Figure of BIR values
values and accident density for different roads.
roads.

InIn Figure
Figure5,5,
InFigure 5,aaagood
good correlation
goodcorrelation between
correlation between
between the data from
the data from surveys
from surveys and
surveys and the
and thestatistical
the statisticaldata
statistical dataabout
data about
about
accidents
accidents appears:
accidentsappears:
appears: the the higher
the higher
higher thethe accident
the accident
accident densitydensity (AD),
(AD), the
(AD), the greater
the greater the
greater the BIR
the BIR values.
BIR values. ROAD1
values. ROAD1and
ROAD1 and
and
ROAD2
ROAD2 have
have thethe highest
highest values
values of of
both both
BIR BIR
and and
AD. AD.
ROAD5ROAD5
does
ROAD2 have the highest values of both BIR and AD. ROAD5 does not comply with this point does
not not
comply comply
with with
this this
point point
because
AD because
and BIR
because AD and
and BIR
ADvalues BIR values
refer refer
refer to
to different
values different
different conditions.
toconditions. Indeed, AD
conditions. Indeed,
refers AD
Indeed, refers
to 2015,
AD to
to 2015,
referswhile while
whiletosurveys
surveys
2015, to
obtain BIR
surveys to
obtain
occurred BIR
in occurred
2017, after in 2017,
safety after
works safety
carried works
out incarried
2016. out in
Therefore, 2016.
obtain BIR occurred in 2017, after safety works carried out in 2016. Therefore, ROAD5 has been Therefore,
ROAD5 has ROAD5
been has
removed been
from
removed
analysis
removed of from
from analysis
correlation
analysis of
of correlation
between between
road defects
correlation between androad
road defects
accidents and
defectsin
and accidents
order to define
accidents in
in order to
to define
Equation
order (14):Equation
define Equation
(14):
(14):
AD = 2 BIR + 9.6 (14)
AD
AD == 22 BIR
BIR ++ 9.6
9.6 (14)
(14)
Moreover,
Moreover,
Moreover,the the analysis
theanalysis
analysisof of SIR
ofSIR
SIRj,rj,r allowed for the identification
allowed identification of of the
the most
mosthazardous
hazardoussections
j,r allowed for the identification of the most hazardous sections
sections
according
according to to Table
Table 7. 7. Figure
Figure 6 6 shows
shows the the results
results obtained
obtained for for ROAD7.
ROAD7.
according to Table 7. Figure 6 shows the results obtained for ROAD7. Its risk map is in Figure Its Its
risk risk
map map
is in is in
Figure Figure
7, 7,
where
7,
it where
is it
possible is possible
to identify to identify
sections sections
4 and 7 4 and
(i.e., 7 (i.e.,
orange orange
and and
yellow yellow
bars,
where it is possible to identify sections 4 and 7 (i.e., orange and yellow bars, respectively). bars, respectively).
respectively).

45
45
40
40
35
35
30
30
(%)

25
j j(%)

25
20
SIR

20
SIR

15
15
10
10
5
5
0
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Number of section
Number of section
SIR Critical Very high High Moderate Low
SIR Critical Very high High Moderate Low

Figure 6. SIRj of ROAD7.

Figure 6.
Figure 6. SIRjj of
of ROAD7.
ROAD7.
Safety 2018, 4, 58 12 of 16
Safety 2018, 4, x 12 of 16
Safety 2018, 4, x 12 of 16

NOT RELEVANT LOW MODERATE HIGH VERY HIGH CRITICAL

NOT RELEVANT LOW MODERATE HIGH VERY HIGH CRITICAL
Figure 7. Map of risk of ROAD7.
Figure 7. Map of risk of ROAD7.
BIR of ROAD7
BIR ROAD7 is equal
equal to
to 9.15%
9.15% (i.e.,
(i.e., not
not relevant
relevant level
level of
of risk);
risk); however,
however, according
according to Figures
Figures 6
BIR ofof ROAD7 is is equal to 9.15% (i.e., not relevant level of risk); however, according to
to Figures 66
and
and 7, it is possible to observe that section n. 4 has a high level of risk (i.e., 31.2%). Table 9 lists data
and 7,
7, it
it is
is possible
possible to
to observe
observe that
that section
section n.n. 44 has
has a
a high
high level
level of
of risk
risk (i.e.,
(i.e., 31.2%). Table 99 lists
31.2%). Table lists data
data
used
used to
to perform
perform the
the risk
risk assessment.
assessment.
used to perform the risk assessment.
Table 9. Inputdata
data forBIR
BIR4,ROAD7 calculation.
Table9.9.Input
Table Input datafor
for BIR4,ROAD7 calculation.
4,ROAD7 calculation.

Defect Code
Bi Bi K1i K2i= =KKP K K333 K4i K5i SFR4,ROAD7
Defect Defect Code Code BKi 1i K1i KK2i 2i = K
PP K K4iK4iK5i KSFR SFR4,ROAD7
5i 4,ROAD7
Missing sidewalk C5 4 1 1.5 2.5 5.0 1.5 112.50
MissingMissing
sidewalk sidewalk C5 4
C5 41 1 1.5
1.5 2.5
2.5 5.05.01.5 1.5 112.50112.50
Hazardous
Hazardous maneuvers maneuvers J3 J3
3 31 1 1.01.0 2.5
2.5 3.0 1.5 33.75
Hazardous maneuvers J3 3 1 1.0 2.5 3.03.01.5 1.5 33.75 33.75
Narrow
Narrow sidewalk
sidewalk (less than(less than 0.5 m-wide)
0.5 m-wide) C6 C6
2 21 1 1.51.5 2.5
2.5 5.05.02.5 2.5 93.75 93.75
Narrow sidewalk radius
Low planimetric (less than 0.5 m-wide)G1 C6
3 20.9 1 1.5 2.5 5.02.52.5 2.5 93.75 42.19
Low planimetric radius G1 3 0.9 1.0 1.0 2.5
2.5 2.5 2.5 42.19
Low
Not visible planimetric
vertical road signsradius S5 G1
3 30.7 0.9 1.0 1.0 2.5
1.5 2.53.02.5 1.5 42.19 23.63
Not visible vertical road signs S5 3 0.7 1.0 1.5 3.0 1.5 23.63
PrivateNot visible
access: lack vertical road signs A2
of visibility S5
3 30.9 0.7 1.0 1.0 1.5
2.5 3.02.51.5 1.5 23.63 25.31
Private access: lack of visibility A2 3 0.9 1.0 2.5 2.5 1.5 25.31
Private access: lack of visibility A2 3 0.9 1.0 2.5 2.5 1.5 25.31
Figure
Figure 8a,b
8a,b represent
represent two
two surveyed
surveyedconditions
conditionswhich
whichbelong
belongto
toSection
Sectionn.4.
n.4.
Figure 8a,b represent two surveyed conditions which belong to Section n.4.

(a) (b)
(a) (b)
Figure 8. Section
Figure Sectionn.4
n.4ofofROAD7:
ROAD7:example
example of of
surveyed conditions.
surveyed (a) Low
conditions. planimetric
(a) Low radius
planimetric G1 (BG1
radius i =
Figure 8. Section n.4 of ROAD7: example of surveyed conditions. (a) Low planimetric radius G1 (Bi =
3);i =
(B (b)3);Hazardous maneuver
(b) Hazardous J3 (BiJ3= (B
maneuver 3).i = 3).
3); (b) Hazardous maneuver J3 (Bi = 3).
The
The showed
showed elements
elements could could cause
cause rear-end
rear-end collisions
collisions (Figure
(Figure 8a),
8a), and
and head-on
head-on and
and lateral
lateral
The showed
collisions elements could cause rear-end collisions (Figure 8a), and head-on andare
lateral
collisions (Figure 8b) along the examined section. Details about geo-referenced accidents are not
(Figure 8b) along the examined section. Details about geo-referenced accidents not
collisions therefore,
available; (Figure 8b) along the examined section. Details about geo-referenced accidents are not
available; therefore, itit is
is not
notpossible
possibleto
tocorrelate
correlatetype
typeof
ofdefects/elements
defects/elements and
and type
type of
of accidents:
accidents: it
it is
is
available;that
desirable therefore,
these it is not
data possible
remain open to correlate
for further type of defects/elements and type of accidents: it is
investigation.
desirable that these data remain open for further investigation.
desirable that these data remain open for further investigation.
Safety 2018, 4, 58 13 of 16

The implementation of the proposed methodology allows the management body to identify and
make decisions about the strategic priorities for interventions of safety improvement at section and
branch levels: geometric, functional, and traffic data of the branches contribute to the assessment
of the accident risk for vulnerable users. Indeed, the results are numerical and synthetic, and in
decision-making processes, they could allow the management body to identify and to determine the
strategic priorities about urban road safety.
Moreover, the proposed model could be modified to assess the index risk of urban intersections.
Indeed, the proposed procedure could be modified (e.g., modification of the extension factor), and new
relevant coefficients could be calibrated. Finally, the implementation of the approach both at section
and at intersection levels could permit us to pursue a network-level approach.

4. Conclusions
The road transport sector is currently adopting growing measures to prevent accidents and reduce
their consequences on people, especially on the most vulnerable users (e.g., pedestrians and cyclists).
This paper presents a quantitative risk analysis of deaths and serious injuries caused by urban
road accidents. The study proposed a methodology based on the visual inspection to interpret the
results from Road Safety Inspections on urban roads, to quantify the safety conditions, and to direct
the competent bodies towards the most appropriate interventions.
The method depends on the assumed ranges of variables and risk classes, as well as on the values
attributed to the variables used for calculating the hazard index of examined homogeneous road
sections and branches. Therefore, both the Section Index Risk (SIR) and the Branch Index Risk (BIR)
depend on geometric, functional, physical, and environmental defects or elements which are potential
source of road accidents. These factors are then related to the involved vulnerable road users and
to existing traffic flows to assess the current levels of risk. The categorization of these values into
six levels of risk allows the identification of the most severe conditions and the prioritization of road
safety works.
The results from surveying 50 km of roads in an Italian municipality demonstrated the good
performance of the proposed tool in identifying, planning, and scheduling all the work required for
improving urban road safety, because it is sensitive to improvements of infrastructure.
Moreover, the proposed methodology has both a diagnostic purpose, in order to evaluate whether
there may be a correlation between the observed defects and the occurred accidents, and a preventive
purpose, in order to correct defects or anomalies that could cause death or serious injuries of road users.

Author Contributions: Conceptualization, G.L.; Data curation, F.D.; Formal analysis, F.D.; Investigation, L.M.;
Methodology, G.L.; Writing—original draft, L.M.; Writing—review & editing, L.M. and G.L.
Funding: This research received no external funding.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. ITF. Road Safety Annual Report 2017; OECD Publishing: Paris, France, 2017.
2. World Health Organization. Global Status Report on Road Safety 2013—Supporting a Decade of Action; World
Health Organization: Geneva, Switzerland, 2013.
3. United Nations. Improving Global Road Safety. Resolution Adopted by the General Assembly on 2 March 2010;
United Nations: New York, NY, USA, 2010.
4. European Commission. Roadmap to a Single European Transport Area—Towards a Competitive and Resource
Efficient Transport System; White Paper, COM(2011) 144 Final; European Commission: Brussel, Belgium, 2011.
5. Canale, S.; Leonardi, S.; Pappalardo, G. The reliability of the urban road network: Accident forecast
models. In Proceedings of the 3rd International Congress SIIV—People, Land, Environment and Transport
Infrastructures, Bari, Italy, 22–24 September 2005.
Safety 2018, 4, 58 14 of 16

6. Cantisani, G.; Fascinelli, G.; Loprencipe, G. Urban road noise: The contribution of pavement discontinuities.
In Proceedings of the 2012 International Conference on Sustainable Design and Construction, Fort Worth,
TX, USA, 7–9 November 2012; pp. 327–334.
7. Di Mascio, P.; Fusco, G.; Grappasonni, G.; Moretti, L.; Ragnoli, A. Geometrical and functional criteria as a
methodological approach to implement a new cycle path in an existing urban road network: A case study in
Rome. Sustainability 2018, 10, 2951. [CrossRef]
8. Cantisani, G.; Loprencipe, G.; Primieri, F. The integrated design of urban road intersections: A case study.
In Proceedings of the 2012 International Conference on Sustainable Design and Construction, Kansas City,
MO, USA, 23–25 March 2012; pp. 722–728.
9. Mauro, R.; Cattani, M. Model to evaluate potential accident rate at roundabouts. J. Transp. Eng. 2004, 130,
602–609. [CrossRef]
10. Beck, B.; Cameron, P.A.; Fitzgerald, M.C.; Judson, R.T.; Teague, W.; Lyons, R.A.; Gabbe, B.J. Road safety:
Serious injuries remain a major unsolved problem. Med. J. Aust. 2017, 207, 244–249. [CrossRef] [PubMed]
11. Corazza, M.V.; Musso, A.; Finikopoulos, K.; Sgarra, V. An analysis on health care costs due to accidents
involving powered two wheelers to increase road safety. Transp. Res. Procedia 2016, 14, 323–332. [CrossRef]
12. Piantini, S.; Baldanzini, N.; Pierini, M.; Mangini, M.; Franci, A.; Peris, A. An Overview on Pedestrians
and Cyclists Serious Injuries in Urban Accidents. In Proceedings of the International Research Council on
Biomechanics of Injury (IRCOBI), Zurich, Switzerland, 9–11 September 2015.
13. Lamm, R.; Mailaender, T.; Psarianos, B. Highway Design & Traffic Sagety Engineering Handbook; Mc Graw Hill:
New York, USA, 1999.
14. Abojaradeh, M.; Jrew, B.; Al-Talafeeh, A. The Effect of Driver Behavior Mistakes on Traffic Safety.
Civ. Environ. Res. 2014, 6, 39–54.
15. Madsen, T.K.O.; Lahrmann, H. Comparison of five bicycle facility designs in signalized intersections using
traffic conflict studies. Transp. Res. Part F 2016, 46, 438–450. [CrossRef]
16. Nitsche, P.; Saleh, P.; Helfert, M. Forgiving Roadside Design Guide: Improving Roadside Design to Forgiven
Human Errors Deliverable n. 3—Annex 1: State of the Art Report on Existing Treatments for the Design of
Forgiving Roadside. 2011. Available online: http://www.cedr.eu/download/other_public_files/research_
programme/eranet_road/call_2009_safety/irdes/04_IRDES_D3_Forgiving_Roadside_Guidelines_V2.0_31.1.
2012.pdf (accessed on 16 November 2018).
17. Corazza, M.V.; Di Mascio, P.; Moretti, L. Managing sidewalk pavement maintenance: A case study to increase
pedestrian safety. J. Traffic Transp. Eng. 2016, 3, 203–214. [CrossRef]
18. Corazza, M.V.; Di Mascio, P.; Moretti, L. Management of sidewalk maintenance to improve walking comfort
for senior citizens. WIT Trans. Built Environ. 2018, 176, 195–206.
19. European Commission. Directive 2008/96/EC of the European Parliament and of the Council of 19 November 2008
on Road Infrastructure Safety Management; European Commission: Brussel, Belgium, 2008.
20. Parlamento Italiano. Decreto Legislativo n. 35/2011. Attuazione della Direttiva 2008/96/CE sulla Gestione della
Sicurezza delle Infrastrutture Stradali (Implementation of the European Directive 2008/96/CE Concerning the Road
Safety Management); Parlamento Italiano: Rome, Italy, 2011.
21. Ministero delle Infrastrutture e dei Trasporti. Linee Guida per la Gestione della Sicurezza delle Infrastrutture
Stradali; Ministero delle Infrastrutture e dei Trasporti: Rome, Italy, 2012.
22. Jamroz, K.; Budzyński, M.; Kustra, W.; Michalski, L.; Gaca, S. Tools for road infrastructure safety
management—Polish experiences. Transp. Res. Procedia 2014, 3, 730–739. [CrossRef]
23. Kustra, W.; Michalski, L. Tools for road infrastructure safety management in Poland. In Proceedings of the
MATEC Web of Conferences 2017, Sun Moon Lake, Taiwan, 29 September–2 October 2017.
24. Alfonsi, R.; Persia, L.; Antonino, T.; Usami, D.S. Advances in Road Safety Management Analysis. Transp. Res.
Procedia 2016, 14, 2064–2073. [CrossRef]
25. Persia, L.; Usami, D.S.; De Simone, F.; Beaumelle, V.F.D.L.; Yannis, G.; Laiou, A.; Han, S.; Machata, K.;
Pennisi, L.; Marchesini, P.; et al. Management of Road Infrastructure Safety. Transp. Res. Procedia 2016, 14,
3426–3445. [CrossRef]
26. Lynam, D. Development of Risk Models for the Road Assessment Programme; International Road Assessment
Programme: London, UK, 2012.
27. American Association of State Highway and Transportation Officials (AASHTO). Highway Safety Manual,
1st ed.; AASHTO: Washington, DC, USA, 2010.
Safety 2018, 4, 58 15 of 16

28. American Association of State Highway and Transportation Officials (AASHTO). An Introduction to the
Highway Safety Manual; AASHTO: Washington, DC, USA, 2010.
29. Appleton, I. Road Infrastructure Safety Assessment. In Proceedings of the 4th IRTAD Conference, Soul,
Korea, 16–17 September 2009.
30. Koorey, G.; Carpenter, M.; Appleton, I. Technical Note—Safety Audits of Existing Roads Database.
In Proceedings of the IPENZ Transportation Group Technical Conference, Christchurch, New Zealand,
17 September 2003.
31. Quezon, E.T.; Wedajo, T.; Mohammed, M. Analysis of Road Traffic Accident Related of Geometric Design
Parameters in Alamata-Mehoni-Hewane Section. Int. J. Sci. Eng. Res. 2017, 8, 874–881.
32. Loprencipe, G.; Moretti, L.; Cantisani, G.; Minati, P. Prioritization methodology for roadside and guardrail
improvement: Quantitative calculation of safety level and optimization of resources allocation. J. Traffic
Transp. Eng. 2018, 5, 348–360. [CrossRef]
33. Loprencipe, G.; Cantisani, G.; Di Mascio, P. Global assessment method of road distresses. Life-Cycle
of Structural System: Design, Assessment, Maintenance and Management. In Proceedings of the 4th
International Symposium on Life-Cycle Civil Engineering, Tokyo, Japan, 16 November 2014; CRC Press:
Leiden, The Netherlands, 2015.
34. EuroRAP 2009. Star Rating Road for Safety: The EuroRAP Methodology; European Road Assessment Programme:
Brussels, Belgium, 2009.
35. Transportation Research Board. Incorporating Reliability Performance Measures into the Transportation Planning
and Programming Processes: Technical Reference; SHRP 2 Report S2-L05-RR-3; The National Academies Press:
Washington, DC, USA, 2014.
36. Cafiso, S.; D’Agostino, C.; Persaud, B. Investigating the influence of segmentation in estimating safety
performance functions for roadway sections. J. Traffic Transp. Eng. 2018, 5, 129–136. [CrossRef]
37. Ministero delle Infrastrutture e dei Trasporti. Norme Funzionali e Geometriche per la Costruzione delle Strade;
Italian Ministry of Transportation Decreto Ministeriale: Rome, Italy, 2001.
38. Ministero delle Infrastrutture e dei Trasporti. Norme Funzionali e Geometriche per la Costruzione delle Intersezioni
Stradali; Italian Ministry of Transportation Decreto Ministeriale: Rome, Italy, 2006.
39. UNI. UNI 10439:2001: Illuminotecnica—Requisiti Illuminotecnici delle Strade con Traffico Motorizzato; Ente Italiano
di normazione: Rome, Italy, 2001.
40. Transportation Research Board. Highway Capacity Manual: A Guide for Multimodal Mobility Analysis, 6th ed.;
The National Academies Press: Washington, DC, USA, 2016.
41. Croft, A.C.; Freeman, M.D. Correlating crash severity with injury risk, injury severity, and long-term
symptoms in low velocity motor vehicle collisions. Med. Sci. Monit. 2005, 11, RA316–RA321. [PubMed]
42. Beck, B.; Stevenson, M.; Newstead, S.; Cameron, P.; Judson, R.; Edwards, E.R.; Bucknill, A.; Johnson, M.;
Gabbe, B. Bicycling crash characteristics: An in-depth crash investigation study. Accid. Anal. Prev. 2016, 96,
219–227. [CrossRef] [PubMed]
43. Kröyer, H.R.G. Is 30 km/h a ‘safe’ speed? Injury severity of pedestrians struck by a vehicle and the relation
to travel speed and age. IATSS Res. 2015, 39, 42–50. [CrossRef]
44. Cameron, M.H.; Elvik, R. Nilsson’s Power Model connecting speed and road trauma: Applicability by road
type and alternative models for urban roads. Accid. Anal. Prev. 2010, 42, 1908–1915. [CrossRef] [PubMed]
45. Biswas, S.; Singh, B.; Saha, A. Assessment of level-of-service on urban arterials: A case study in Kolkata
metropolis. Int. J. Traffic Transp. Eng. 2016, 6, 303–312. [CrossRef]
46. Wramborg, P. A new Approach to a Safe and Sustainable Traffic Planning and Street Design for Urban Areas.
In Proceedings of the Road Safety in Four Continents Conference, Warsaw, Poland, 5–7 October 2005.
47. Bonin, G.; Folino, N.; Loprencipe, G.; Oliverio Rossi, G.; Polizzotti, S.; Teltayev, B. Development of a road
asset management system in Kazakhstan. In Proceedings of the 1st TIS 2017 International Congress on
Transport Infrastructure and Systems, Rome, Italy, 10–12 April 2017.
48. Di Mascio, P.; Loprencipe, G. Risk analysis in the surrounding areas of one-runway airports: A methodology
to preliminary calculus of PSZs dimensions. J. Eng. Appl. Sci. 2016, 11, 13641–13649.
49. Moretti, L.; Cantisani, G.; Caro, S. Airport veer-off risk assessment: An Italian case study. ARPN J. Eng.
Appl. Sci. 2017, 12, 900–912.
Safety 2018, 4, 58 16 of 16

50. Moretti, L.; Cantisani, G.; Di Mascio, P.; Nichele, S.; Caro, S. A runway veer-off risk assessment based on
frequency model: Part I. Probability analysis. In Proceedings of the 1st AIIT International Congress on
Transport Infrastructure and Systems, Rome, Italy, 10–12 April 2017.
51. Moretti, L.; Cantisani, G.; Di Mascio, P.; Nichele, S.; Caro, S. A runway veer-off risk assessment based on
frequency model: Part II. Risk analysis. In Proceedings of the 1st AIIT International Congress on Transport
Infrastructure and Systems, Rome, Italy, 10–12 April 2017.
52. Moretti, L.; Di Mascio, P.; Nichele, S.; Cokorilo, O. Runway veer-off accidents: Quantitative risk assessment
and risk reduction measures. Saf. Sci. 2018, 104, 157–163. [CrossRef]
53. Di Mascio, P.; Perta, G.; Loprencipe, G. The public safety zones around small and medium airports. Aerospace
2018, 5, 46. [CrossRef]
54. Mooney, C.Z. Monte Carlo Simulation; No. 116; Sage: London, UK, 1997.
55. Scozzafava, R. Primi Passi in Probabilità e Statistica (Basics of Probability and Statistics); Zanichelli Editore:
Bologna, Italy, 1995.
56. ISTAT. Incidenti Stradali in Italia; Istituto Nazionale di Statistica: Rome, Italy, 2016.

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).