Content uploaded by Veronique Cerezo
Author content
All content in this area was uploaded by Veronique Cerezo on Jan 17, 2016
Content may be subject to copyright.
22nd ITS World Congress, Bordeaux, France, 5–9 October 2015
Paper number ITS-2273
Heavy Commercial Vehicle Greening, Safety and Compliance
Bernard Jacob1*, Véronique Cerezo2
1. IFSTTAR, France
Cité Descartes - 14-20, bd Newton, 77447 Champs-sur-Marne, Marne-la-Vallée Cedex 2,
bernard.jacob@ifsttar.fr - tel : +33 1 81 66 83 12
2. IFSTTAR, France
Abstract
A series of ITS technologies were developed for heavy commercial vehicles (HCVs), to reduce the
consumption of fossil fuel and the CO2 emission, to increase the road safety and protect vulnerable
road users, and to ensure a better compliance of HCVs with the weights and dimensions regulations.
Advanced eco-driving strategies taking into account the road profile and the traffic conditions may be
implemented on connected HCVs in interaction with the infrastructure. Effective stability programmes
(e.g. anti roll-over and jackknifing) were also developed for such connected vehicles. Diagnosis tools
are derived to identify the road sections which are the most risky for HCVs. Weigh-in-motion systems
were developed and used to improve the vehicle compliance, to protect the infrastructures and to
ensure road safety and fair completion in transport. This paper presents an overview of these solutions
for HCVs.
Keywords: Heavy commercial vehicles, safety, compliance, emission.
1. Context and Challenges
In Europe and in many other regions of the World, freight transport is mainly operated on road with
heavy commercial/good vehicles (HCV). In Europe more than 75% of the freight transport is done by
road and this share is still increasing quicker than the other modes.
The white paper of the European Commission (EC, 2011) required a 20% cut off of the fossil fuel
consumption and CO2 emission by 2020 for road transport, and a reduction of 60% of CO2 emission
by 2050, compared to 1990. HCVs are responsible for 26% of the CO2 emission of the road transport
sector, which represent 71% of the emissions of the whole transport. Moreover, rolling resistance due
to tire-road interaction is a key parameter in heavy vehicle fuel consumption which can be influenced
by pavement design and management. As an example, at a speed of 90 km/h, fuel consumption
increases, per unit increase of IRI (evenness) and MPD (macro-texture), accordingly for a car by 0.8%
and 2.8%, for a heavy truck by 1.3% and 3.4% and for a truck with trailer by 1.7% and 5.3%
(Hammarström et al., 2012).
Heavy Commercial Vehicle Greening, Safety and Compliance
2
In parallel, road safety is still a major issue with more than 25,000 road fatalities per year in the EU
(EC, 2014), and 1.3 million worldwide. HCVs are involved in 5% of the accidents in the EU, but in
15% of the accidents with fatalities, while they represent 1.5 to 2% of the motor vehicles (cars, vans
and lorries). The risk of fatality is multiplied by 2.5 when a lorry is involved in an accident. In addition
roll-over and jackknifing of HCVs induced large congestions which have a severe impact on the traffic,
travel time and CO2 emissions of all other vehicles.
Then, 10 to 15% of HCVs are overloaded in developed countries (EU, North America, Australia…)
while this proportion may reach 20 to 60% in developing countries. The overload rates also vary from
5-10% for the majority of non compliant vehicles in the EU, to 50-100% in some other regions of the
World. Large overloads deeply affect the road infrastructures (pavements and bridges), either
shortening their lifetime by fatigue or leading to collapses in the most severe cases. Overloads also
deteriorate the road safety level while lorries are not designed for the heaviest loads and because the
kinetic energy, proportional to the load, is a main factor of severity in collisions. Overloads also induce
unfair competition between transport modes and between road transport companies.
Therefore research works are carried out to provide solutions, including ITS solutions, to mitigate the
above listed adverse effects. In the last 10 years, quick progresses were done thanks to a better
knowledge and accounting of the vehicle/driver/infrastructure interact-tions and the connected heavy
vehicle allows implementing a broad spectrum of ITS solutions (Janin, 2008), (Kolski, 2011). This
paper will present an overview of research results and solutions developed either in National,
European or International projects. The fields of investigation comprise stability programme such as
anti roll-over and jackknifing devices involving vehicle dynamic assessment, the infrastructure
geometry and early warnings delivered to the drivers, eco-driving and routing strategies based on an
advance knowledge of the infrastructure profile and the traffic and environmental conditions (such as
visibility), and weigh-in-motion (WIM) tools used for axle and vehicle loads assessment, self control
and enforcement. These solutions also contribute to improve the HCVs compliance with respect to the
evolving regulation, e.g. on weights and dimension such as the European Directive 96/53/EC.
2. HCV Greening
Very significant progresses were achieved over the last 20 years in the HCV design and motorization.
The powertrain, the suspension, the aerodynamics, etc. of HCVs were continuously improved,
allowing the vehicles to meet the classes Euro IV, V and VI (European Directives 2005/55/EC and
2008/74/EC).
However, in addition to the vehicle design and component optimization, its operation also contributes
to the emission and consumption performances. It was proven in driver trainings and trials that for the
same journey and lorry, in the same environmental conditions and travel time, the difference of
consumed fuel may vary by 20% between an experienced driver and a beginner.
Heavy Commercial Vehicle Greening, Safety and Compliance
3
Figure 1 – Eco-driving strategy
Eco-
driving strategies adapted to HCVs were developed
(Figure 1) or are under investigation (Chaari and Ballot,
2012). A promising way consists of taking into account the
local road profile, i.e. the longitudinal slopes and radius of
curvature, the sk
id resistance, the rolling resistance, the
traffic and environmental conditions (e.g. the visibility) to
continuously adapt the speed to an optimum in the safe
domain. HCVs may store a significant amount of energy in
the downhill slopes or during the braking periods, and reuse
it in uphill slopes or to restart or accelerate (Nouveliere et
al., 2010). The most advanced vehicles may be equipped
with temporary ener
gy storage (batteries, inertia wheels,
etc.), but all the vehicles may store some kinetic energy and
reuse it immediately (Kapoor and Parveen, 2013). The
challenge is maximizing this stored kinetic energy without
taking any risk for the vehicle safety. That required a high
degree of anticipation, both of the forthcoming road profile
and of the traffic condition in front of the vehicle. Coupling
a global positioning system (GPS) and information collected
by/on the infrastructure a few hundred meter in front of the
lorry, even on out of visibility zone, may help assessing the
optimal velocity and delivering advices to the driver through
a human machine interface (HMI), as advanced driver
assistance system (ADAS).
More advanced and automated driving systems ma
y also
directly adjust the vehicle speed and gear, under the control
of cruise control and obstacle detection systems.
Some research projects were conducted in France and in
Europe on that topic, such as eCoMove1
, in which HCVs
were instrumented to monito
r the instantaneous fuel
consumption while eco-
driving ADAS were implemented.
Spacing measurements collected by WIM systems were
used together with an optimization module of velocity and
gear to develop an eco-cruise control. Fuel and CO2 savings
are expected to reach up to 10 to 12%, while the vehicles better comply with the speed limitation.
1 http://www.ecomove-project.eu/
Heavy Commercial Vehicle Greening, Safety and Compliance
4
Routing strategies of HCVs may also contribute to save fuel and CO2. For a given journey with fixed
origin and destination, a dynamic routing program based on the traffic conditions, the vehicle loading
and performances, the road and environmental conditions may compare different routes and assess the
optimal one with respect to the emission of CO2. The European project Heavyroute investigated these
issues and suggested some solutions (Ihs et al., 2008), (Adesiyun and Ihs, 2009). The French project
VIF (Interactive vehicle of the future) provided additional results.
3. Safety and stability of HCVs
Losses of control of HCVs result either in collision, lane departure, roll-over or jackknifing. The last
event only occurs with articulated multibody vehicles. The last 3 types of accidents mainly involve a
single lorry but may have severe consequences on the traffic congestion or may induce accident with
following vehicles. Collisions with vulnerable road users, above all in urban area, are responsible for a
large number of fatalities in cities. Better identifying the road sections with higher risk level for HCVs
allows preventing accidents and implementing suitable solutions, such as targeted warnings.
Articulated lorries represent more than 55% of the HCVs in Europe, and are exposed to
roll-over and jackknifing when their trajectory, including speed and acceleration, does not
comply to the infrastructure conditions. Most of the marketed stability programme tools (ESP,
etc.) only consider the current instantaneous dynamic state of the vehicle and therefore use
instant stability indicators such as the LTR (Load transfer ratio) for roll-over to issue a
warning to the driver. But this warning mostly comes too late to avoid the loss of control,
because of a rather high threshold chosen for this indicator. Otherwise, a large number of
false warnings would make the system unreliable. Smart anti lane departure, roll-over and
jackknifing systems were developed in a couple of research projects (VIF, Corolla in France,
European project HeavyRoute) which combine the dynamic states of the vehicle both at the
instant t, but also at a future instant t+h (h = a few seconds, typically 5 to 15 s), using a
predictive model taking into account the road profile (radius of curves, transverse and
longitudinal slopes, etc.) in front of the vehicle available in a database accessible by remote or
integrated in the on-board computer. A combination of on-board sensors and state observers
feed the dynamic model which enables the system to calculate the dynamic state at t+h and to
issue an early warning if needed. On-board WIM systems were proposed to measure the
instantaneous wheel loads and to assess the LTR for anti roll-over systems.
3.1 Anti-rollover system
Rollover is a lateral instability due to a lateral load transfer. Several rollover indicators can be found in
the literature such as SRT (static rollover threshold), CSV (critical sliding velocity), etc. Nevertheless,
LTR (load transfer ratio) is the most widely used (Ackermann et al., 1999a) (Gaspar et al., 2006)
(Imine et al., 2007), (Bouteldja et al., 2010). In practice, when LTR is equal to 0, the heavy vehicle has
stable roll dynamics. The risk becomes high as this indicator goes towards ±1.
Heavy Commercial Vehicle Greening, Safety and Compliance
5
FF
FF
rightzleftz
rightzleftz
LTR
,,
,,
+
−
=
(1)
where Fz,left and Fz,right are the vertical forces applied on the left and right wheels.
The dynamical model uses data available in BUS-CAN and sliding mode observers to estimate lacking
parameters (Fig. 2). Several experiments were realized to validate the warning system.
Ve hic le
Obs erver 1
Obs erver 2
Obs erver 3
Sw it ch
Observation
Estimation
Rollover
detection
0,0 ==
yx
vv
yx
vv
T
δ
Ve hic le
Obs erver 1
Obs erver 2
Obs erver 3
Sw it ch
Observation
Estimation
Rollover
detection
0,0 ==
yx
vv
yx
vv
T
δ
u
Figure 2 - Rollover detection and prediction model (Bouteldja et al., 2010)
Lastly, these researches were completed by introducing reliability methods and stochastic approaches.
Indeed, parameters involved in the model present uncertainties that can be taken into account. Thus,
failure probability can be correctly approximated (Sellami et al., 2008) (Badi et al., 2012).
3.2 Anti-jackknifing system
Jack-knifing is a yaw instability of an articulated vehicle, leading to a folding of the semi-trailer or
trailer along the tractor (Ervin et al., 1998), (Weidner, 2007). It mainly occurs in curves and on
slippery road while braking and more frequently when the semi-trailer is empty. Some works proposed
a yaw control active systems of tractor and semi-trailer by acting independently or by combining the
braking or steering actions (Chen et al., 1997), (Ackermann et al., 1999b), (Peng et al., 1999). Then
(Timothy et al., 1981) proposed a mathematical model describing the jackknifing situation, which was
adapted in (Bouteldja et al.,2006).
Two indicators were developed: (i) based on the instantaneous assessment of the lorry dynamic
behaviour, (ii) based on the skid resistance mobilized by the vehicle on each tire/pavement contact
point. Again, on-board sensors deliver data to calculate the vehicle dynamic state and a digital map
provides the road characteristics. State observers and estimators complement the sensor measurements.
The risk is assessed by a model and an algorithm using these data, and a warning is elaborated to be
sent to the driver.
Thus, (Bouteldja et al., 2013) designed such system for jackknifing detection and prediction by using
on one hand, a 12-dof nominal model of articulated vehicle and on the other hand, detection algorithm
based on a jackknifing criterion and on the prediction function, in view of estimating the time to
jackknifing. The jackknifing criterion Cm is defined as follows:
0 5 10 15 20 25 30 35 40
Ti m e [s ]
Load Trans fer Ra ti o
-0.6
-1.3
-0.3
0
1
1.3
-1
-0.9
0.6
0.3
0.9
Rollover detection
0 5 10 15 20 25 30 35 40
Ti m e [s ]
Load Trans fer Ra ti o
-0.6
-1.3
-0.3
0
1
1.3
-1
-0.9
0.6
0.3
0.9
Rollover detection
Heavy Commercial Vehicle Greening, Safety and Compliance
6
( ) ()
85tan
85tan <
+
−
=<−
ψ
ψ
xy
yx
Cm
(2)
with ẋ: speed along axle Ox, Y: position along axle Oy, and ψ: yaw speed.
It is used with a Kalman method of prediction to define time to accident. Accidents risks prediction
method (jackknifing) is based on the evolution of the criterion, calculated with respect to the predicted
state. The tangent line of the evolution curve is determined for each time path. Thus, on each point of
the curve, which defines the risk level, the duration between the current time and the instant
corresponding to the intersection axis between the tangent line and the threshold value is calculated.
Figure 3 presents the definition and evolution of the indicator Cm, which tends to exceed the
thresholds value after 10 seconds. So, Cm is a good indicator for jackknifing detection.
Crit er ion evolu tion
+ threshold
time
y
∆t
1
: stat e predic tio n time.
∆t
2
: risk criterion prediction time.
m : the tangent point after a time
∆
t1.
m
∆t1∆t2
Crit er ion evolu tion
+ threshold
time
y
∆t
1
: stat e predic tio n time.
∆t
2
: risk criterion prediction time.
m : the tangent point after a time
∆
t1.
m
∆t1
∆t1∆t2
∆t2
Crite r io n e vo lut ion
time
-thr es ho ld
+ threshold
x0
y
y
x
∆t
Crite r io n e vo lut ion
time
-thr es ho ld
+ threshold
Crite r io n e vo lut ion
time
-thr es ho ld
+ threshold
x0
y
y
x
∆t
Figure 3 - Risk situation prediction principle and criterion prediction (Bouteldja et al., 2013)
Figure 4 - Yaw angular velocity and yaw angle evolution for a 5-axles tractor semi-trailer
These systems as well as the anti roll-over system were validated by simulation using a multibody
vehicle software (Figure 4). It was proved that the first anti jackknifing indicator was performing and
could be implemented in an operational tool.
3.3 Road safety diagnosis tool
A diagnosis tool (AlertinfraPL) was also developed for road managers and operators by the CEREMA
and IFSTTAR (Cerezo et al., 2012). It allows detecting risky zones also called “warnings” along a
02 4 6810 12 14
-10
0
10
20
30
40
Yaw angular
velocity in [°/s]
0 2 4 6 8 10 12 14
-20
0
20
40
60
80
Time in [s]
Yaw* angle in [°]
Tre nd to wa rds t he
jackknif ing situ ation
Jackkn ifing situa tio n
*: Yaw angle between tractor an d trailer at fi ft h-wh e el
0 2 4 6 8 10 12 14
-10
0
10
20
30
40
Yaw angular
velocity in [°/s]
0 2 4 6 8 10 12 14
-10
0
10
20
30
40
Yaw angular
velocity in [°/s]
0 2 4 6 8 10 12 14
-20
0
20
40
60
80
Time in [s]
Yaw* angle in [°]
02 4 6 8 10 12 14
-20
0
20
40
60
80
Time in [s]
Yaw* angle in [°]
Tre nd to wa rds t he
jackknif ing situ ation
Jackkn ifing situa tio n
*: Yaw angle between tractor an d trailer at fi ft h-wh e el
Heavy Commercial Vehicle Greening, Safety and Compliance
7
road infrastructure, with respect to geometrical and surface road characteristics, and to the type of road
network. It is dedicated to HCVs and takes into account their specific dynamic behaviour (sensitivity
to roll-over, jackknifing, etc.). The warnings are a combination of values taken by road characteristics
(crossfall, radius, skid resistance, slope), which exceed at the same time thresholds values that were
defined (Figure 5). A three steps methodology was applied to develop this safety tool. In a first step, a
literature study is conducted in view of listing the most dangerous situations for heavy vehicles. In a
second step, a numerical model is used to perform simulations in view of defining maximum speed
Vmax as a function of geometrical characteristics, surface characteristics and heavy vehicle
characteristics (type and load). Then, the difference between Vmax and a reference speed called Vref
is assessed. This difference represents the risk of accident. When it becomes too small the
corresponding thresholds values of the infrastructure characteristics are determined. In a third step, the
warnings are validated through statistical tests performed both on bidirectional and dual carriageways.
Figure 5 - Example of assessment of the thresholds value vs radius of curvature in a warning
On dual carriageways, six warnings in curves and two in ramps/descents are proposed (Table 1). The
combinations include parameters such as radius of curvature, skid resistance, slope and crossfall. On
single carriageways, five warnings are implemented in curves and five in roundabouts.
Lastly, a validation process based on accidents ratio analyses and use of statistical tests is applied.
Road characteristics and accidents data are collected on primary and secondary roads network. More
than 1000 km of motorways and around 500 km of single carriageway roads are included in the
database.
Tabl e 1 - Example of warning definition in curves on single carriageway roads
Warning
Radius of curvature (m)
Skid resistance
Crossfall (%)
1
R < 120 m
SFC < 0.40
-
2
R < 120 m
SFC < 0.60
3
R < 120 m
-
-
4
R < 120 m
-
< 5%
5
R < 150 m with geometrical problems
-
-
The network is divided in homogeneous areas based on infrastructure characteristics. Homogeneous
section is a part of the road network presenting similar geometrical characteristics. As it is not possible
to have exactly the same value for the relevant parameters, it was decided to define classes presenting
a range of values for a given parameter. A two step method is used. First, curves and straight lines are
separated. Then, a split is done on the longitudinal profile (flat section, ramp, down slope). Afterward,
0100 2 00 30 0 4 00 500 6 00
0
10
20
30
40
50
60
70
80
90
100
Radius of cur vatu re [m]
S pee d [km/h]
S pee d limit
V 85
H igh risk of accident
fo r heavy vehicles
0100 2 00 30 0 4 00 500 6 00
0
10
20
30
40
50
60
70
80
90
100
Radius of cur vatu re [m]
S pee d [km/h]
S pee d limit
V 85
0100 2 00 30 0 4 00 500 6 00
0
10
20
30
40
50
60
70
80
90
100
Radius of cur vatu re [m]
S pee d [km/h]
S pee d limit
V 85
S pee d limit
V 85
H igh risk of accident
fo r heavy vehicles
Heavy Commercial Vehicle Greening, Safety and Compliance
8
boundaries are proposed for the relevant parameters (slope, radius) to define classes in which data are
aggregating. Accidents rates are calculated by adding all the accidents data obtained on a given class
of parameter. Statistical tests allow a comparison between the rates calculated on the various classes to
estimate if one class of characteristics represents a significant risk for heavy vehicles. A level of
confidence of 95% is used to compare the risk levels.
Figure 6 - Validation of warning 2 (see table 1) in curves on single carriageway
An example of validation is provided in Figure 6. Calculations are performed with three situations of
accidents: accidents involving at least one heavy vehicle, accidents where heavy vehicle is considered
as responsible, accidents involving a heavy vehicle alone. For each type of accidents, the left bin
category indicates the accident rates calculated on sections with warnings and the right bin category
indicates the accident rates calculated on sections without warning. On each case, the accidents rates
are higher on sections with warnings than on other sections. Moreover, the levels of confidence of the
rates calculated on sections with and without warning are separated, which means that they
statistically differ. Thus, the risk of accident is higher on sections with warning than on other sections,
which allows validating this configuration for the diagnosis tool.
Another way to use this diagnosis tool, is to couple it with an on-board warning system. Indeed, it may
help drivers, if their vehicles are geo-located, identifying risky zones and adapt their driving behavior.
The limit of this application lies in the fact that the road network database need to be accurate and
complete. However, the development of network mapping (“Google cars”, etc.) and the new
possibilities offered by vehicle/infrastructure communication provide interesting perspectives.
4. Compliance of HCVs
HCVs must comply with the weight and dimension regulations (among others) in force in the territory
and road network of operation. In each country the driving law specifies overall dimension limits,
length, width and mostly height, as well as more specific dimensions (e.g. trailer or semi-trailer length,
etc.). Gross vehicle weight and axle loads are also limited by type of vehicles (number of axles, or
Heavy Commercial Vehicle Greening, Safety and Compliance
9
other categories), by type of axles or bogies (tandem, tridem, single or dual tire), and sometime by
type of suspension (air or steel). Derogations may be given to special vehicles or special transport. In
Europe and North America (United States and Canada), there are National or State regulations which
apply for National (EU Member States) or intra-state (province in Canada) transport, and European or
Federal (in North America) regulation which applies for International (in the EU) or Interstate (in
North America) transport. These regulations may significantly differ even if applicable in the same
territory or road network. E.g. the maximum gross vehicle weight for the largest combinations in the
EU Member States varies from 40 to 60 t, while the European Directive 96/53EC limits it to 40 t for
International transport. In the US, the Federal regulation limits the gross weight to 80,000 lbs, i.e. 36.3
t, while some State regulations allow up to 50 t and more.
Collecting HCV weight and dimension data during the journey by WIM systems allows to get
statistics over the whole traffic flow, but also large samples of individual data per vehicle, which are
very useful for infrastructure and road safety purposes and to ensure fair competition between
transport companies and modes (Jacob and Feypell, 2010). There is almost no limitation in the sample
size, but the storage and data transfer capacity of the system and network. WIM systems may be
divided into several types depending on the type of sensors used, their installation and the operation of
the system:
(i) Low speed WIM (LS-WIM) mainly used load cell plates mounted on a flat and smooth platform,
30 to 60 m in length (i.e. app. twice the vehicle length), installed outside the traffic lane, e.g. on a
parking lot or an adjacent lane nearby the highway, and the vehicle are diverted on the system and
cross it at 5 to 15 km/h, avoiding all the dynamic effects (Figure 7, left). The accuracy is very good
within the metrological tolerances of the OIML class 5 or 10, i.e. ±5 or 10% for the gross weight, and
±8 or 15% for the axle loads. These systems are only used for direct enforcement or trade, but are not
capable to check all the HCVs while they only may weigh 30 to 60 vehicles per hour.
(ii) High speed (HS-) WIM) mainly uses road sensors installed in the traffic lane, scales (load cell or
bending plate scales) or strip sensors (piezoelectric usually, polymer, ceramic or quarz) (Figure 7 right).
The sensors may weigh the whole axles if the plate or strip length is equal to the lane width, or a half
axle if the length if half of the lane width. In that case, a pair of sensors may be mounted to weigh
individually the left and right wheels. The accuracy depends on the sensor type, on the pavement
conditions and evenness, on the vehicle suspension and speed, etc. The most accurate systems are in
the COST323 accuracy class B(10), i.e. app. 95% of the measurements are within ±10% for the gross
weight, and ±15% for the axle loads (COST323, 1999). Most of these systems are in class C(15)
(±15% for the gross weight, and ± 20% for the axle loads), but sometime less accurate.
(iii) High speed bridge WIM (B-WIM): a bridge or part of it, e.g. the upper deck of a slab bridge, or
the main girders of a concrete or composite bridge, or the longitudinal stiffeners of steel orthotropic
deck bridges, is instrumented with extensometers or strain gauges, which measure the strain of the
bridge element while a lorry is crossing it (Figure 9). An algorithm adapted to the bridge type performs
a back calculation of the axle load and vehicle weight, and also identify the number of axles and the
Heavy Commercial Vehicle Greening, Safety and Compliance
10
speed. Not all the bridges are suitable for that, but with common slab or girder bridges fitting some
known criteria B-WIM is possible and gives rather accurate results, at least for the gross vehicle
weights, in the accuracy class B(10) or C(15). In some cases the class B+(7) is reached. The advantage
of B-WIM is to be undetectable by the drivers and the systems may be installed and removed without
stopping the traffic. The vertical dynamic motions of the HCVs are averaged all along the
instrumented span of the bridge, which increases the accuracy compared to a road sensor WIM system.
Figure 7 - Low speed (LS-) (left) and high speed (HS-) (right) weigh-in-motion systems
Figure 8 - Bridge weigh-in-motion on an integral concrete bridge in France
(iv) On-board WIM: HCVs are instrumented with accelerometers, strain gauges, displacement or
pressure sensors, etc., to collect data on the vertical forces applied on the wheel and tires. A dynamic
model of the vehicle is generally used and fed with these data, to compute more accurately the wheel
loads, above all for multi-body combinations. However, because of the complex dynamic behavior of
HCVs, most of the marketed systems only works while the vehicle is stopped on a flat surface, which
occurs quite often (rest area, traffic light, gas station, congestion, etc.). A few instrumented lorries
were developed for research or WIM system calibration purposes, and are rather accurate for on-board
WIM, in the accuracy class A(5). But they are very costly and require sophisticated dynamic models
and algorithms.
WIM systems were installed since the early 60s, but they were very inaccurate in the early stages.
Heavy Commercial Vehicle Greening, Safety and Compliance
11
Since the mid-90s and early 2000s, the performances of WIM systems dramatically increased because
of the progresses of the sensor technologies, of the electronic and softwares, but also because of a
better knowledge of the vehicle/pavement interaction and vehicle dynamics. The calibration methods
were adapted, e.g. automatic self-calibration, a smart technique which uses some specific HCVs
having one or two axles with almost constant static load, e.g. the steer axle of the fully loaded 5 axle
articulated combination, a 2 axle tractor and a semi-trailer with a tridem axle.
In several countries, a network of WIM systems coupled with cameras and automatic vehicle
identification (AVI) tools was installed on the highway and motorway network, to collect data 24 hrs
all along the year, helping to identify the overloaded vehicles (Jacob et al., 2015). Screening and
preselecting these vehicles upstream to a check point (equipped with static scales or LS-WIM) is
common in many countries, but some countries (e.g. The Netherland and France) also use the WIM
systems to collect data on the suspicious vehicles to make company profiling, to send warning to some
companies and to target the check in time and location. Two countries already pioneered the use of
WIM for direct enforcement (Taiwan and Czech Republic). A National research project is carried out
in France on this subject to prove the feasibility of direct enforcement by WIM on a large scale (Jacob
et al., 2015). These initiatives effectively contribute reducing overloads and getting a better
compliance of HCVs.
In Australia, the intelligent access program (IAP) was developed and is implemented by Transport
Certification Australia (TCA) (Koniditsiotis, 2008). The aim is to deliver access authorization to the
compliant vehicles on specific road section or itineraries, depending on the infrastructures, traffic and
environmental conditions. WIM systems are used to check the weight and dimensions of the applicant
vehicles during their journey. This is a typical example of ITS solution and communication between
HCVs, infrastructure and operation centers. In UK and a few other countries, managers of partially
damaged bridges sensitive to the traffic loads, installed WIM systems prior to the bridge to check if
the axle loads and gross weight could be accepted without any risk on the bridge, and if not the
vehicles may be either stopped or diverted to another route. For some long span old bridges (e.g.
suspended bridges), WIM systems may be installed at both ends of the bridge to continuously assess
the total load on the bridge, and to stop the HCVs outside the bridge if this load exceeds a given
threshold. B-WIM and other WIM systems are more and more commonly used to monitor sensitive
and damaged bridges to relate the bridge strains and stresses to the applied traffic loads and to detect
some modification of the bridge behavior which could announce more severe deteriorations or risk of
collapse.
5. Conclusions
While the road share is the dominant mode of freight transport worldwide, and still increasing, ITS
solutions and technologies are more and more used to mitigate negative impacts of heavy commercial
vehicles or to improve their efficiency.
Heavy Commercial Vehicle Greening, Safety and Compliance
12
To reduce the fuel consumption and CO2 emission, eco-driving and energy saving and management
strategies are developed, which use on-board and external data, incl. global positioning systems and
road databases, to optimize and anticipate the driving decisions.
Advanced stability programs based on dynamic assessment and interaction between vehicle and
infrastructure are developed to mitigate losses of control and accidents. Early warnings and driving
assistance may help the drivers using accurate stability indicators to prevent roll-over, jackknifing or
lane departure, and the same sensors and methods may also be used for an accurate diagnosis of the
infrastructure, to identify risky road sections for heavy vehicles.
Monitoring heavy vehicle (axle and gross) weights by WIM not only provides relevant data for
infrastructure design and maintenance and statistics of transport, but now is widely used to ensure
compliance of weights and dimensions, by detecting and checking overloads. The next stage will be
using WIM for direct enforcement, a very challenging objective.
References
1. Ackermann, J. and Odenthal, D. (1999a) Damping of vehicle roll dynamics by gain scheduled
active steering, in Proc. European Control Conf. Karlsruhe, Germany.
2. Ackermann, J., Bünte, T. and Odenthal, D. (1999b). Advantages of active steering for dynamic
control, Report, German Aerospace Center, 99ME013.
3. Adesiyun, A. and Ihs, A. (2009), Heavyroute - Intelligent Route Guidance for Heavy Vehicles, Final
Report, September, 64 pp. http://cordis.europa.eu/documents/documentlibrary/123823861EN6.pdf
4. Badi, A., Bouteldja, M. and Bernardin, F. (2012), A reliability analysis of rollover and jackknifing
risks, in Proc. of the 12th International Conference on Heavy Vehicle Transport Technology
(HVTT12), Stockholm, Sweden, September.
5. Bouteldja, M. and Cerezo, V. (2010), Driving situations analysis based on multi-models observers:
application to risk accident of heavy vehicles, in Proc. of the 11th Int. Conference on Heavy Vehicle
Transport Technology (HVTT11), Melbourne, Australia, March.
6. Bouteldja, M. and Cerezo, V. (2011), Jackknifing warning for articulated vehicles based on a
detection and prediction system, 3rd International Conference on Road Safety and Simulation,
Indianapolis, USA, September.
7. Bouteldja, M. Koita, A., Cadiou, J. C. and Dolcemascolo, V. (2006), Prediction and Detection of
Truck Jack-knifing, IEEE Vehicle Power and Propulsion Conference IEEE/VPPC, Windsor, UK.
8. Cerezo, V., Bouteldja, M. and Conche, F. (2012), AlertinfraPL: a safety tool for infrastructure
diagnosis devoted to heavy vehicles, in. Proc. of the 12th International Conference on Heavy
Vehicle Transport Technology (HVTT12), Stockholm, Sweden, September.
Heavy Commercial Vehicle Greening, Safety and Compliance
13
9. Chaari, H. and Ballot, E. (2012), Assessment of Trucks Emissions and Impact of Eco-driving in
Delivery Tours, 4th International Conference on Information Systems, Logistics and Supply Chain -
ILS 2012, August, Québec, Canada.
10. Chen, C. and Tomizuka, M. (1997), Modelling and control of articulated vehicles, Technical
report, UCB.ITS-PRR-97-42, University of California, Berkeley.
11. COST323 (1999), European Specification on Weigh-in-Motion of Road Vehicles, EUCO-COST
/323/8/99, LCPC, Paris, August, 66 pp., http://www.is-wim.org/doc/wim_eu_specs_cost323.pdf
12. European Commission (2011), Roadmap to a Single European Transport Area – Towards a
competitive and resource efficient transport system, White paper, 30 pp.
13. European Commission (2014), Road Safety Vademecum: Road safety trends, statistics and chal-
lenges in the EU 2010-2013, DG for Mobility and Transport, Unit C4–Road Safety, March, 14 pp.
14. Ervin, R.D., and Mallikarjunarao, C. (1981), A study of the yaw stability of tractor semi-trailer
combinations, Proc. 7th IASVD Symposium, Cambridge, UK, 111-126.
15. Gaspar, P., Szaszi, I. and Bokor, J. (2006), Rollover stability control in steer-by-wire vehicles
based on an Ipv method, Int. Journal of Heavy Vehicle Systems, Vol. 13, No 1, 125-143.
16. Hammarström, U., Eriksson, J., Karlsson, R. and Reza Yahya, M. (2012), Rolling resistance
model, fuel consumption model and the traffic energy saving potential from changed road surface
conditions, VTI report 748A, 100 pp.
17. Ihs, A., Omasits, D. and Lindberg, G. (2008), HeavyRoute - Intelligent Route Guidance for Heavy
Vehicles, in Proc. of the 10th Int. Conference on Heavy Vehicles Transport Technology (HVTT10),
Eds. B. Jacob, EJ Obrien, P. Nordengen, J.Wiley-ISTE (ISBN 978-1-84821-058-5), 260-269.
18. Imine, H. and Dolcemascolo, V. (2007), Rollover risk prediction of heavy vehicle in interaction
with infrastructure, International Journal of Heavy Vehicle Systems, Vol. 14, No 3.
19. Jacob, B. and Feypell de La Beaumelle, V. (2010), Improving truck safety: Potential of
weigh-in-motion technology, IATSS Research, Vol. 34, Issue 1, July, 9–15.
20. Jacob, B., Cottineau, L-M. and Simonin, J-M (2015), Wei gh -in-motion for controlling excess
loads, Routes-Roads, N°367, July, 46-55.
21. Janin, J-F. (2008), Intelligent Freight, in Proc. of the 10th Int. Conf. on Heavy Vehicles Transport
Technology (HVTT10), Eds. B. Jacob et al., J.Wiley-ISTE (ISBN 978-1-84821-058-5), 34-39.
22. Kapoor, R. and Parveen, M. (2013), Comparative Study on Various KERS, in Proc. of the World
Congress on Engineering, Vol III, WCE 2013, London, UK, July.
23. Kolski, C. (2011), Human-Computer Interaction in Transport, chapter 6, J. Wiley/ ISTE, 397 pp.
Heavy Commercial Vehicle Greening, Safety and Compliance
14
24. Koniditsiotis, C. (2008), Intelligent Access Program, in Proc. of the 10th International Conference
on Heavy Vehicles Transport Technology (HVTT10), Eds. B. Jacob, EJ Obrien, P. Nordengen,
J.Wiley-ISTE (ISBN 978-1-84821-058-5), 280-287.
25. Nouveliere, L., Luu, H-T., Duval, F-R., Jacob, B. and Mammar, S. (2010), Eco-driving assistance
system for low fuel consumption of a heavy vehicle: advisor system, in Proc. of the 11th Int.
Conference on Heavy Vehicle Transport Technology (HVTT11), March, Melbourne, Australia.
26. Peng, H. and Chen, B. (1999), Rollover warning of articulated vehicles based on a time to
rollover metric, ASME International Congress.
27. Sellami, Y. Imine, H. Jacob, B. Bernardin, F. and Cadiou, J.-C. (2008), Rollover risk prevention of
heavy vehicles by reliability-based analysis, in Proc. of the 10th Int. Conf. on Heavy Vehicles
Transport Technology, Eds. B. Jacob et al., J.Wiley-ISTE (ISBN 978-1-84821-058-5), 199-208.
28. Timothy, V. Fossum, V. Gilbert, N. and Lewis, A. (1981), Mathematical model for trailer-truck
jackknifing, SIAM Review, vol. 23 (1).
29. Weidner, J. L. (2007), Tractor trailer instability of variably loaded container trucks, PHD thesis,
Pennsylvania State University.
