European Journal of Operational Research 286 (2020) 401–416

Contents lists available at ScienceDirect

European Journal of Operational Research

journal homepage: www.elsevier.com/locate/ejor

Invited Review

A concise guide to existing and emerging vehicle routing problem

variants
Thibaut Vidal a,∗, Gilbert Laporte b, Piotr Matl c
a
Departamento de Informática, Pontifícia Universidade, Católica do Rio de Janeiro, Brazil
b
CIRRELT, Canada Research Chair in Distribution Management and HEC Montréal, Canada
c
Department of Business Decisions and Analytics, University of Vienna, Austria

a r t i c l e i n f o a b s t r a c t

Article history: Vehicle routing problems have been the focus of extensive research over the past sixty years, driven
Received 16 June 2019 by their economic importance and their theoretical interest. The diversity of applications has motivated
Accepted 4 October 2019
the study of a myriad of problem variants with different attributes. In this article, we provide a concise
Available online 10 October 2019
overview of existing and emerging problem variants. Models are typically refined along three lines: con-
Keywords: sidering more relevant objectives and performance metrics, integrating vehicle routing evaluations with
Transportation other tactical decisions, and capturing fine-grained yet essential aspects of modern supply chains. We
Combinatorial optimization organize the main problem attributes within this structured framework. We discuss recent research di-
Vehicle routing problem rections and pinpoint current shortcomings, recent successes, and emerging challenges.
Challenges and perspectives
© 2019 Elsevier B.V. All rights reserved.

1. Introduction problem variants has grown rapidly, reflecting the diversity of ap-
plications. Vehicle routing algorithms are no longer used only to
Vehicle routing problems (VRPs) have been the subject of inten- produce daily routes but also serve as evaluation tools for other
sive and fast-growing research over sixty years. This is due to their strategic and tactical decisions such as facility location, fleet sizing,
economic importance and their theoretical interest. Using efficient production, and inventory management (Andersson, Hoff, Chris-
vehicle routes provides a direct competitive advantage to trans- tiansen, Hasle, & Løkketangen, 2010; Hoff, Andersson, Christiansen,
portation companies, which usually operate with limited profitabil- Hasle, & Løkketangen, 2010).
ity margins. Moreover, the fact that these problems share a simple The goal of this article is to draw a succinct picture of cur-
yet rich structure, generalizing the traveling salesman problem, has rent research in the field of vehicle routing. It is addressed to re-
helped to elevate the VRP family into one of the main testbeds searchers and practitioners who wish to consult a concise review
for studies in combinatorial optimization and heuristics. The VRP of existing problem features and applications. We discuss within
family can be seen as combinatorial in two senses: (1) because of a structured framework the main problem attributes and research
the number of possible solutions, which grows exponentially with directions in the field of vehicle routing as of 2020, pinpointing
the size of the instances, and (2) because the number of conceiv- current shortcomings, recent successes and emerging challenges.
able problem variants also grows exponentially with the variety of Given the breadth of the field, a description of every available
problem attributes, i.e., the specific constraints, decision sets and study is now impractical. This paper therefore does not claim to
objectives arising from real applications (Vidal, Crainic, Gendreau, be exhaustive in its coverage. Instead, we have opted for a struc-
& Prins, 2013). ture based on themes rather than on VRP variants, as is the case of
The VRP research landscape has dramatically evolved over the several books or review papers, and refer to the books of Golden,
past two decades. Until the early 20 0 0s, most methodological stud- Raghavan, and Wasil (2008) and Toth and Vigo (2014) for a more
ies were centered around a limited subset of operational prob- detailed coverage of specific problem variants. This work is orga-
lems with attributes such as time windows, multiple depots, mul- nized according to application-centered goals and concerns. From
tiple periods, and heterogeneous fleets. Since then, the number of a high-level perspective, a VRP model can be extended along three
main lines: (1) considering relevant side metrics, objectives, or
combinations of objectives; (2) integrating routing optimization
∗
Corresponding author.
with other business decisions; (3) progressing toward more precise
E-mail addresses: [email protected] (T. Vidal), [email protected]
(G. Laporte), [email protected] (P. Matl).
and fine-grained models.

https://doi.org/10.1016/j.ejor.2019.10.010
0377-2217/© 2019 Elsevier B.V. All rights reserved.
402 T. Vidal, G. Laporte and P. Matl / European Journal of Operational Research 286 (2020) 401–416

We discuss the academic problem variants and studies ac- optimizer has the authority to select some of the deliveries, giv-
cording to these three classifications in Sections 2–4. Then, ing rise to the class of VRPs with profits (Archetti, Speranza, &
we highlight some important challenges and conclude in Vigo, 2014b). In most of these problems, customers are associated
Section 5. with individual prizes, and the objective is to maximize the total
profit as the difference between collected prizes and routing costs.
2. Emerging objectives – measuring as a step toward Other problem variants maximize profit subject to distance or time
optimizing constraints. These problems are connected to numerous applica-
tions in production planning and logistics (Aksen, Kaya, Salman, &
Measurement and quantification are central to any optimization Akça, 2012), manufacturing (Lopez, Carter, & Gendreau, 1998), mil-
algorithm for business processes. Most of the VRP literature con- itary reconnaissance (Mufalli, Batta, & Nagi, 2012), and the design
siders cost as the main objective, but this does not capture all rele- of tourist itineraries (Vansteenwegen & Souffriau, 2009), among
vant performance criteria and metrics arising in practice, and many others.
solutions based on cost optimization alone turn out to be impos- Outsourcing. To respond to growing delivery volumes while lim-
sible to apply in practice. In these contexts, other metrics must iting the impact of high variance in shipping patterns, many freight
be considered, either as additional objectives or as constraints. We forwarding companies regularly outsource a portion of their busi-
subdivide these metrics into seven main categories: ness to subcontractors. This practice has led to the VRP with pri-
vate fleet and common carrier (see, e.g., Côté & Potvin, 2009),
(1) profitability: performance ratios, profits, outsourcing; which can be viewed as a special case of VRP with profits in
(2) service quality: cumulative objectives, inconvenience mea- which each customer’s prize represents its outsourcing cost. Sev-
sures, service levels; eral variants of this problem have recently been studied. Krajewska
(3) equity: workload balance, service equity, collaborative plan- and Kopfer (2009) discuss the impact of considering heteroge-
ning; neous subcontractors and distinguish three types of direct out-
(4) consistency: temporal, person-oriented, regional, or delivery sourcing cost: per tour, flat rate per day, and flow-based de-
consistency, inconsistency; pending on distance and weight. Ceschia, Gaspero, and Schaerf
(5) simplicity: compactness, separation, navigation complexity; (2011) consider a cost function with coefficients that depend not
(6) reliability: expected cost or loss, probability of failure; only on distance and load but also on geographic aspects relat-
(7) externalities: emissions, safety risks. ing to the most distant customer on a tour. Stenger, Vigo, Enz,
and Schwind (2013b) solve a variant with multiple depots, while
This section will discuss each of these criteria and the related Gahm, Brabänder, and Tuma (2017); Stenger, Schneider, and Goeke
VRP variants. For some applications, multiple criteria may appear (2013a) and Dabia, Lai, and Vigo (2019) consider nonlinear cost
as objectives (using a weighted sum, hierarchical or multi-objective functions arising from volume discounts. Finally, Goeke, Gschwind,
formulation) or as constraints. We analyze how each criterion has and Schneider (2019a) design a state-of-the-art branch-and-price
been integrated in academic problems, citing key methodological algorithm.
contributions and case studies.
2.2. Service quality
2.1. Profitability
Although operational efficiency is important, providing superior
It is safe to say that profitability or cost optimization is the pri- service quality helps businesses to differentiate themselves from
mary concern in the overwhelming majority of VRP studies. Most their competitors. Furthermore, profit measures are not the pri-
articles consider the minimization of total routing costs, which mary concern in some contexts, such as humanitarian relief op-
may include a fixed cost per route (e.g., vehicle cost, insurance, erations, public transportation, and home healthcare logistics.
daily wages) as well as variable costs proportional to distance or Cumulative objectives. In the cumulative VRP (Ngueveu, Prins, &
travel duration (e.g., fuel consumption, maintenance costs, hourly Calvo, 2010a; Silva, Subramanian, Vidal, & Ochi, 2012), the classi-
wages). Moreover, as outlined below, profitability also extends be- cal total cost objective is replaced with the sum of individual ar-
yond operational costs. rival times at the customers. Objectives of this type can be seen as
Performance ratios. In some situations, the optimization of rout- more service-focused, and they are often proposed as relevant op-
ing costs is not meaningful and can even be counter-productive if timization criteria for relief effort operations (Campbell, Vanden-
it is not balanced with other performance measures. Especially for bussche, & Hermann, 2008; Golden, Kovacs, & Wasil, 2014). The
problems posed on a rolling horizon, there is a need to consider components of a cumulative objective can also be weighted in or-
short-term surrogate objectives that approximate long-term perfor- der to further bias the route structure toward a customer-centered
mance goals. A practical example is the class of inventory-routing perspective (Huang, Smilowitz, & Balcik, 2012). In general, cumula-
problems, for which several authors have emphasized the need to tive objectives are more appropriate when the distribution of the
optimize the logistic ratio. This is the ratio of routing cost to deliv- arrival times, travel times, transported load, etc. is more important
ered quantity over the planning horizon (Archetti, Desaulniers, & than their sum.
Speranza, 2017b; Benoist, Gardi, Jeanjean, & Estellon, 2011; Song & Inconvenience measures. Service quality is particularly important
Savelsbergh, 2007), which measures the average cost to deliver one for passengers transportation. Common examples include the
unit of product. This objective prevents myopic behavior that could planning of school bus routes (Park & Kim, 2010) and dial-a-ride
arise from pure cost minimization (Archetti et al., 2017b). Another services organized by home healthcare providers (Cordeau &
practical example can be found in mobility-on-demand services Laporte, 2007). Since service quality is multi-dimensional, many
and in the maximization of the occupancy rate, i.e., the ratio of criteria have been proposed to measure its different aspects
total passenger travel times to total vehicle travel times (Garaix, (Paquette, Bellavance, Cordeau, & Laporte, 2012). From an op-
Artigues, Feillet, & Josselin, 2011). VRPs with other fractional ob- timization perspective, it is common to introduce measures of
jectives, such as profit over time, have been studied in Baldacci, customer inconvenience, for example by minimizing the maximum
Lim, Traversi, and Calvo (2018). ride time; the maximum time loss defined as the difference
Profit. Cost minimization often competes with profit maximiza- between the ride time and the corresponding shortest possible
tion in tactical business decisions. This is especially true when the ride; or the deviation from a preset time window in case of
 T. Vidal, G. Laporte and P. Matl / European Journal of Operational Research 286 (2020) 401–416 403

earliness or lateness. Importantly, the target levels of these criteria quality can also be interpreted as equity constraints. However,
may vary for different sets of customers. For example, emerging as discussed and analyzed by Huang et al. (2012) in the context
mobility-on-demand platforms aim to satisfy different service of disaster relief, there can be discrepancies between efficacy
quality thresholds for different customer segments (e.g., business, (quality of coverage) and equity, as these issues may concern
standard, budget) (Beirigo, Schulte, & Negenborn, 2019). different dimensions, e.g., the quantity of supplies may satisfy
Service levels. As a general trend worldwide, logistics activities the full demand at all the service points while the timeliness of
are being increasingly outsourced to third-party logistics (3PL) ser- delivery may be very inequitable, or vice versa. Moreover, we note
vice providers. Due to large volumes and unforeseen events, 3PL that unless the quality of the worst service is tightly constrained,
providers can rarely service all requests. To guarantee a certain satisfying the corresponding min-max constraint does not imply
service level, most 3PL providers establish contracts that stipu- an equitable distribution of service quality. To date, these issues
late a minimum ratio of on-time deliveries. This gives rise to the have received limited attention in the VRP literature.
VRP with service-level constraints (Bulhões, Há, Martinelli, & Vi- Collaborative planning. Due to strong competitive pressures and
dal, 2018a). In this problem, the deliveries are partitioned into falling profit margins in the logistics sector, carriers have an in-
groups, and a minimum percentage of the deliveries (or deliv- centive to form horizontal collaborations that pool their capacities
ery load) must be fulfilled for each group. Orlis, Laganá, Dullaert, and increase their overall efficiency (Cruijssen, Dullaert, & Fleuren,
and Vigo (2019) describe an application to automated teller ma- 2007; Gansterer & Hartl, 2018). In such coalitions, the planning
chine replenishment. In this study, the service levels are treated of logistics operations is performed jointly through the exchange
as soft constraints, and there are penalties for non-fulfillment. Ser- or consolidation of transportation requests and a redistribution of
vice fulfillment can even be the primary optimization objective in costs or gains, which leads to problems of fair resource allocation
some home healthcare applications (Rasmussen, Justesen, Dohn, & and profit sharing (Guajardo & Rönnqvist, 2016; Padilla Tinoco,
Larsen, 2012). Creemers, & Boute, 2017). The collaboration should be stable in the
sense that each partner’s individual cost is reduced by joining the
2.3. Equity partnership and the benefit of these reductions is fairly distributed.
Since many of the proposed cost allocation methods relate the
Efficient solutions are not necessarily equitable. Their accep- distributed cost or gain to the contributed resources (Guajardo
tance and implementation may be contingent on a sufficiently fair & Rönnqvist, 2016), the routing decisions help determining the
distribution of resources, responsibilities, and benefits among dif- achievable savings and the fairest benefit distribution.
ferent stakeholders. These concerns have led to a variety of equity
criteria in the routing literature, reviewed in Balcik, Iravani, and 2.4. Consistency
Smilowitz (2011) and Matl, Hartl, and Vidal (2018, 2019).
Workload balance. In routing problems in the private sector, the Cost-optimal routing plans may turn out to be of limited value
most common equity considerations concern internal stakeholders, if they vary too much over time. Customers appreciate being
i.e., the drivers or other personnel providing the service. The aim served by familiar faces at regular intervals; service providers are
is to balance the workload allocation in order to ensure accep- more effective and can personalize their service when they know
tance of operational plans, to maintain employee satisfaction and their customers’ requirements and preferences; drivers are more
morale, to reduce overtime, and to reduce bottlenecks in resource efficient and drive more safely when they are familiar with the pe-
utilization. Practical examples include balancing the workload of culiarities of their routes (Kovacs, Golden, Hartl, & Parragh, 2014a).
service technicians (Blakeley, Bozkaya, Cao, Hall, & Knolmajer, Establishing and maintaining these aspects of familiarity requires
2003), home healthcare professionals (Liu, 2013), and volunteers routing and service plans to be consistent with respect to various
(Goodson, 2014). Balancing criteria also appear in periodic settings metrics over multiple time periods.
(Groër & Golden, 2009; Mendoza, Medaglia, & Velasco, 2009) Temporal consistency. One aspect of consistency concerns the
and in tactical planning problems such as service territory design timing of the service provision to individual customers. The aim is
(Butsch, Kalcsics, & Laporte, 2014; Kalcsics, 2015). The workload to provide service at roughly the same time of day and at regular
Wr of a route or service unit is usually operationalized through intervals. Initial studies by Groër and Golden (2009), Tarantilis,
its total service duration, total demand, number of customers, Stavropoulou, and Repoussis (2012), and Kovacs et al. (2014a) han-
or some combination of these metrics. The degree of balance is dle this feature by imposing a maximum difference between the
then quantified by applying an inequality function to the vector latest and earliest arrival times at any customer location. The re-
of workloads, the most common functions being minimization of sulting consistent VRP (ConVRP) is often solved by metaheuristics,
the maximum workload (min {max {Wr }}) and the minimization of since the time constraints create route interdependencies which
the range (min{max{Wr } − min{Wr }}). Care should be taken when pose considerable challenges for exact solution approaches (Goeke,
modeling equity criteria, because certain combinations of work- Roberti, & Schneider, 2019b). Feillet, Garaix, Lehuede, Péton, and
load metrics and equity functions are not appropriate for guiding Quadri (2014) suggest an alternative approach that discretizes
mathematical optimization methods. In particular, equity functions the day into disjoint time segments and imposes consistency by
that are not monotonic with respect to all workloads (e.g., the bounding the number of different time segments during which a
range or standard deviation) can lead optimization methods to customer is served. Some recent works propose self-imposed time
unnecessarily increase the workload (e.g., longer distance or time) windows, whereby the service provider selects for each customer
of every route in an effort to artificially satisfy the ill-posed equity a fixed time window before the demand is known, communicates
criterion (Matl et al., 2018). this information to the customer, and subsequently generates
Service equity. In contrast to private and profit-oriented logis- routing plans respecting these commitments during the planning
tics businesses, public and nonprofit organizations also generally horizon (Jabali, Leus, Woensel, & de Kok, 2015; Spliet & Gabor,
have an obligation of equitable service provision to their external 2015). Other authors have set minimum and maximum time
stakeholders, i.e., the users or customers (Balcik et al., 2011). The intervals between consecutive customer visits in periodic settings
most common application areas are public transportation and (Coelho & Laporte, 2013; Gaudioso & Paletta, 1992).
humanitarian logistics. There exists a close connection between Person-oriented consistency. Another form of consistency relates
the service quality measures discussed in the previous section to the assignment of drivers to customers. If the same driver
and service equity. In fact, the min-max constraints on service regularly visits the same customers, the quality and efficiency of
404 T. Vidal, G. Laporte and P. Matl / European Journal of Operational Research 286 (2020) 401–416

the customer service improves as personal relationships become 2.5. Simplicity

established and the driver becomes more familiar with the cus-
tomers (Smilowitz, Nowak, & Jiang, 2013). This type of consistency In complex real-life systems, the acceptance and efficient re-
is particularly important in home healthcare logistics, where alization of vehicle routing plans often depends on their simplic-
the quality of the service depends on the nurses’ knowledge of ity and their intuitive appeal (Poot, Kant, & Wagelmans, 2002). As
the preference and needs of their patients (Eveborn, Flisberg, & non-experts in combinatorial optimization, drivers and dispatch-
Rönnqvist, 2006). Personal consistency is often handled at the ers may be reluctant to trust and implement solutions that ap-
tactical level by creating a fixed route for each driver (Beasley, pear overly complex and counter-intuitive or that require a high
1984; Christofides, 1971) or a set of template routes that are degree of coordination, even if these solutions are technically op-
adjusted into daily operational routes (Groër & Golden, 2009; timal with respect to cost. In routing contexts, visual appeal is of-
Kovacs, Parragh, & Hartl, 2014b; Tarantilis et al., 2012). More ten synonymous with compact and non-overlapping tours (Hollis
flexible alternatives focus on directly maximizing the number & Green, 2012). This corresponds to the two fundamental no-
of times a unique driver visits each customer (Haughton, 2007; tions of clustering in the machine learning literature (Jain, 2010):
Smilowitz et al., 2013), ensuring that each driver visits at least a group together elements in such a way that similar elements
certain fraction of their assigned customers (Spliet & Dekker, 2016) (nearby deliveries) belong to the same cluster (compactness) and
or bounding the number of different drivers serving any customer different elements (distant deliveries) belong to different clusters
(Braekers & Kovacs, 2016; Luo, Qin, Che, & Lim, 2015). (separation).
Regional consistency. In practice, and especially in urban con- Compactness. A route whose customers are geographically
texts, the efficiency of a route depends on the driver’s familiar- clustered is intuitive, because its compactness serves as a visual
ity with the addresses and buildings in the area, the typical traffic surrogate for low cost (Rossit, Vigo, Tohmé, & Frutos, 2019). This
conditions on important streets or junctions, possible shortcuts or intuition is typically exploited by optimization algorithms within
detours, etc. (Holland, Levis, Nuggehalli, Santilli, & Winters, 2017). the cluster-first, route-second category. In numerical terms, com-
As a result, it is desirable to maintain some form of regional con- pactness is usually optimized at the route level by minimizing a
sistency so that drivers can become familiar with their assigned measure of geographic spread. For example, Poot et al. (2002) min-
or most common service regions and hence benefit from the asso- imize the average pairwise distance between all customers in a
ciated learning effects (Haughton, 2002; Zhong, Hall, & Dessouky, route, or the average distance of the customers to the route’s cen-
2007). This can be seen as a generalization of person-oriented con- ter of gravity. A different definition of “route center” is proposed
sistency, and the previously mentioned fixed routes can also be a by Tang and Miller-Hooks (2006), who define the median of a
way to delineate fixed regions (Wong & Beasley, 1984). Similarly, tour to be the customer that minimizes the maximum distance
regional consistency can be enforced by constructing routes that to any other customer in the same tour. These ideas correspond
maximize the number of visited nodes within some threshold dis- to the more general k-means and k-median clustering approaches.
tance to fixed master routes (Sungur, Ren, Ordóñez, Dessouky, & Likewise, although route compactness measures concern the
Zhong, 2010), or maximize the number of times a driver repeat- operational level, they are closely connected to tactical decisions
edly visits the same region (Smilowitz et al., 2013). related to the design of compact distribution territories such as
Delivery consistency. In contexts such as vendor-managed in- those commonly used in postal deliveries (see Section 3.1). The
ventories (Day, Wright, Schoenherr, Venkataramanan, & Gaudette, corresponding compactness criteria are similar, e.g., minimizing
2009), it may be desirable to deliver a consistent quantity of the maximum distance of a customer to their territory’s (route’s)
materials or provide a consistent level of service. Since cost- center (Ríos-Mercado & Fernández, 2009), the maximum distance
minimizing solutions do not typically possess these properties, between any pair of customers in the same territory (Lin, Chin, Fu,
Coelho, Cordeau, and Laporte (2012) propose to constrain delivery & Tsui, 2017), or the ratio of the territory’s (route’s) perimeter to
quantities within lower and upper bounds, or to follow an order- the total perimeter of the service area (Lei, Laporte, Liu, & Zhang,
up-to policy. 2015). Other criteria are based on geometric ratios to ideal shapes
Inconsistency. Finally, inconsistency can be desired in some ap- like squares and circles (Kalcsics, 2015) or even temporal char-
plications. For cash-in-transit operations, the routes should be acteristics such as time-window differences (Schneider, Stenger,
unpredictable from day to day to reduce the risk of robberies Schwahn, & Vigo, 2015).
(Bozkaya, Salman, & Telciler, 2017; Constantino, Mourão, & Pinto, Separation. Routes that are geographically separate make coor-
2017). For the transportation of hazardous materials, safe backup dination easier, because local changes (e.g., unexpected demand)
routes should be available in case of adverse weather condi- do not impact the rest of the plan (Lum, Cerrone, Golden, & Wasil,
tions or to spread the accident risk geographically (Akgün, Erkut, 2017). Moreover, if the geographic separation is done at the tactical
& Batta, 20 0 0). In the m-peripatetic VRP, complete dissimilarity level, then processes such as sorting can be executed in parallel
is ensured by requiring alternative solutions to be edge-disjoint with routing, reducing delivery times and improving competitive-
(Ngueveu et al., 2010a; 2010b). The k-dissimilar VRP relaxes this ness (Janssens, Bergh, Sörensen, & Cattrysse, 2015). Unlike com-
constraint and minimizes the average ratio between the length pactness, separation measures are calculated at the solution level,
of edges shared by any pair of alternative solutions and the as they concern the relationship between multiple routes. They all
length of the routes containing those edges (Talarico, Sörensen, minimize some measure of overlap. Example metrics include the
& Springael, 2015). Martí, Velarde, and Duarte (2009) proposed number of customers that are closer to another route’s center or
a vertex-based dissimilarity measure, maximizing for each pair median than to their own (Poot et al., 2002; Tang & Miller-Hooks,
of paths the average distance between the vertices in one path 2006), the number of edges shared by two or more routes in an
and their closest neighbor in the other. Zajac (2016) considers arc routing context (Constantino, Gouveia, Mourão, & Nunes, 2015),
geographic dissimilarity explicitly by minimizing the intersection the number of customers contained in the convex hull of a route
of the geographic units visited in different solutions. Hoogeboom that is not their own (Poot et al., 2002), the average overlap of the
and Dullaert (2019); Michallet, Prins, Amodeo, Yalaoui, and Vitry routes’ convex hulls (Lum et al., 2017), the number of times dif-
(2014) and Soriano, Vidal, Gansterer, and Doerner (2019) enforce ferent routes cross paths (Poot et al., 2002), and others (Corberán,
temporal inconsistency by using time-window penalties when the Golden, Lum, Plana, & Sanchis, 2017). Although these metrics
arrival times of consecutive visits at the same customer do not dif- may initially appear somewhat ad hoc, they can have meaningful
fer by more than a given constant. properties. For example, Tang and Miller-Hooks (2006) show that
 T. Vidal, G. Laporte and P. Matl / European Journal of Operational Research 286 (2020) 401–416 405

if no customer is closer to another route’s median than to its own, These approaches significantly differ in how they model stochastic
then the convex hulls of the routes cannot overlap. parameters. Chance constraints still rely on distributional infor-
Navigation complexity. Routes should be easy to follow and exe- mation to evaluate and bound the probabilities of failure. This
cute. Distribution companies such as UPS prefer simple route struc- paradigm has been commonly used to solve VRPs with stochastic
tures so that drivers spend less effort on spatial route cognition travel times and time windows (Laporte et al., 1992; Li, Tian, &
and instead concentrate on driving safely (Holland et al., 2017). Leung, 2010). However, as highlighted in Errico, Desaulniers, Gen-
Users of consumer navigation systems prefer routes that are con- dreau, Rei, and Rousseau (2018), there is a thin line between model
cisely described and can be easily followed, especially when trav- assumptions that allow for efficient calculations (e.g., convolutions
eling through unfamiliar environments (Shao, Kulik, Tanin, & Guo, and dominance properties) and those that lead to intractable
2014). In practice, metrics for quantifying the navigation complex- problems. In contrast, robust models rely on an uncertainty set
ity of a route are commonly based on the number and type of (e.g., a polytope) to represent reasonable parameter variations
turns encountered. Turn restrictions and turn penalties frequently and seek solutions that are feasible for any parameter realization
arise in arc routing applications (Assad & Golden, 1995; Benavent & within this set (Ben-Tal, Ghaoui, & Nemirovski, 2009; Bertsimas,
Soler, 1999; Corberán, Martí, Martínez, & Soler, 2002; Vidal, 2017) Brown, & Caramanis, 2011). Robust models are especially useful in
and can be refined by considering different types of intersections situations where no complete distribution information is available,
as well as the road network hierarchy (Duckham & Kulik, 2003). and they are typically easier to solve than their chance-constrained
counterparts (Gounaris, Wiesemann, & Floudas, 2013; Pessoa, Poss,
2.6. Reliability Sadykov, & Vanderbeck, 2018b; Sungur, Ordóñez, & Dessouky,
2008). Since these models are completely risk-averse, research
Deterministic VRPs consider that all problem information is continues on alternative models of uncertainty (e.g., distribution-
available and accurate. However, data are always subject to approx- ally robust models) that are meaningful in practice and remain
imations, and unexpected events can render “optimal” determin- tractable (Jaillet, Qi, & Sim, 2016; Zhang, Baldacci, Sim, & Tang,
istic routing plans inefficient or impracticable. As a consequence, 2019).
finding reliable routing solutions that remain effective in the pres-
ence of uncertainty has become a major concern (Gendreau, Jabali, 2.7. Externalities
& Rei, 2014; 2016). Under uncertainty, a natural but cost-ineffective
strategy is to use a deterministic model to generate reliable solu- Although transportation is essential for modern businesses
tions that contain some slack (e.g., capacity or time). A better op- and society, it also has undesirable consequences (Demir, Huang,
tion is to exploit additional knowledge of the uncertain events, in Scholts, & Woensel, 2015). A more holistic optimization of logis-
the form of representative scenarios, probability distributions, or tics and mobility is needed to mitigate the impacts of externalities
uncertainty sets, giving rise to stochastic or robust VRP models. while maintaining efficient transportation systems.
Beyond a mere choice of objective function, defining a stochastic Emissions. Road transportation is a major contributor to increas-
VRP requires to specify when and how stochastic parameter values ing atmospheric pollution caused by greenhouse gases and partic-
are observed, and when decisions are taken. Two main groups of ulates. Reflecting also the broader societal concerns about sustain-
approaches can be distinguished: (1) stochastic programming mod- ability, the past decade has seen a rapid growth in studies falling
els, which typically focus on minimizing the expected cost of the under the class of green VRPs that account for emissions in the
routes and recourse actions made as a consequence of uncertain optimization model (Demir, Bektaş, & Laporte, 2014b). It has in-
events; and (2) chance constraints or robust formulations, which deed been recognized that classical cost-minimizing objectives (in
impose constraints on the failure probabilities. terms of distance or time) do not lead to minimal emissions or
Expected cost or loss. Stochastic models based on a priori opti- fuel consumption, although there is a correlation (Bektaş & La-
mization (Bertsimas, Jaillet, & Odoni, 1990) assume that the routing porte, 2011). A variety of fuel consumption models and solution
decisions are made in a first stage based on partial knowledge of methods have therefore been put forward (Demir, Bektaş, & La-
future events (before any stochastic parameters are observed), and porte, 2011; 2014b; Fukasawa, He, Santos, & Song, 2018; Kramer,
that prespecified recourse policies will be used in a second stage Subramanian, Vidal, & Cabral, 2015). Due to the complexity of the
when unexpected events occur (e.g., a direct return to the depot in emissions functions, optimization methods need to handle vari-
the case of excess demand). Most models in this family focus on ous factors, e.g., load-dependency (Kara, Kara, & Yetis, 2007), time-
optimizing the expected cost of first-stage routes and second-stage dependency (Franceschetti, Honhon, Woensel, Bektaş, & Laporte,
recourse actions. There are three main sources of uncertainty: cus- 2013; Jabali, Woensel, & de Kok, 2012b), heterogeneous fleets (Koç,
tomer demands (Bertsimas, 1992), service requests (Jaillet, 1988), Bektaş, Jabali, & Laporte, 2015), and modal choice (Bauer, Bektaş,
and travel times (Laporte, Louveaux, & Mercure, 1992). Yet, de- & Crainic, 2010). Although direct speed optimization is difficult to
spite considerable algorithmic progress over the last four decades, plan for road-based operations, it is an important concern and
the solution methods (metaheuristics and mathematical program- easier to achieve in maritime transportation (Fagerholt, Laporte,
ming methods alike) are limited by the necessity to evaluate the & Norstad, 2009; Hvattum, Norstad, Fagerholt, & Laporte, 2013;
expected cost of the recourse actions. Therefore, strong assump- Norstad, Fagerholt, & Laporte, 2011). From a practical perspective,
tions are typically made to keep the evaluations tractable: simplis- it is worth noting that by allowing a small increase in distance
tic recourse policies are used, and the probability distributions as- or time, one can significantly reduce emissions, which motivates
sociated with the random events are assumed to be independent. the consideration of fuel consumption as a side objective (Demir,
Ongoing research is exploring more sophisticated recourse policies Bektaş, & Laporte, 2014a).
(Ak & Erera, 2007; Louveaux & Salazar-González, 2018; Salavati- Safety risks. When transporting hazardous materials (hazmat)
Khoshghalb, Gendreau, Jabali, & Rei, 2019; Yang, Mathur, & Ballou, such as nuclear waste, chemical agents, or noxious gases, risk mit-
20 0 0), correlated random events (Rostami, Desaulniers, Errico, & igation is a priority. Since the degree of risk and the severity of a
Lodi, 2017), and multiple decision stages (Dror, Laporte, & Trudeau, potential accident are closely related to the selected route, clas-
1989; Goodson, Ohlmann, & Thomas, 2013). sical VRP models must be carefully extended to properly incor-
Risk of failure. Models based on chance constraints or robust porate various aspects of risk. For example, Tarantilis and Kira-
optimization impose constraints on the probability of failure as noudis (2001) consider population exposure risk on each link of
opposed to optimizing the expected cost of the uncertain events. the network, Ma, Cheang, Lim, Zhang, and Zhu (2012) propose the
406 T. Vidal, G. Laporte and P. Matl / European Journal of Operational Research 286 (2020) 401–416

inclusion of link-specific risk capacities, and Taslimi, Batta, and that define the district boundaries through geometric arguments,
Kwon (2017) examine a bilevel problem in which a regulator de- as in Carlsson (2012) and Carlsson and Delage (2013). One obvious
cides which links to close for hazmat transportation while consid- advantage of the first technique is that it lends itself to the use
ering the expected alternative routes then chosen by the hazmat of local search-based metaheuristics in which basic units are itera-
carriers. Accident risk is also considered along the temporal di- tively relocated or swapped between adjacent districts and allows
mension by Meng, Lee, and Cheu (2005) and Toumazis and Kwon efficient evaluations of the objective function.
(2013), who consider time-dependent risk models, and by Zografos We focus on applications in which a traveling salesman prob-
(2004), who examines the trade-off between travel time and risk. lem (TSP) or VRP must eventually be solved within each district. A
Note that some hazmat VRP models can be generalized to different common case arises in the planning of sales districts, where each
types of undesirable externalities, such as noise, disturbance, and district is assigned to a vendor or a team of vendors. When de-
pedestrian safety (Bronfman, Marianov, Paredes-Belmar, & Lüer- signing the districts, one must take into account the routing cost
Villagra, 2015; Grabenschweiger, Tricoire, & Doerner, 2018). Finally, and also ensure a level of equity between the routes of different
consumer-oriented routing applications optimize safety from the districts. If, as is usually the case, a local search technique is used
opposite perspective, aiming to generate routes that are safe for to optimize the districts, it can be prohibitively long to optimize
the user (Kim, Cha, & Sandholm, 2014; Shah, Bao, Lu, & Chen, the vehicle routes associated with the districts at each step (i.e.,
2011). move evaluation). To circumvent this issue, most solution meth-
ods rely on closed-form formulas to approximate the routing costs
3. Integrated problems – routing as an evaluation tool without actually determining the routes. Two such formulas are
the Beardwood, Halton and Hammersley (1959) (BHH) formula for
Vehicle routing decisions are fundamentally operational but the TSP and the Daganzo (1984) formula for the VRP. The BHH
are often linked with other decisions taken at a strategic or formula approximates the routing cost through n independently
tactical level over a longer planning horizon (Crainic, 2002). In and
√ identically distributed points in a compact area of size A as
such contexts, generating VRP solutions or at least evaluating their β nA, where β is a constant. Appropriate constant values are pro-
characteristics becomes essential to evaluate the cost of plan- vided in Applegate, Bixby, Chvátal, and Cook (2011). Combining this
ning decisions made at a higher level, which can be districting, formula with a simple geometrical partitioning strategy, Daganzo √
facility location, fleet composition, or inventory and production (1984) approximates the cost of a VRP solution as 2rm + 0.57 nA,
management. Two main approaches are typically used: continuous where m is the number of vehicle tours and r is the average dis-
approximation and regression models (Franceschetti, Jabali, & tance between the depot and the barycenters of the districts. The
Laporte, 2017b), or fast versions of VRP algorithms adapted to first term in this expression represents the “line-haul” distance to
stochastic or scenario-based problem variants. The former aims reach the districts, and the second term measures the routing costs
to give a good estimate of the routing costs based on geometric within the districts. Continuous approximation formulas are still
considerations, while the latter samples demand patterns resulting being refined and generalized (see, e.g., Çavdar & Sokol, 2015; Mer-
from distribution or scenario information. While stochastic and chán & Winkenbach, 2019), and approaches based on regression
scenario-based approaches offer greater precision, they generally or neural networks (see, e.g., Kwon, Golden, & Wasil, 1995) may
lead to large-scale integrated problems which challenge the capa- soon achieve even better trade-offs between estimation accuracy
bilities of current solution methods. This section surveys the main and computational effort.
applications and methods arising in integrated two-level problems
of which routing is one of the components. 3.2. Routing and facility location

3.1. Routing and districting Many applications require the evaluation of routing costs dur-
ing the facility location decisions. This has led to the development
Districting is the process of partitioning a territory for polit- of a vast literature dedicated to combined location and routing
ical, administrative or commercial purposes (Kalcsics, 2015). The problems (Laporte, Nickel, & Gama, 2015; Prodhon & Prins, 2014;
best-known application is the design of political districts (see, e.g., Schneider & Drexl, 2017). Facility location decisions are strategic
Bozkaya, Erkut, & Laporte, 2003), but logistics applications are also or tactical in most applications. They concern warehouses, cross-
common. These include the design of sales territories (Drexl & docks, or satellite facilities in city logistics, whereas vehicle routes
Haase, 1999; Lei et al., 2015; Skiera & Albers, 1998) and distri- are operational decisions that can change dynamically over time. In
bution management applications, for example those encountered these contexts, continuous approximation formulas can be used to
in mail delivery systems (Bruno, Cavola, Diglio, Laporte, & Piccolo, estimate the routing cost (Campbell, 1990; Laporte & Dejax, 1989;
2019; Novaes, de Cursi, & Graciolli, 20 0 0; Rosenfield, Engelstein, Ouyang & Daganzo, 2006; Xie & Ouyang, 2015), and facility catch-
& Feigenbaum, 1992). Fixed districts ensure regional consistency ment areas may be represented as polygons in a Voronoi diagram
and facilitate delivery operations (see Section 2.4). Districting plans (Laporte & Dejax, 1989). Continuous approximation methods can
are typically subject to hard constraints such as contiguity as well also be extended to integrate a variety of constraints and objec-
as soft constraints such as size, compactness, population balance, tives, such as backbone costs in hub networks (Campbell, 2013;
homogeneity, and fairness. These soft constraints, which are often Carlsson & Jia, 2013; 2015).
nonlinear, are eventually aggregated into a multi-criteria objective Another approach for location and routing is to rely on Monte
function with suitable user-defined weights. The districts are often Carlo scenario generation as a basis for routing cost evaluations
expected to change over time because of population shifts, for ex- (Klibi, Lasalle, Martel, & Ichoua, 2010). This approach, however, can
ample. In such cases, robustness with respect to future stochastic lead to challenging scenario generation and optimization problems.
or dynamic changes is also deemed to be a desirable property (Lei, This may explain why most studies on combined location and
Wang, & Laporte, 2016). routing problems have opted for a deterministic “single routing
There are two main techniques for constructing districts. The scenario” approach, giving rise to the canonical location-routing
most common aggregates cells, usually called basic units, for which problem (LRP), recently surveyed in Schneider and Drexl (2017).
geographic, demographic or socio-economic data are available. It The LRP model is mainly relevant in contexts where the delivery
is common to define basic units as census tracts (Bozkaya et al., routes are fixed over a long time, or where both location and rout-
2003). A second technique divides a planar area by drawing lines ing decisions are operational, e.g., when locating transfer points
 T. Vidal, G. Laporte and P. Matl / European Journal of Operational Research 286 (2020) 401–416 407

between two vehicles or vehicle reception points such as tempo- sition is taking place between transporter-managed and crowd-
rary parking places and postal boxes (Boudoin, Morel, & Gardat, sourced delivery systems. Crowdsourcing involves paying daily
2014). The canonical LRP represents a challenge for exact algo- commuters and ad hoc drivers for last-mile deliveries in an effort
rithms (Baldacci, Mingozzi, & Calvo, 2011; Contardo, Cordeau, & to use their residual capacity, leading to a new generation of tac-
Gendron, 2014), since these approaches must consider many candi- tical fleet composition and multi-modal transportation problems
date subsets of locations. In contrast, metaheuristics currently pro- (Archetti, Savelsbergh, & Speranza, 2016; Arslan, Agatz, Kroon, &
duce good solutions for large-scale instances (Schneider & Löffler, Zuidwijk, 2019; Cleophas, Cottrill, Ehmke, & Tierney, 2019; Mourad,
2019). In the future, these methods may be extended to sophis- Puchinger, & Chu, 2019).
ticated settings with multiple routing scenarios in an attempt to
improve the accuracy and applicability of tactical location routing
models.
3.4. Routing, inventory, and production management
3.3. Routing and fleet composition
Inventory-routing problems (IRPs) arise in the context of
Tactical fleet sizing and composition problems occur across all vendor-managed inventory management in which a supplier
transportation modalities, when renewing vehicles, adapting to jointly optimizes vehicle routes, delivery schedules, and quantities.
market fluctuations, and evaluating business changes (e.g., com- The field is rooted in the work of Bell et al. (1983) and has since
pany mergers). Fleet size adjustments can be done via long-term seen a phenomenal growth, discussed in the survey of Coelho,
vehicle acquisitions and sales or short-term leasing. Typical plan- Cordeau, and Laporte (2014). Multiple versions of the problem ex-
ning horizons vary among applications: horizons are generally ist, varying in the planning horizon (finite or infinite), the delivery
longer in maritime operations than in land-based transportation structure (1-1, 1-M, M-M, or 1-M-M-1: see Section 4.3), the rout-
because of the long lifetime of ships and the large capital costs in- ing patterns (back and forth routes or multi-customer routes), the
curred (Hoff et al., 2010). As a result, maritime fleet sizing models inventory policy (maximum level or up-to-order), the inventory
usually consider fixed trade lanes for strategic planning (Pantuso, decisions (lost sales or backlogging), the fleet composition (homo-
Fagerholt, & Hvattum, 2014; Wang, Fagerholt, & Wallace, 2018). For geneous or heterogeneous), and the fleet size (single, multiple, or
land-based transportation, two main approaches are generally used unconstrained). Since the planning horizons tend to be shorter in
to evaluate the routing costs within fleet composition models: con- IRPs than in the other strategic problems discussed in this section,
tinuous approximations or (multi-period or stochastic) heteroge- a larger part of the literature combines inventory management
neous VRP solution methods. with route generation within integrated VRP models, although
Continuous approximation models stem from the observation continuous routing-cost approximations are also sometimes used
that it is difficult to obtain accurate demand scenarios and even (Baller, Dabia, Dullaert, & Vigo, 2019). Most models are defined on
harder to solve the resulting VRPs. Time-consuming route eval- a rolling horizon, so the choice of objective is nontrivial. In par-
uations can therefore be avoided by the use of approximation ticular, optimizing the logistic ratio (Archetti et al., 2017b) can be
formulas to focus the optimization on the fleet sizing decisions better in practice than pure cost minimization. IRPs are notoriously
(Campbell, 1995; Franceschetti, Honhon, Laporte, Woensel, & Fran- challenging for exact methods (Desaulniers, Rakke, & Coelho, 2016),
soo, 2017a; Jabali, Gendreau, & Laporte, 2012a; Nourinejad & Ro- but there are efficient hybrid metaheuristics (see, e.g., Archetti,
orda, 2017). Boland, & Speranza, 2017a; Chitsaz, Cordeau, & Jans, 2019).
Heterogeneous VRP models, in contrast, require the joint de- As discussed in the surveys of Christiansen, Fagerholt, Nygreen,
termination of vehicle types and routes. Each vehicle type may and Ronen (2013) and Papageorgiou, Nemhauser, Sokol, Cheon, and
possess distinct characteristics, e.g., capacity, fixed and variable Keha (2014), IRPs have often been applied to maritime routing,
costs, customer-service restrictions, or even specific travel costs particularly for the transportation of liquefied natural gas (see,
and speeds. Two canonical problems are generally distinguished: e.g., Andersson, Christiansen, & Desaulniers, 2016; Ghiami, Demir,
the fleet size and mix VRP (FMVRP) and the heterogeneous fixed Woensel, Christiansen, & Laporte, 2019; Halvorsen-Weare & Fager-
fleet VRP (HFVRP). The FMVRP assumes that an unlimited num- holt, 2013; Stålhane et al., 2012). Another important application is
ber of vehicles of each type is available, whereas maximum limits the transportation of perishable products (see, e.g., Coelho & La-
are set in the HFVRP. As illustrated in the survey of Koç, Bektaş, porte, 2014; Crama, Rezaei, Savelsbergh, & Woensel, 2018). More-
Jabali, and Laporte (2016), research on heterogeneous VRPs is ex- over, recent years have seen the emergence of inventory-routing
tensive but usually focused on a single period in the presence problems related to the management of shared mobility systems,
of a fixed set of customer requests. This case corresponds to ap- mostly in the case of bikes and cars. In these challenging prob-
plications in which the fleet is already acquired (or rented for a lems, one must simultaneously optimize the inventory levels at the
short term) or where the demand is stable over a long time pe- stations and the itineraries used to reposition the shared vehicles
riod. Kilby and Urli (2016), Pasha, Hoff, and Hvattum (2016), and (Chemla, Meunier, & Calvo, 2013; Laporte, Meunier, & Calvo, 2018).
Bertoli, Kilby, and Urli (2020) have recently extended the FMVRP Finally, supply chain integration extends well beyond inven-
to multi-period and stochastic settings, helping to bridge the gap tory routing. As demonstrated by Chandra (1993) and Chandra and
between the heterogeneous VRP and its tactical fleet composition Fisher (1994), the joint optimization of routing, inventory manage-
applications. ment, and production can lead to substantial savings over a se-
Finally, the emergence of vehicles with alternative fuels and quential approach. The resulting production-routing problem (PRP)
the growing focus on (locally) emission free deliveries have led aims to coordinate a production schedule with product deliver-
to new fleet composition problems involving battery-powered and ies at customer locations (Adulyasak, Cordeau, & Jans, 2015). Re-
conventional vehicles (Felipe, Ortuño, Righini, & Tirado, 2014; Hier- cent algorithms and case studies are presented in Absi, Archetti,
mann, Puchinger, Ropke, & Hartl, 2016; Pelletier, Jabali, & Laporte, Dauzére-Pérés, and Feillet (2015); Adulyasak, Cordeau, and Jans
2016). Cities around the world are gradually restricting the vehi- (2014); Neves-Moreira, Almada-Lobo, Cordeau, Guimarães, and Jans
cle types allowed in city centers. To cope with these challenges, (2019) and Qiu, Qiao, and Pardalos (2019). Ongoing research is con-
there have been studies of fleet composition models with city cen- sidering integrating a wider set of supply chain decisions, e.g., as-
ter restrictions (Davis & Figliozzi, 2013; Franceschetti et al., 2017a; sembly, production, inventory, and routing (Chitsaz et al., 2019), or
Hiermann, Hartl, Puchinger, Schiffer, & Vidal, 2019a). Another tran- production, location, and inventory (Darvish & Coelho, 2018).
408 T. Vidal, G. Laporte and P. Matl / European Journal of Operational Research 286 (2020) 401–416

4. Refined problems – precise and applicable plans recent survey of Rincon-Garcia, Waterson, and Cherrett (2018) re-
ports that the inadequate management of time-dependent travel
In parallel with studies that concern the integration of VRP times in routing software remains a major barrier to application.
models with other tactical supply chain decisions, significant re- Time-dependent effects are commonly modeled via travel-time or
search is being conducted to refine the models and integrate fine- travel-speed functions (Gendreau et al., 2015). Furthermore, the
grained problem attributes that can have a large impact on solu- speed on an arc may be computed at its entry time (frozen link
tion quality and feasibility. This section reviews some important model) or may vary on the arc as time passes (elastic link model).
problem refinements in relation to the transportation network, the In an elastic link model with strictly positive speeds, the FIFO
drivers and vehicles, and the customer requests. property is always satisfied, i.e., a later departure leads to a later
arrival time (Orda & Rom, 1990). This model has been used in the
4.1. Specificities of the transportation network seminal work of Ichoua, Gendreau, and Potvin (2003).
Most studies on TDVRPs rely on a complete graph represen-
Arc attributes. Transportation networks are usually character- tation of the network in which each origin-destination pair is
ized by multiple attributes, including driving time, driving cost represented by a single link and travel time function. In prac-
(and tolls), transportation mode, attractiveness, safety, emissions, tice, however, the time-dependent travel times are specific to
and energy consumption. In these conditions, a single best path each street or neighborhood of an urban network. To account
may not be readily definable between each origin and destina- for this, some studies have defined time-dependent speed func-
tion, and several trade-off paths should be considered. For ex- tions at the network level (Huang, Zhao, Woensel, & Gross, 2017;
ample, the canonical VRP with time windows has been exten- Maden, Eglese, & Black, 2009; Vidal, Martinelli, Pham, & Há,
sively studied with the fundamental assumption that one time 2019). It is important to note that most existing vehicle rout-
unit corresponds to one cost unit. In such situations without any ing heuristics can be adapted to the TDVRP under the condi-
trade-off, the search can be limited to a single shortest path for tion that a fast mechanism is available for time-dependent travel
each origin and destination. However, time and cost are not di- time queries. Yet, despite the development of sophisticated quick-
rectly proportional in real transportation networks: these resources est path algorithms (Bast et al., 2016; Batz, Geisberger, Sanders, &
can even be negatively correlated when tolls or access restric- Vetter, 2013), producing accurate speed predictions and perform-
tions are imposed (Reinhardt, Jepsen, & Pisinger, 2016). Research ing rapid travel-time queries on large-scale networks (typically
on this topic is fairly recent. Accounting for these effects gives within a fraction of a millisecond) raise significant methodological
rise to a class of VRPs on multi-graphs (Ben Ticha, Absi, Feil- challenges.
let, & Quilliot, 2017; 2018; 2019; Hiermann et al., 2019a; Soriano Access restrictions. Turn restrictions, delays at intersections, tolls,
et al., 2019) linked to critical applications in multi-modal trans- and limited parking availability are a significant part of the reality
portation, long-haul transportation, and city logistics, among oth- of urban logistics. The inadequate management of these aspects is
ers (Caramia & Guerriero, 2009; Garaix, Artigues, Feillet, & Josselin, another important barrier to the application of routing software
2010). Solution methods must jointly optimize the visit sequences in practice (Rincon-Garcia et al., 2018). Nielsen, Frederiksen, and
and the paths between them. In the worst case, the number of Simonsen (1998) estimate that turns and delays at intersections
trade-off paths between any two points grows exponentially. Still, represent 30% of the total transit time in cities, so an accurate
empirical analyses have shown that this number remains small model of turn restrictions is critical for mail delivery, waste col-
in practice for transportation networks with time-window con- lection, snow plowing, and street maintenance operations, among
straints (Ben Ticha et al., 2017; Müller-Hannemann & Weihe, 2006), others (Irnich, 2008; Perrier, Langevin, & Amaya, 2008). Likewise,
and the set of paths could otherwise be heuristically restricted an excessive number of turns in warehouse operations can lead
(Hiermann et al., 2019a). to increased chances of vehicle tipovers, congestion, and collisions
Two-echelon structures. Studies on distribution networks pos- (Çelik & Süral, 2016).
sessing a two-echelon structure can be traced back to the work Accounting for these detailed effects is not straightforward. In
of Jacobsen and Madsen (1980), in which intermediate facilities the case of turn restrictions, for example, joining turn-feasible
are used to transfer newspapers from large vehicles to smaller shortest paths may still lead to forbidden turns at their junc-
ones. Nowadays, as reviewed in Cuda, Guastaroba, and Speranza tions. Solution approaches for such problem variants rely on graph
(2015) and Guastaroba, Speranza, and Vigo (2016), e-retailers com- transformations (Clossey, Laporte, & Soriano, 2001; Corberán et al.,
monly adopt a two-echelon structure to deliver orders from dis- 2002; Vanhove & Fack, 2012) or exploit a mode selection subprob-
tribution centers to cross-docking facilities for consolidation, and lem to optimize the arrival direction at each service location dur-
thence to customers. In city logistics, transfer points are typi- ing route evaluations (Vidal, 2017). Exact algorithms may require
cally located on the outskirts of cities to reduce noise, pollu- dedicated pricing and cut separation procedures to consider costs
tion and traffic (Soysal, Bloemhof-Ruwaard, & Bektaş, 2015). Re- based on consecutive edge pairs (Martinelli & Contardo, 2015). The
search on two-echelon VRPs is now very active since the joint limited amount of space in city centers also leads truck drivers to
optimization of two route levels and the related time constraints rely on double parking. Some recent studies have modeled the im-
and synchronization issues pose substantial methodological chal- pact of such practices (Figliozzi & Tipagornwong, 2017; Morillo &
lenges (Grangier, Gendreau, Lehuédé, & Rousseau, 2016). We refer Campos, 2014), yet parking considerations remain largely unrepre-
to Breunig, Schmid, Hartl, and Vidal (2016) and Marques, Sadykov, sented in VRP models.
Deschamps, and Dupas (2019) for state-of-the-art heuristic and ex-
act algorithms. 4.2. Specificities of drivers and vehicles
Congestion and time dependency. Congestion is a major factor in
city logistics, since it causes massive economic losses (400 billion Heterogeneous vehicles and delivery modes. As discussed in
dollars per year in the United States according to Cookson & Pishue Section 3.3, vehicle fleets are rarely homogeneous (Koç et al., 2016;
2018) and has numerous negative effects. As noted in the survey of Pantuso et al., 2014). Individual vehicle specificities (e.g., variable
Gendreau, Ghiani, and Guerriero (2015), VRPs with time-dependent costs, specific equipments, or access restrictions) must often be
travel times (TDVRPs) may arise as a consequence of congestion, explicitly considered to obtain accurate operational plans, giving
weather conditions, road closures, roaming targets, and other fac- rise to FMVRP and HFVRP variants. Many efficient metaheuris-
tors. TDVRPs have been the focus of extensive research, but the tics and exact algorithms have been proposed for these prob-
 T. Vidal, G. Laporte and P. Matl / European Journal of Operational Research 286 (2020) 401–416 409

lems (see, e.g., Koç et al., 2015; Penna, Subramanian, Ochi, Vi- used to directly filter some feasible or infeasible loads. Moreover,
dal, & Prins, 2019; Pessoa, Sadykov, & Uchoa, 2018a; Vidal, Crainic, the loading constraints go well beyond the search for a feasible
Gendreau, & Prins, 2014). Recent studies have extended the scope packing of items: some applications require precedence constraints
of heterogeneous VRPs to multi-modal transportation systems in- (e.g., LIFO or FIFO) between services to make unloading possible
volving bikes, scooters, vans, as well as alternative propulsion (Cordeau, Iori, Laporte, & González, 2009) or integrate handling
modes (Felipe et al., 2014; Hiermann, Hartl, Puchinger, & Vidal, constraints for on-board load rearrangement (Battarra, Erdoǧan, La-
2019b; Nocerino, Colorni, Lia, & Lué, 2016). Beyond this, the recent porte, & Vigo, 2010), while other applications, e.g., for hazardous
growth of e-commerce has given rise to new distribution prac- materials or food transportation, impose incompatibility or sepa-
tices, including the use of drones. In the simplest case, drones ration constraints (Battarra, Monaci, & Vigo, 2009; Hamdi-Dhaoui,
make back-and-forth deliveries from a warehouse to customer lo- Labadie, & Yalaoui, 2014). The loading area can also be unique, split
cations. More sophisticated distribution modes involve the com- into different compartments (Derigs et al., 2010), or even separated
bined use of delivery vehicles and drones. For example, Murray into a truck and a trailer (Villegas, Prins, Prodhon, Medaglia, & Ve-
and Chu (2015), Dorling, Heinrichs, Messier, and Magierowski lasco, 2013). The trailer can be parked and retrieved to facilitate
(2017), Poikonen, Wang, and Golden (2017), and Agatz, Bouman, access to some customers, leading to two-echelon problem vari-
and Schmidt (2018) consider a delivery configuration in which a ants.
drone, mounted on a vehicle, detaches itself to perform deliveries Recharging stops. There has been a rapid growth of research
while the vehicle keeps moving. The resulting problems are grad- into VRP variants for battery-powered electric vehicles. Because
ually giving rise to a rich research area. of their limited range, early electric models often required en
Working hours regulations. Hours-of-service (HOS) regulations route recharging stops, and these intermediate stops (Schiffer,
are ubiquitous, and they should be taken into account when long- Schneider, Walther, & Laporte, 2019) became a defining feature of
haul routes are generated for several days or weeks. Transporta- most electric VRPs (EVRP). The EVRP literature has quickly grown
tion companies, in particular, have the responsibility of ensuring to take into account the numerous characteristics of real appli-
that driving plans can be safely performed with regulatory break cations. Studies have been conducted on EVRPs with heteroge-
and rest periods. Typical HOS regulations in the United States, the neous fleets and charging infrastructure (Felipe et al., 2014; Hi-
European Union, Canada, and Australia impose daily and weekly ermann et al., 2019b; Hiermann et al., 2016); more realistic en-
rest periods as well as limits on the driving and working hours. ergy consumption functions (Goeke & Schneider, 2015) and charg-
Their numerous clauses, conditions, and exclusions make it ex- ing profiles (Keskin & Çatay, 2016; Montoya, Guéret, Mendoza,
tremely difficult to check that a compliant schedule exists, even & Villegas, 2017); limited charging capacity (Froger, Mendoza, Ja-
for a fixed sequence of visits. Prescott-Gagnon, Desaulniers, Drexl, bali, & Laporte, 2017); and time-dependent energy costs (Pelletier,
and Rousseau (2010) and Goel and Vidal (2014) have studied these Jabali, & Laporte, 2018). As battery technology progresses, the
rule sets for different countries and proposed efficient routing and range of electric vehicles is becoming sufficient for daily deliv-
scheduling algorithms. The latter study, in particular, used the op- ery operations in metropolitan areas. Therefore, en route recharg-
timized routing plans to compare various regulations in terms of ing is gradually disappearing from these applications. It may still
their impact on drivers’ fatigue. be necessary for lightweight vehicles (e.g., drones) or vehicles
HOS regulations also extend beyond classical single-driver day performing round-the-clock operations. Also note that the lim-
operations, and specific provisions exist for night work (Goel, ited supply of some materials (e.g., rare earths) and the lack
2018) and team-driving (Goel, Vidal, & Kok, 2019). Schiffer, Laporte, of a good recycling process can limit the availability of large
Schneider, and Walther (2017) recently highlighted the benefits of batteries (Hwang, Myung, & Sun, 2017), so the development of
jointly planning rest periods and recharging actions for electric ve- a more efficient recharging infrastructure remains a plausible
hicles. A key challenge of HOS regulations relates to the purposeful scenario.
use of optimization: transportation companies should verify that a
feasible schedule exists, but most decisions on break and rest pe- 4.3. Specificities of customer requests
riods lie with the drivers. In such situations, a simple simulation
of driver behavior may be more reliable than a full-blown opti- Some customer-request specificities arising in the form of
mization algorithm considering all regulatory aspects and excep- customer-oriented objectives have been discussed in Section 2.
tions. Beyond this, there is a thin line between regulatory aspects Here we discuss other aspects of customer requests that do not
that can be optimized and those that should be used as a recourse arise as an optimization goal but are nevertheless essential for use-
when facing unforeseen events (e.g., the extended driving time de- ful routing plans.
fined by regulation (EC) 561/2006 should likely be kept as a re- Service types. VRP applications can involve very different ser-
course). vice types, depending on the number of commodities involved and
Loading constraints and compartments. Trucks, ships, and air- on the origin and destination points (depot or customer location).
planes have many specific load restrictions which must be taken Four main types can generally be distinguished:
into account during optimization (Pollaris, Braekers, Caris, Janssens,
& Limbourg, 2015). The papers considering these aspects are pri- • 1-M-1 (including 1-M and M-1). One-to-many-to-one problems
marily classified by geometry, e.g., pallet loading (Pollaris, Braek- include depot-to-customer and customer-to-depot transporta-
ers, Caris, Janssens, & Limbourg, 2017), 2D packing (Iori, González, tion as special cases. Applications of 1-M-1 services arise, e.g.,
& Vigo, 2007), and 3D packing constraints (Gendreau, Iori, Laporte, in small-package delivery, where deliveries are made early in
& Martello, 2006), but other constraints related to fragility, orienta- the routes and are followed by pickups later in the day (Holland
tion, or equilibrium often come into play. Specialized applications et al., 2017).
such as car hauling require dedicated feasibility-checking mecha- • 1-1. One-to-one problems represent transportation settings in
nisms to ensure that a load can be feasibly placed on the truck which each service is unique and associated with a fixed ori-
(Dell’Amico, Falavigna, & Iori, 2015) and that axle-weight limits gin and destination. A typical application is taxi fleet operations
are respected (Pollaris et al., 2017). Many of these VRP variants (Doerner & Salazar-González, 2014).
share the common trait that load-feasibility checking, even for a • M-M. Many-to-many problems involve one or several resources
fixed route, is an NP-hard problem. To speed up this critical evalua- at multiple locations. The goal is to move some of these re-
tion step, a variety of packing heuristics, bounds, and rules may be sources toward the locations where they are most needed. Typ-
410 T. Vidal, G. Laporte and P. Matl / European Journal of Operational Research 286 (2020) 401–416

ical applications concern bike repositioning (Bulhões, Subrama- constraints; these are notoriously difficult to solve (Drexl, 2012).
nian, Erdogan, & Laporte, 2018b; Chemla et al., 2013) and lateral Synchronization may even be imposed between different techni-
transshipments (Hartl & Romauch, 2016; Paterson, Kiesmüller, cians at different locations, as in an application to electric network
Teunter, & Glazebrook, 2011). recovery studied by Goel and Meisel (2013). In this setting, a local
• 1-M-M-1. Finally, some applications may involve a combination change in one route can impact the entire daily schedule, violating
of the M-M and 1-M-1 cases. One such problem was investi- the synchronization constraints or delaying other routes.
gated by van Anholt, Coelho, Laporte, and Vis (2016) in the con-
5. Challenges and prospects
text of the replenishment of automated teller machines: money
has to be transferred from a central office to automated tellers, As we have shown, extensive research has been conducted over
among these tellers, and back to the office. 60 years to better connect vehicle routing models and application
cases. This close proximity between academic research, software
As noted in Battarra, Cordeau, and Iori (2014), the first two cat- companies, and transportation actors has led to a multitude of suc-
egories of problems (1-M-1 and 1–1) have been extensively dis- cessful applications (see, e.g., Hall & Partyka, 2018; Toth & Vigo,
cussed in the VRP literature. In contrast, studies on M-M or 1- 2014). Nonetheless, vehicle routing research is far from a closed
M-M-1 settings, especially with multiple commodities, are not as topic. Technologies and business models evolve at a rapid pace.
common. The continuing growth of e-commerce and home deliveries, in-
Applications also differ in terms of whether or not split ship- creased access to on-demand transportation via mobility applica-
ments are allowed. Split loads (Archetti & Speranza, 2012) typi- tions, and ongoing urbanization have put city transportation net-
cally occur when transporting many units of the same commod- works and supply chains under an unprecedented strain. To meet
ity (e.g., bikes) or when delivering or collecting a divisible product these challenges, companies and governing authorities seek true
(e.g., food or liquids). In such situations, a customer may be vis- shifts of transportation paradigms rather than incremental opti-
ited multiple times in order to fulfill its request. The resulting split mizations of existing systems. These changes may be linked to new
delivery VRP is a relaxation of the capacitated VRP (CVRP) but is transportation modes, e.g., drones (Poikonen et al., 2017) or au-
more complicated to solve. Because of a lack of an efficient route- tonomous vehicles (Fagnant & Kockelman, 2015), or to the redesign
based decomposition, due to the customer demands which act as of business models and supply chains, e.g., crowdsourced deliveries
linking constraints, exact methods (Archetti, Bianchessi, & Sper- (Arslan et al., 2019) or the physical Internet (Montreuil, 2011). Re-
anza, 2014a) struggle to optimally solve instances of a size (e.g., gardless of the technology adopted, whereas products, drivers, and
50–100 deliveries) that can easily be handled in a canonical CVRP customers were typically aggregated into a route in classical VRPs,
setting. Finally, applications involving pickups and deliveries with future applications will increasingly differentiate, synchronize, and
split loads have been considered in Nowak, Ö. Ergun, and White optimize multiple flows associated with products, customers, and
(2008), Sahin, Çavuşlar, Öncan, Sahin, and Aksu (2013) and Haddad vehicles. The efficient coordination of such systems is a challenging
et al. (2018). This setting is unexpectedly challenging: it is possible task, and the associated rise in complexity rests on a fine equi-
to create a family of benchmark instances for which any optimal librium: while optimization models and their data requirements
solution requires a number of split pickups and deliveries that is should be as sophisticated as required, they should also remain
an exponential function of the instance size (Haddad et al., 2018). as simple as possible.
Time constraints. A wide range of VRP variants arising from the With respect to methodology, the development of heuristics
addition of time restrictions were surveyed in Vidal, Crainic, Gen- and mathematical programming algorithms that are simple and
dreau, and Prins (2015). Time constraints can arise as customer- or efficient yet general enough to cope with a wide gamut of VRPs
self-imposed time windows for deliveries (Agatz, Campbell, Fleis- remains a crucial topic. Significant progress has been achieved
chmann, & Savelsbergh, 2011; Bruck, Cordeau, & Iori, 2018; Jabali by disciplined research built on problem-structure analysis and
et al., 2015; Solomon, 1987). In addition, release and due dates decision-set decompositions (see, e.g., Pessoa, Sadykov, Uchoa, &
for commodities can be imposed at the depot (Cattaruzza, Absi, Vanderbeck, 2019; Toffolo, Vidal, & Wauters, 2019; Vidal, 2017; Vi-
& Feillet, 2016; Shelbourne, Battarra, & Potts, 2017). Both of these dal et al., 2014). There is also a need to scale up VRP research.
settings can be viewed as time-window constraints on pickup or Current algorithms are usually evaluated on benchmark instances
delivery locations. Other applications impose response-time lim- with a few hundred delivery points. This size could be strategically
its between a request and its fulfillment by a vehicle. This is crit- increased to thousands of visits to reflect emerging applications
ical for customer satisfaction in mobility-on-demand systems, or (Arnold, Gendreau, & Sörensen, 2019; Uchoa et al., 2017). Multi-
for the delivery of perishable products (Pillac, Gendreau, Guéret, ple planning periods and scenarios should also be considered when
& Medaglia, 2013). Finally, in dial-a-ride transportation services relevant, e.g., for districting or location-routing.
where passengers with distinct pickup and delivery locations share Finally, it is important to focus our energy on problem vari-
the same vehicle, ride-time constraints are typically imposed to ants that are truly of methodological and practical interest. In-
limit detours for each customer (Cordeau & Laporte, 2007; Paque- deed, solving a new VRP variant made up of an arbitrary com-
tte et al., 2012). These constraints, however, make feasibility checks bination of attributes is certainly a technical achievement, but it
more complex. Research is ongoing into efficient solution evalua- does not necessarily constitute a significant methodological ad-
tion procedures for these problems, to speed up heuristic search vance. Reproducibility and benchmarking are other important con-
using preprocessing, incremental evaluations, and concatenations cerns. Methodological issues such as over-tuning, as well as dif-
(Gschwind & Drexl, 2019; Tang, Kong, Lau, & Ip, 2010; Vidal et al., ferences in coding protocols and in hardware have been raised by
2015). several researchers, but are not yet fully resolved. The fact that
Skills. Finally, maintenance or home care services may re- some flagship journals now require that codes be submitted as a
quire specific skills. These requirements must be taken into ac- condition for paper acceptance should, in all likelihood, foster the
count when assigning technicians and vehicles to tasks (Cappanera, enforcement of stricter experimental standards.
Gouveia, & Scutellá, 2013; Paraskevopoulos, Laporte, Repoussis, &
Tarantilis, 2017; Xie, Potts, & Bektaş, 2017). In complex situations, Acknowledgments
several vehicles and workers with different skills and equipment
(Eveborn et al., 2009; Parragh & Doerner, 2018) may be jointly This research was partly funded by the Canadian Natural and
requested for a single task, leading to VRPs with synchronization Engineering Research Council [grant number 2015–06189] as well
 T. Vidal, G. Laporte and P. Matl / European Journal of Operational Research 286 (2020) 401–416 411

as the National Council for Scientific and Technological Develop- Battarra, M., Erdoǧan, G., Laporte, G., & Vigo, D. (2010). The traveling salesman prob-
ment [grant number 308498/2015-1], CAPES and FAPERJ [grant lem with pickups, deliveries, and handling costs. Transportation Science, 44(3),
383–399.
number E-26/203.310/2016] in Brazil. This support is gratefully ac- Battarra, M., Monaci, M., & Vigo, D. (2009). An adaptive guidance approach for the
knowledged. Thanks are also due to the referees for their valuable heuristic solution of a minimum multiple trip vehicle routing problem. Comput-
comments. ers & Operations Research, 36(11), 3041–3050.
Batz, G. V., Geisberger, R., Sanders, P., & Vetter, C. (2013). Minimum time-dependent
travel times with contraction hierarchies. Journal of Experimental Algorithmics,
References 18(1), 1–43.
Bauer, J., Bektaş, T., & Crainic, T. G. (2010). Minimizing greenhouse gas emissions in
Absi, N., Archetti, C., Dauzére-Pérés, S., & Feillet, D. (2015). A two-phase iterative intermodal freight transport: An application to rail service design. Journal of the
heuristic approach for the production routing problem. Transportation Science, Operational Research Society, 61(3), 530–542.
49(4), 784–795. Beardwood, J., Halton, J. H., & Hammersley, J. M. (1959). The shortest path through
Adulyasak, Y., Cordeau, J.-F., & Jans, R. (2014). Formulations and branch-and-cut al- many points. Mathematical Proceedings of the Cambridge Philosophical Society,
gorithms for multivehicle production and inventory routing problems. INFORMS 55(9), 299–327.
Journal on Computing, 26(1), 103–120. Beasley, J. E. (1984). Fixed routes. Journal of the Operational Research Society, 35(1),
Adulyasak, Y., Cordeau, J.-F., & Jans, R. (2015). The production routing problem: A re- 49–55.
view of formulations and solution algorithms. Computers & Operations Research, Beirigo, B., Schulte, F., & Negenborn, R. R. (2019). A business class for autonomous
55, 141–152. mobility-on-demand: Modeling service quality in dynamic ridesharing systems.
Agatz, N., Bouman, P., & Schmidt, M. (2018). Optimization approaches for the trav- In IEEE intelligent transportation systems magazine.
eling salesman problem with drone. Transportation Science, 52(4), 965–981. Bektaş, T., & Laporte, G. (2011). The pollution-routing problem. Transportation Re-
Agatz, N., Campbell, A. M., Fleischmann, M., & Savelsbergh, M. W. P. (2011). search Part B: Methodological, 45(8), 1232–1250.
Time slot management in attended home delivery. Transportation Science, 45(3), Bell, W. J., Dalberto, L. M., Fisher, M. L., Greenfield, A. J., Jaikumar, R., Kedia, P.,
435–449. Mack, R. G., & Prutzman, P. J. (1983). Improving the distribution of industrial
Ak, A., & Erera, A. L. (2007). A paired-vehicle recourse strategy for the vehicle-rout- gases with an on-line computerized routing and scheduling optimizer. Inter-
ing problem with stochastic demands. Transportation Science, 41(2), 222–237. faces, 13(6), 4–23.
Akgün, V., Erkut, E., & Batta, R. (20 0 0). On finding dissimilar paths. European Journal Ben-Tal, A., Ghaoui, L. E., & Nemirovski, A. (2009). Robust optimization. Princeton, NJ:
of Operational Research, 121(2), 232–246. Princeton University Press.
Aksen, D., Kaya, O., Salman, F. S., & Akça, Y. (2012). Selective and periodic inventory Ben Ticha, H., Absi, N., Feillet, D., & Quilliot, A. (2017). Empirical analysis for the
routing problem for waste vegetable oil collection. Optimization Letters, 6(6), VRPTW with a multigraph representation for the road network. Computers &
1063–1080. Operations Research, 88(1), 103–116.
Andersson, H., Christiansen, M., & Desaulniers, G. (2016). A new decomposition al- Ben Ticha, H., Absi, N., Feillet, D., & Quilliot, A. (2018). Vehicle routing prob-
gorithm for a liquefied natural gas inventory routing problem. International Jour- lems with road-network information: State of the art. Networks, 72(3), 393–
nal of Production Research, 54(2), 564–578. 406.
Andersson, H., Hoff, A., Christiansen, M., Hasle, G., & Løkketangen, A. (2010). Indus- Ben Ticha, H., Absi, N., Feillet, D., & Quilliot, A. (2019). Multigraph modeling and
trial aspects and literature survey: Combined inventory management and rout- adaptive large neighborhood search for the vehicle routing problem with time
ing. Computers & Operations Research, 37(9), 1515–1536. windows. Computers & Operations Research, 104(1), 113–126.
van Anholt, R. G., Coelho, L. C., Laporte, G., & Vis, I. F. A. (2016). An inventory-routing Benavent, E., & Soler, D. (1999). The directed rural postman problem with turn
problem with pickups and deliveries arising in the replenishment of automated penalties. Transportation Science, 33(4), 408–418.
teller machines. Transportation Science, 50(3), 1077–1091. Benoist, T., Gardi, F., Jeanjean, A., & Estellon, B. (2011). Randomized local search for
Applegate, D. L., Bixby, R. E., Chvátal, V., & Cook, W. J. (2011). The traveling salesman real-life inventory routing. Transportation Science, 45(3), 381–398.
problem: A computational study. Princeton: Princeton University Press. Bertoli, F., Kilby, P., & Urli, T. (2020). A column-generation-based approach to fleet
Archetti, C., Bianchessi, N., & Speranza, M. G. (2014a). Branch-and-cut algorithms design problems mixing owned and hired vehicles. International Transactions in
for the split delivery vehicle routing problem. European Journal of Operational Operational Research, 27(2), 899–923.
Research, 238(3), 685–698. Bertsimas, D., Brown, D. B., & Caramanis, C. (2011). Theory and applications of ro-
Archetti, C., Boland, N., & Speranza, M. G. (2017a). A matheuristic for the multivehi- bust optimization. SIAM Review, 53(3), 464–501.
cle inventory routing problem. INFORMS Journal on Computing, 29(3), 377–387. Bertsimas, D. J. (1992). A vehicle routing problem with stochastic demand. Opera-
Archetti, C., Desaulniers, G., & Speranza, M. G. (2017b). Minimizing the logistic ratio tions Research, 40(3), 574–585.
in the inventory routing problem. EURO Journal on Transportation and Logistics, Bertsimas, D. J., Jaillet, P., & Odoni, A. (1990). A priori optimization. Operations Re-
6(4), 289–306. search, 38(6), 1019–1033.
Archetti, C., Savelsbergh, M. W. P., & Speranza, M. G. (2016). The vehicle rout- Blakeley, F., Bozkaya, B., Cao, B., Hall, W., & Knolmajer, J. (2003). Optimizing peri-
ing problem with occasional drivers. European Journal of Operational Research, odic maintenance operations for Schindler elevator corporation. Interfaces, 33(1),
254(2), 472–480. 67–79.
Archetti, C., & Speranza, M. G. (2012). Vehicle routing problems with split deliveries. Boudoin, D., Morel, C., & Gardat, M. (2014). Supply chains and urban logistics.
International Transactions in Operational Research, 19(1–2), 3–22. In J. Gonzalez-Feliu, F. Semet, & J. L. Routhier (Eds.), Sustainable urban logis-
Archetti, C., Speranza, M. G., & Vigo, D. (2014b). Vehicle routing problems with prof- tics: Concepts, methods and information systems (pp. 1–20). Berlin Heidelberg:
its. In P. Toth, & D. Vigo (Eds.), Vehicle routing: Problems, methods, and applica- Springer.
tions (pp. 273–297). Philadelphia: Society for Industrial and Applied Mathemat- Bozkaya, B., Erkut, E., & Laporte, G. (2003). A tabu search heuristic and adaptive
ics. Chap. 10 memory procedure for political districting. European Journal of Operational Re-
Arnold, F., Gendreau, M., & Sörensen, K. (2019). Efficiently solving very large scale search, 144(1), 12–26.
routing problems. Computers & Operations Research, 107(1), 32–42. Bozkaya, B., Salman, F. S., & Telciler, K. (2017). An adaptive and diversified vehi-
Arslan, A. M., Agatz, N., Kroon, L., & Zuidwijk, R. (2019). Crowdsourced delivery – cle routing approach to reducing the security risk of cash-in-transit operations.
A dynamic pickup and delivery problem with ad hoc drivers. Transportation Sci- Networks, 69(3), 256–269.
ence, 53(1), 222–235. Braekers, K., & Kovacs, A. A. (2016). A multi-period dial-a-ride problem with driver
Assad, A. A., & Golden, B. L. (1995). Arc routing methods and applications. In consistency. Transportation Research Part B: Methodological, 94, 355–377.
M. Ball, T. L. Magnanti, C. L. Monma, & G. L. Nemhauser (Eds.), Network rout- Breunig, U., Schmid, V., Hartl, R. H., & Vidal, T. (2016). A large neighbourhood based
ing (pp. 375–483). Amsterdam: Elsevier. Vol. 8 heuristic for the two-echelon vehicle routing problem. Computers & Operations
Balcik, B., Iravani, S. M. R., & Smilowitz, K. (2011). A review of equity in nonprofit Research, 76, 208–225.
and public sector: A vehicle routing perspective. In J. J. Cochran, L. A. Cox, P. Ke- Bronfman, A., Marianov, V., Paredes-Belmar, G., & Lüer-Villagra, A. (2015). The max-
skinocak, J. P. Kharoufeh, & J. C. Smith (Eds.), Wiley encyclopedia of operations imin HAZMAT routing problem. European Journal of Operational Research, 241(1),
research and management science. Wiley. 15–27.
Baldacci, R., Lim, A., Traversi, E., & Calvo, R. W. (2018). Optimal solution of ve- Bruck, B. P., Cordeau, J.-F., & Iori, M. (2018). A practical time slot management and
hicle routing problems with fractional objective function. Tech. rep., ArXiv routing problem for attended home services. Omega, 81, 208–219.
1804.03316. Bruno, G., Cavola, M., Diglio, A., Laporte, G., & Piccolo, C. (2019). Reorganizing postal,
Baldacci, R., Mingozzi, A., & Calvo, R. W. (2011). An exact method for the capacitated collection operations in urban areas as a result of declining mail volumes – A
location-routing problem. Operations Research, 59(5), 1284–1296. case study in Bologna. Tech. rep. CIRRELT-2019-19.
Baller, A. C., Dabia, S., Dullaert, W. E. H., & Vigo, D. (2019). The dynamic-demand Bulhões, T., Há, M. H., Martinelli, R., & Vidal, T. (2018a). The vehicle routing problem
joint replenishment problem with approximated transportation costs. European with service level constraints. European Journal of Operational Research, 265(2),
Journal of Operational Research, 276(3), 1013–1033. 544–558.
Bast, H., Delling, D., Goldberg, A., Müller-Hannemann, M., Pajor, T., Sanders, P., Wag- Bulhões, T., Subramanian, A., Erdogan, G., & Laporte, G. (2018b). The static bike relo-
ner, D., & Werneck, R. F. (2016). Route planning in transportation networks. In cation problem with multiple vehicles and visits. European Journal of Operational
L. Kliemann, & P. Sanders (Eds.), Algorithm engineering: Selected results and sur- Research, 264(2), 508–523.
veys (pp. 19–80). Berlin Heidelberg: Springer. Butsch, A., Kalcsics, J., & Laporte, G. (2014). Districting for arc routing. INFORMS Jour-
Battarra, M., Cordeau, J.-F., & Iori, M. (2014). Pickup-and-delivery problems for goods nal on Computing, 26(4), 809–824.
transportation. In P. Toth, & D. Vigo (Eds.), Vehicle routing: Problems, methods, Campbell, A. M., Vandenbussche, D., & Hermann, W. (2008). Routing for relief ef-
and applications (pp. 161–191). Philadelphia: SIAM. forts. Transportation Science, 42(2), 127–145.
412 T. Vidal, G. Laporte and P. Matl / European Journal of Operational Research 286 (2020) 401–416

Campbell, J. F. (1990). Locating transportation terminals to serve an expanding de- Crama, Y., Rezaei, M., Savelsbergh, M. W. P., & Woensel, T. V. (2018). Stochastic in-
mand. Transportation Research Part B: Methodological, 24(3), 173–192. ventory routing for perishable products. Transportation Science, 52(3), 526–546.
Campbell, J. F. (1995). Using small trucks to circumvent large truck restrictions: Cruijssen, F., Dullaert, W., & Fleuren, H. (2007). Horizontal cooperation in transport
Impacts on truck emissions and performance measures. Transportation Research and logistics: A literature review. Transportation Journal, 46(3), 22–39.
Part A: Policy and Practice, 29(6), 445–458. Cuda, R., Guastaroba, G., & Speranza, M. G. (2015). A survey on two-echelon routing
Campbell, J. F. (2013). A continuous approximation model for time definite many– problems. Computers & Operations Research, 55, 185–199.
to-many transportation. Transportation Research Part B: Methodological, 54, Dabia, S., Lai, D., & Vigo, D. (2019). An exact algorithm for a rich vehicle routing
100–112. problem with private fleet and common carrier. Transportation Science, 53(4),
Cappanera, P., Gouveia, L., & Scutellá, M. G. (2013). Models and valid inequalities 986–10 0 0.
to asymmetric skill-based routing problems. EURO Journal on Transportation and Daganzo, C. F. (1984). The distance traveled to visit n points with a maximum of c
Logistics, 2(1–2), 29–55. stops per vehicle: An analytic model and an application. Transportation Science,
Caramia, M., & Guerriero, F. (2009). A heuristic approach to long-haul freight trans- 18(4), 331–350.
portation with multiple objective functions. Omega, 37, 600–614. Darvish, M., & Coelho, L. C. (2018). Sequential versus integrated optimization: Pro-
Carlsson, J. G. (2012). Dividing a territory among several vehicles. INFORMS Journal duction, location, inventory control, and distribution. European Journal of Oper-
on Computing, 24(4), 565–577. ational Research, 268(1), 203–214.
Carlsson, J. G., & Delage, E. (2013). Robust partitioning for stochastic multivehicle Davis, B. A., & Figliozzi, M. A. (2013). A methodology to evaluate the competitiveness
routing. Operations Research, 61(3), 727–744. of electric delivery trucks. Transportation Research Part E: Logistics and Trans-
Carlsson, J. G., & Jia, F. (2013). Euclidean hub-and-spoke networks. Operations Re- portation Review, 49(1), 8–23.
search, 61(6), 1360–1382. Day, J. M., Wright, P. D., Schoenherr, T., Venkataramanan, M., & Gaudette, K. (2009).
Carlsson, J. G., & Jia, F. (2015). Continuous facility location with backbone network Improving routing and scheduling decisions at a distributor of industrial gasses.
costs. Transportation Science, 49(3), 433–451. Omega, 37(1), 227–237.
Cattaruzza, D., Absi, N., & Feillet, D. (2016). The multi-trip vehicle routing problem Dell’Amico, M., Falavigna, S., & Iori, M. (2015). Optimization of a real-world auto–
with time windows and release dates. Transportation Science, 50(2), 676–693. carrier transportation problem. Transportation Science, 49(2), 402–419.
Çavdar, B., & Sokol, J. (2015). A distribution-free TSP tour length estimation Demir, E., Bektaş, T., & Laporte, G. (2011). A comparative analysis of several vehicle
model for random graphs. European Journal of Operational Research, 243(2), emission models for road freight transportation. Transportation Research Part D:
588–598. Transport and Environment, 16(5), 347–357.
Çelik, M., & Süral, H. (2016). Order picking in a parallel-aisle warehouse with turn Demir, E., Bektaş, T., & Laporte, G. (2014a). The bi-objective pollution-routing prob-
penalties. International Journal of Production Research, 54(14), 1–16. lem. European Journal of Operational Research, 232(3), 464–478.
Ceschia, S., Gaspero, L., & Schaerf, A. (2011). Tabu search techniques for the het- Demir, E., Bektaş, T., & Laporte, G. (2014b). A review of recent research on green
erogeneous vehicle routing problem with time windows and carrier-dependent road freight transportation. European Journal of Operational Research, 237(3),
costs. Journal of Scheduling, 14(6), 601–615. 775–793.
Chandra, P. (1993). A dynamic distribution model with warehouse and customer Demir, E., Huang, Y., Scholts, S., & Woensel, T. V. (2015). A selected review on the
replenishment requirements. Journal of the Operational Research Society, 44(7), negative externalities of the freight transportation: Modeling and pricing. Trans-
681–692. portation Research Part E: Logistics and Transportation Review, 77, 95–114.
Chandra, P., & Fisher, M. L. (1994). Coordination of production and distribution plan- Derigs, U., Gottlieb, J., Kalkoff, J., Piesche, M., Rothlauf, F., & Vogel, U. (2010). Vehicle
ning. European Journal of Operational Research, 72, 503–517. routing with compartments: Applications, modelling and heuristics. OR Spec-
Chemla, D., Meunier, F., & Calvo, R. W. (2013). Bike sharing systems: Solving the trum, 33(4), 885–914.
static rebalancing problem. Discrete Optimization, 10(2), 120–146. Desaulniers, G., Rakke, J. G., & Coelho, L. C. (2016). A branch-price-and-cut algorithm
Chitsaz, M., Cordeau, J.-F., & Jans, R. (2019). A unified decomposition matheuristic for the inventory-routing problem. Transportation Science, 50(3), 1060–1076.
for assembly, production, and inventory routing. INFORMS Journal on Computing, Doerner, K. F., & Salazar-González, J. J. (2014). Pickup-and-delivery problems for peo-
31(1), 134–152. ple transportation. In P. Toth, & D. Vigo (Eds.), Vehicle routing: Problems, methods,
Christiansen, M., Fagerholt, K., Nygreen, B., & Ronen, D. (2013). Ship routing and and applications (pp. 193–212). Philadelphia: SIAM.
scheduling in the new millennium. European Journal of Operational Research, Dorling, K., Heinrichs, J., Messier, G. G., & Magierowski, S. (2017). Vehicle routing
228(3), 467–483. problems for drone delivery. IEEE Transactions on Systems, Man, and Cybernetics:
Christofides, N. (1971). Fixed routes and areas for delivery operations. International Systems, 47(1), 70–85.
Journal of Physical Distribution, 1(2), 87–92. Drexl, A., & Haase, K. (1999). Fast approximation methods for sales force deploy-
Cleophas, C., Cottrill, C., Ehmke, J. F., & Tierney, K. (2019). Collaborative urban trans- ment. Management Science, 45(10), 1307–1323.
portation: Recent advances in theory and practice. European Journal of Opera- Drexl, M. (2012). Synchronization in vehicle routing – A survey of VRPs with multi-
tional Research, 273(3), 801–816. ple synchronization constraints. Transportation Science, 46(3), 297–316.
Clossey, J., Laporte, G., & Soriano, P. (2001). Solving arc routing problems with turn Dror, M., Laporte, G., & Trudeau, P. (1989). Vehicle routing with stochastic de-
penalties. Journal of the Operational Research Society, 52(4), 433–439. mands: Properties and solution frameworks. Transportation Science, 23(3),
Coelho, L. C., Cordeau, J.-F., & Laporte, G. (2012). Consistency in multi-vehicle inven- 166–176.
tory-routing. Transportation Research Part C: Emerging Technologies, 24, 270–287. Duckham, M., & Kulik, L. (2003). Simplest paths: Automated route selection for nav-
Coelho, L. C., Cordeau, J.-F., & Laporte, G. (2014). Thirty years of inventory routing. igation. In W. Kuhn, M. F. Worboys, & S. Timpf (Eds.), Proceedings of the interna-
Transportation Science, 48(1), 1–19. tional conference on spatial information theory (pp. 169–185). Kartause Ittingen,
Coelho, L. C., & Laporte, G. (2013). The exact solution of several classes of invento- Switzerland: Springer.
ry-routing problems. Computers & Operations Research, 40(2), 558–565. Errico, F., Desaulniers, G., Gendreau, M., Rei, W., & Rousseau, L. M. (2018). The vehi-
Coelho, L. C., & Laporte, G. (2014). Optimal joint replenishment, delivery and in- cle routing problem with hard time windows and stochastic service times. EURO
ventory management policies for perishable products. Computers & Operations Journal on Transportation and Logistics, 7(3), 223–251.
Research, 47, 42–52. Eveborn, P., Flisberg, P., & Rönnqvist, M. (2006). LAPS CARE – an operational system
Constantino, M., Gouveia, L., Mourão, M. C., & Nunes, A. C. (2015). The mixed ca- for staff planning of home care. European Journal of Operational Research, 171(1),
pacitated arc routing problem with non-overlapping routes. European Journal of 962–976.
Operational Research, 244(2), 445–456. Eveborn, P., Rönnqvist, M., Einarsdóttir, H., Eklund, M., Lidén, K., & Alm-
Constantino, M., Mourão, M. C., & Pinto, L. S. (2017). Dissimilar arc routing problems. roth, M. (2009). Operations research improves quality and efficiency in home
Networks, 70(3), 233–245. care. Interfaces, 39(1), 18–34.
Contardo, C., Cordeau, J.-F., & Gendron, B. (2014). An exact algorithm based on Fagerholt, K., Laporte, G., & Norstad, I. (2009). Reducing fuel emissions by optimiz-
cut-and-column generation for the capacitated location-routing problem. IN- ing speed on shipping routes. Journal of the Operational Research Society, 61(3),
FORMS Journal on Computing, 26(1), 88–102. 523–529.
Cookson, G., & Pishue, B. (2018). INRIX global traffic scorecard. Tech. rep., INRIX Fagnant, D. J., & Kockelman, K. (2015). Preparing a nation for autonomous vehi-
Research. cles: Opportunities, barriers and policy recommendations. Transportation Re-
Corberán, A., Golden, B. L., Lum, O., Plana, I., & Sanchis, J. M. (2017). Aesthetic con- search Part A: Policy and Practice, 77, 167–181.
siderations for the min-max k-windy rural postman problem. Networks, 70(3), Feillet, D., Garaix, T., Lehuede, F., Péton, O., & Quadri, D. (2014). A new consistent
216–232. vehicle routing problem for the transportation of people with disabilities. Net-
Corberán, A., Martí, R., Martínez, E., & Soler, D. (2002). The rural postman problem works, 63(3), 211–224.
on mixed graphs with turn penalties. Journal of the Operational Research Society, Felipe, A., Ortuño, M. T., Righini, G., & Tirado, G. (2014). A heuristic approach for the
29(7), 887–903. green vehicle routing problem with multiple technologies and partial recharges.
Cordeau, J.-F., Iori, M., Laporte, G., & González, J. J. S. (2009). A branch-and-cut algo- Transportation Research Part E: Logistics and Transportation Review, 71, 111–128.
rithm for the pickup and delivery traveling salesman problem with LIFO load- Figliozzi, M., & Tipagornwong, C. (2017). Impact of last mile parking availability on
ing. Networks, 55(1), 46–59. commercial vehicle costs and operations. Supply Chain Forum: An International
Cordeau, J.-F., & Laporte, G. (2007). The dial-a-ride problem: Models and algorithms. Journal, 18(2), 60–68.
Annals of Operations Research, 153(1), 29–46. Franceschetti, A., Honhon, D., Laporte, G., Woensel, T. V., & Fransoo, J. C. (2017a).
Côté, J.-F., & Potvin, J. Y. (2009). A tabu search heuristic for the vehicle routing prob- Strategic fleet planning for city logistics. Transportation Research Part B: Method-
lem with private fleet and common carrier. European Journal of Operational Re- ological, 95, 19–40.
search, 198(2), 464–469. Franceschetti, A., Honhon, D., Woensel, T. V., Bektaş, T., & Laporte, G. (2013).
Crainic, T. G. (2002). Long-haul freight transportation. In R. W. Hall (Ed.), Handbook The time-dependent pollution-routing problem. Transportation Research Part B:
of transportation science (pp. 451–516). Boston: Springer. Methodological, 56, 265–293.
 T. Vidal, G. Laporte and P. Matl / European Journal of Operational Research 286 (2020) 401–416 413

Franceschetti, A., Jabali, O., & Laporte, G. (2017b). Continuous approximation models Halvorsen-Weare, E. E., & Fagerholt, K. (2013). Routing and scheduling in a liquefied
in freight distribution management. TOP, 25(3), 413–433. natural gas shipping problem with inventory and berth constraints. Annals of
Froger, A., Mendoza, J. E., Jabali, O., & Laporte, G. (2017). A matheuristic for the Operations Research, 203(1), 167–186.
electric vehicle routing problem with capacitated charging stations. Tech. rep., Hamdi-Dhaoui, K., Labadie, N., & Yalaoui, A. (2014). The bi-objective two-dimen-
CIRRELT-2017-31. sional loading vehicle routing problem with partial conflicts. International Jour-
Fukasawa, R., He, Q., Santos, F., & Song, Y. (2018). A joint vehicle routing and speed nal of Production Research, 52(19), 5565–5582.
optimization problem. INFORMS Journal on Computing, 30(4), 694–709. Hartl, R. F., & Romauch, M. (2016). Notes on the single route lateral transhipment
Gahm, C., Brabänder, C., & Tuma, A. (2017). Vehicle routing with private fleet, multi- problem. Journal of Global Optimization, 65(1), 57–82.
ple common carriers offering volume discounts, and rental options. Transporta- Haughton, M. A. (2002). Measuring and managing the learning requirements of
tion Research Part E: Logistics and Transportation Review, 97, 192–216. route reoptimization on delivery vehicle drivers. Journal of Business Logistics,
Gansterer, M., & Hartl, R. F. (2018). Collaborative vehicle routing: A survey. European 23(2), 45–66.
Journal of Operational Research, 268(1), 1–12. Haughton, M. A. (2007). Assigning delivery routes to drivers under variable cus-
Garaix, T., Artigues, C., Feillet, D., & Josselin, D. (2010). Vehicle routing problems tomer demands. Transportation Research Part E: Logistics and Transportation Re-
with alternative paths: An application to on-demand transportation. European view, 43(1), 157–172.
Journal of Operational Research, 204, 62–75. Hiermann, G., Hartl, R. F., Puchinger, J., Schiffer, M., & Vidal, T. (2019a). Sustainable
Garaix, T., Artigues, C., Feillet, D., & Josselin, D. (2011). Optimization of occupancy city logistics via access restrictions? An impact assessment of city center poli-
rate in dial-a-ride problems via linear fractional column generation. Computers cies. Tech. rep. University of Vienna.
& Operations Research, 38(10), 1435–1442. Hiermann, G., Hartl, R. F., Puchinger, J., & Vidal, T. (2019b). Routing a mix of con-
Gaudioso, M., & Paletta, G. (1992). A heuristic for the periodic vehicle routing prob- ventional, plug-in hybrid, and electric vehicles. European Journal of Operational
lem. Transportation Science, 26(2), 86–92. Research, 272(1), 235–248.
Gendreau, M., Ghiani, G., & Guerriero, E. (2015). Time-dependent routing problems: Hiermann, G., Puchinger, J., Ropke, S., & Hartl, R. F. (2016). The electric fleet size
A review. Computers & Operations Research, 64, 189–197. and mix vehicle routing problem with time windows and recharging stations.
Gendreau, M., Iori, M., Laporte, G., & Martello, S. (2006). A tabu search algorithm European Journal of Operational Research, 252(3), 995–1018.
for a routing and container loading problem. Transportation Science, 40(3), Hoff, A., Andersson, H., Christiansen, M., Hasle, G., & Løkketangen, A. (2010). Indus-
342–350. trial aspects and literature survey: Fleet composition and routing. Computers &
Gendreau, M., Jabali, O., & Rei, W. (2014). Stochastic vehicle routing problems. In Operations Research, 37(12), 2041–2061.
P. Toth, & D. Vigo (Eds.), Vehicle routing: Problems, methods, and applications. Holland, C., Levis, J., Nuggehalli, R., Santilli, B., & Winters, J. (2017). UPS optimizes
Philadelphia: Society for Industrial and Applied Mathematics. delivery routes. Interfaces, 47(1), 8–23.
Gendreau, M., Jabali, O., & Rei, W. (2016). Future research directions in stochastic Hollis, B. L., & Green, P. J. (2012). Real-life vehicle routing with time windows for vi-
vehicle routing. Transportation Science, 50(4), 1163–1173. sual attractiveness and operational robustness. Asia-Pacific Journal of Operational
Ghiami, Y., Demir, E., Woensel, T. V., Christiansen, M., & Laporte, G. (2019). A dete- Research, 29(4), 1–29.
riorating inventory routing problem for an inland liquefied natural gas distribu- Hoogeboom, M., & Dullaert, W. (2019). Vehicle routing with arrival time diversifica-
tion network. Transportation Research Part B: Methodological, 126, 45–67. tion. European Journal of Operational Research, 275(1), 93–107.
Goeke, D., Gschwind, T., & Schneider, M. (2019a). Upper and lower bounds for the Huang, M., Smilowitz, K., & Balcik, B. (2012). Models for relief routing: Equity, ef-
vehicle-routing problem with private fleet and common carrier. Discrete Applied ficiency and efficacy. Transportation Research Part E: Logistics and Transportation
Mathematics, 264, 43–61. Review, 48(1), 2–18.
Goeke, D., Roberti, R., & Schneider, M. (2019b). Exact and heuristic solution of the Huang, Y., Zhao, L., Woensel, T. V., & Gross, J. P. (2017). Time-dependent vehicle rout-
consistent vehicle-routing problem. Transportation Science, 53(4), 1023–1042. ing problem with path flexibility. Transportation Research Part B: Methodological,
Goeke, D., & Schneider, M. (2015). Routing a mixed fleet of electric and conventional 95, 169–195.
vehicles. European Journal of Operational Research, 245(1), 81–99. Hvattum, L. M., Norstad, I., Fagerholt, K., & Laporte, G. (2013). Analysis of an ex-
Goel, A. (2018). Legal aspects in road transport optimization in europe. Transporta- act algorithm for the vessel speed optimization problem. Networks, 62(2), 132–
tion Research Part E: Logistics and Transportation Review, 114, 144–162. 135.
Goel, A., & Meisel, F. (2013). Workforce routing and scheduling for electricity net- Hwang, J.-Y., Myung, S.-T., & Sun, Y. K. (2017). Sodium-ion batteries: Present and
work maintenance with downtime minimization. European Journal of Opera- future. Chemical Society Reviews, 46(12), 3485–3856.
tional Research, 231(1), 210–228. Ichoua, S., Gendreau, M., & Potvin, J. Y. (2003). Vehicle dispatching with time-de-
Goel, A., & Vidal, T. (2014). Hours of service regulations in road freight transport: pendent travel times. European Journal of Operational Research, 144(2), 379–396.
An optimization-based international assessment. Transportation Science, 48(3), Iori, M., González, J. J. S., & Vigo, D. (2007). An exact approach for the vehicle rout-
391–412. ing problem with two-dimensional loading constraints. Transportation Science,
Goel, A., Vidal, T., & Kok, A. L. (2019). To team up or not – Single versus team driving 41(2), 253–264.
in European road freight transport. Rio de Janeiro, PUC–Rio. Tech. rep. Irnich, S. (2008). Solution of real-world postman problems. European Journal of Op-
Golden, B., Raghavan, S., & Wasil, E. (2008). The vehicle routing problem: Latest ad- erational Research, 190(1), 52–67.
vances and new challenges. New York: Springer. Jabali, O., Gendreau, M., & Laporte, G. (2012a). A continuous approximation model
Golden, B. L., Kovacs, A. A., & Wasil, E. A. (2014). Vehicle routing applications in for the fleet composition problem. Transportation Research Part B: Methodologi-
disaster relief. In P. Toth, & D. Vigo (Eds.), Vehicle routing: Problems, methods, cal, 46(10), 1591–1606.
and applications (pp. 409–436). Philadelphia: SIAM. Jabali, O., Leus, R., Woensel, T. V., & de Kok, T. (2015). Self-imposed time windows
Goodson, J. C. (2014). Election day routing of rapid response attorneys. INFOR: Infor- in vehicle routing problems. OR Spectrum, 37(2), 331–352.
mation Systems and Operational Research, 52(1), 1–9. Jabali, O., Woensel, T. V., & de Kok, A. G. (2012b). Analysis of travel times and CO2
Goodson, J. C., Ohlmann, J. W., & Thomas, B. W. (2013). Rollout policies for dynamic emissions in time-dependent vehicle routing. Production and Operations Man-
solutions to the multivehicle routing problem with stochastic demand and du- agement, 21(6), 1060–1074.
ration limits. Operations Research, 61(1), 138–154. Jacobsen, S. K., & Madsen, O. B. G. (1980). A comparative study of heuristics for
Gounaris, C. E., Wiesemann, W., & Floudas, C. A. (2013). The robust capacitated a two-level routing-location problem. European Journal of Operational Research,
vehicle routing problem under demand uncertainty. Operations Research, 61(3), 5(6), 378–387.
677–693. Jaillet, P. (1988). A priori solution of a traveling salesman problem in which a ran-
Grabenschweiger, J., Tricoire, F., & Doerner, K. F. (2018). Finding the trade-off be- dom subset of the customers are visited. Operations Research, 36(6), 929–936.
tween emissions and disturbance in an urban context. Flexible Services and Man- Jaillet, P., Qi, J., & Sim, M. (2016). Routing optimization under uncertainty. Operations
ufacturing Journal, 30(3), 554–591. Research, 64(1), 186–200.
Grangier, P., Gendreau, M., Lehuédé, F., & Rousseau, L. M. (2016). An adaptive large Jain, A. K. (2010). Data clustering: 50 years beyond k-means. Pattern Recognition Let-
neighborhood search for the two-echelon multiple-trip vehicle routing problem ters, 31(8), 651–666.
with satellite synchronization. European Journal of Operational Research, 254(1), Janssens, J., Bergh, J. V. d., Sörensen, K., & Cattrysse, D. (2015). Multi-objective mi-
80–91. crozone-based vehicle routing for courier companies: From tactical to opera-
Groër, C., & Golden, B. L. (2009). The balanced billing cycle vehicle routing problem. tional planning. European Journal of Operational Research, 242(1), 222–231.
Networks, 54(4), 243–254. Kalcsics, J. (2015). Districting problems. In G. Laporte, S. Nickel, & F. S. d. Gama
Gschwind, T., & Drexl, M. (2019). Adaptive large neighborhood search with a con- (Eds.), Location science (pp. 595–622). Berlin: Springer.
stant-time feasibility test for the dial-a-ride problem. Transportation Science, Kara, I., Kara, B. Y., & Yetis, M. K. (2007). Energy minimizing vehicle routing prob-
53(2), 480–491. lem. In Proceedings of the international conference on combinatorial optimization
Guajardo, M., & Rönnqvist, M. (2016). A review on cost allocation methods in collab- and applications (pp. 62–71). Berlin Heidelberg: Springer.
orative transportation. International Transactions in Operational Research, 23(3), Keskin, M., & Çatay, B. (2016). Partial recharge strategies for the electric vehicle
371–392. routing problem with time windows. Transportation Research Part C: Emerging
Guastaroba, G., Speranza, M. G., & Vigo, D. (2016). Intermediate facilities in freight Technologies, 65, 111–127.
transportation planning: A survey. Transportation Science, 50(3), 763–789. Kilby, P., & Urli, T. (2016). Fleet design optimisation from historical data using con-
Haddad, M. N., Martinelli, R., Vidal, T., Martins, S., Ochi, L. S., Souza, M. J. F., straint programming and large neighbourhood search. Constraints, 21, 2–21.
& Hartl, R. F. (2018). Large neighborhood-based metaheuristic and Kim, J., Cha, M., & Sandholm, T. (2014). Socroutes: Safe routes based on tweet sen-
branch-and-price for the pickup and delivery problem with split loads. timents. In Proceedings of the twenty-third international conference on world wide
European Journal of Operational Research, 270(3), 1014–1027. web (pp. 179–182). Association for Computing Machinery.
Hall, R., & Partyka, J. (2018). Higher expectations drive transformation. OR/MS Today, Klibi, W., Lasalle, F., Martel, A., & Ichoua, S. (2010). The stochastic multiperiod loca-
45(1), 40–47. tion transportation problem. Transportation Science, 44(2), 221–237.
414 T. Vidal, G. Laporte and P. Matl / European Journal of Operational Research 286 (2020) 401–416

Koç, c., Bektaş, T., Jabali, O., & Laporte, G. (2015). A hybrid evolutionary algorithm Montreuil, B. (2011). Toward a physical internet: Meeting the global logistics sus-
for heterogeneous fleet vehicle routing problems with time windows. Computers tainability grand challenge. Logistics Research, 3(2–3), 71–87.
& Operations Research, 64, 11–27. Morillo, C., & Campos, J. M. (2014). On-street illegal parking costs in urban areas.
Koç, c., Bektaş, T., Jabali, O., & Laporte, G. (2016). Thirty years of heterogeneous ve- Procedia – Social and Behavioral Sciences, 160, 342–351.
hicle routing. European Journal of Operational Research, 249(1), 1–21. Mourad, A., Puchinger, J., & Chu, C. (2019). A survey of models and algorithms for
Kovacs, A. A., Golden, B. L., Hartl, R. F., & Parragh, S. N. (2014a). Vehicle routing optimizing shared mobility. Transportation Research Part B: Methodological, 123,
problems in which consistency considerations are important: A survey. Net- 323–346.
works, 64(3), 192–213. Mufalli, F., Batta, R., & Nagi, R. (2012). Simultaneous sensor selection and routing
Kovacs, A. A., Parragh, S. N., & Hartl, R. F. (2014b). A template-based adaptive of unmanned aerial vehicles for complex mission plans. Computers & Operations
large neighborhood search for the consistent vehicle routing problem. Networks, Research, 39(11), 2787–2799.
63(1), 60–81. Müller-Hannemann, M., & Weihe, K. (2006). On the cardinality of the Pareto set in
Krajewska, M. A., & Kopfer, H. (2009). Transportation planning in freight forwarding bicriteria shortest path problems. Annals of Operations Research, 147(1), 269–286.
companies. European Journal of Operational Research, 197, 741–751. Murray, C. C., & Chu, A. G. (2015). The flying sidekick traveling salesman problem:
Kramer, R., Subramanian, A., Vidal, T., & Cabral, L. A. F. (2015). A matheuristic ap- Optimization of drone-assisted parcel delivery. Transportation Research Part C:
proach for the pollution-routing problem. European Journal of Operational Re- Emerging Technologies, 54, 86–109.
search, 243(2), 523–539. Neves-Moreira, F., Almada-Lobo, B., Cordeau, J.-F., Guimarães, L., & Jans, R. (2019).
Kwon, O., Golden, B. L., & Wasil, E. A. (1995). Estimating the length of the optimal Solving a large multi-product production-routing problem with delivery time
TSP tour: An empirical study using regression and neural networks. Computers windows. Omega, 86, 154–172.
& Operations Research, 22(10), 1039–1046. Ngueveu, S. U., Prins, C., & Calvo, R. W. (2010a). A hybrid tabu search for the
Laporte, G., & Dejax, P. J. (1989). Dynamic location-routeing problems. Journal of the m-peripatetic vehicle routing problem. In V. Maniezzo, T. Stützle, & S. Voß
Operational Research Society, 40(5), 471–482. (Eds.), Matheuristics (pp. 253–266). Boston, MA: Springer.
Laporte, G., Louveaux, F. V., & Mercure, H. (1992). The vehicle routing problem with Ngueveu, S. U., Prins, C., & Calvo, R. W. (2010b). Lower and upper bounds for the
stochastic travel times. Transportation Science, 26(3), 161–170. m-peripatetic vehicle routing problem. 4OR, 8(4), 387–406.
Laporte, G., Meunier, F., & Calvo, R. W. (2018). Shared mobility systems: An updated Nielsen, O. A., Frederiksen, R. D., & Simonsen, N. (1998). Using expert system rules
survey. Annals of Operations Research, 271, 105–126. to establish data for intersections and turns in road networks. International
Laporte, G., Nickel, S., & Gama, F. S. d. (2015). Location science. Berlin: Springer. Transactions in Operational Research, 5(6), 569–581.
Lei, H., Laporte, G., Liu, Y., & Zhang, T. (2015). Dynamic design of sales territories. Nocerino, R., Colorni, A., Lia, F., & Lué, A. (2016). E-bikes and e-scooters for smart
Computers & Operations Research, 56, 84–92. logistics: Environmental and economic sustainability in pro-e-bike italian pilots.
Lei, H., Wang, R., & Laporte, G. (2016). Solving a multi-objective dynamic stochastic Transportation Research Procedia, 14, 2362–2371.
districting and routing problem with a co-evolutionary algorithm. Computers & Norstad, I., Fagerholt, K., & Laporte, G. (2011). Tramp ship routing and scheduling
Operations Research, 67, 12–24. with speed optimization. Transportation Research Part C: Emerging Technologies,
Li, X., Tian, P., & Leung, S. C. H. (2010). Vehicle routing problems with time win- 19(5), 853–865.
dows and stochastic travel and service times: Models and algorithm. Interna- Nourinejad, M., & Roorda, M. J. (2017). A continuous approximation model for the
tional Journal of Production Economics, 125(1), 137–145. fleet composition problem on the rectangular grid. OR Spectrum, 39(2), 373–401.
Lin, M., Chin, K.-S., Fu, C., & Tsui, K. L. (2017). An effective greedy method for the Novaes, A. G. N., de Cursi, J. E. S., & Graciolli, O. D. (20 0 0). A continuous approach
meals-on-wheels service districting problem. Computers & Industrial Engineering, to the design of physical distribution systems. Computers & Operations Research,
106, 1–19. 27(9), 877–893.
Liu, S. (2013). A hybrid population heuristic for the heterogeneous vehicle routing Nowak, M., Ö. Ergun, C. C., & White, I. (2008). Pickup and delivery with split loads.
problems. Transportation Research Part E: Logistics and Transportation Review, 54, Transportation Science, 42(1), 32–43.
67–78. Orda, A., & Rom, R. (1990). Shortest-path and minimum-delay algorithms in net-
Lopez, L., Carter, M. W., & Gendreau, M. (1998). The hot strip mill production works with time-dependent edge-length. Journal of the Association for Comput-
scheduling problem: A tabu search approach. European Journal of Operational ing Machinery, 37(3), 607–625.
Research, 106(2–3), 317–335. Orlis, C., Laganá, D., Dullaert, W., & Vigo, D. (2019). Distribution with quality of ser-
Louveaux, F. V., & Salazar-González, J. J. (2018). Exact approach for the vehicle rout- vice considerations: The capacitated routing problem with profits and service
ing problem with stochastic demands and preventive returns. Transportation Sci- level requirements. In Omega, articles in advance.
ence, 52(6), 1463–1478. Ouyang, Y., & Daganzo, C. F. (2006). Discretization and validation of the continuum
Lum, O., Cerrone, C., Golden, B. L., & Wasil, E. A. (2017). Partitioning a street net- approximation scheme for terminal system design. Transportation Science, 40(1),
work into compact, balanced, and visually appealing routes. Networks, 69(3), 89–98.
290–303. Padilla Tinoco, S. V., Creemers, S., & Boute, R. N. (2017). Collaborative shipping un-
Luo, Z., Qin, H., Che, C., & Lim, A. (2015). On service consistency in multi-period der different cost-sharing agreements. European Journal of Operational Research,
vehicle routing. European Journal of Operational Research, 243(3), 731–744. 263(3), 827–837.
Ma, H., Cheang, B., Lim, A., Zhang, L., & Zhu, Y. (2012). An investigation into the ve- Pantuso, G., Fagerholt, K., & Hvattum, L. M. (2014). A survey on maritime fleet size
hicle routing problem with time windows and link capacity constraints. Omega, and mix problems. European Journal of Operational Research, 235(2), 341–349.
40(3), 336–347. Papageorgiou, D. J., Nemhauser, G. L., Sokol, J., Cheon, M.-S., & Keha, A. B. (2014).
Maden, W., Eglese, R., & Black, D. (2009). Vehicle routing and scheduling with time– MIRPLib – A library of maritime inventory routing problem instances: Survey,
varying data: A case study. Journal of the Operational Research Society, 61(3), core model, and benchmark results. European Journal of Operational Research,
515–522. 235(2), 350–366.
Marques, G., Sadykov, R., Deschamps, J.-C., & Dupas, R. (2019). An improved branch- Paquette, J., Bellavance, F., Cordeau, J.-F., & Laporte, G. (2012). Measuring quality
cut-and-price algorithm for the two-echelon capacitated vehicle routing prob- of service in dial-a-ride operations: The case of a canadian city. Transportation,
lem. Tech. rep. University of Bordeaux. 39(3), 539–564.
Martí, R., Velarde, J. L. G., & Duarte, A. (2009). Heuristics for the bi-objective path Paraskevopoulos, D. C., Laporte, G., Repoussis, P. P., & Tarantilis, C. D. (2017). Re-
dissimilarity problem. Computers & Operations Research, 36(11), 2905–2912. source constrained routing and scheduling: Review and research prospects. Eu-
Martinelli, R., & Contardo, C. (2015). Exact and heuristic algorithms for capacitated ropean Journal of Operational Research, 263(3), 737–754.
vehicle routing problems with quadratic costs structure. INFORMS Journal on Park, J., & Kim, B. I. (2010). The school bus routing problem: A review. European
Computing, 27(4), 658–676. Journal of Operational Research, 202(2), 311–319.
Matl, P., Hartl, R. F., & Vidal, T. (2018). Workload equity in vehicle routing problems: Parragh, S. N., & Doerner, K. F. (2018). Solving routing problems with pairwise syn-
A survey and analysis. Transportation Science, 52(2), 239–260. chronization constraints. Central European Journal of Operations Research, 26(2),
Matl, P., Hartl, R. F., & Vidal, T. (2019). Workload equity in vehicle routing: The im- 443–464.
pact of alternative workload resources. Computers & Operations Research, 110, Pasha, U., Hoff, A., & Hvattum, L. M. (2016). Simple heuristics for the multi-period
116–129. fleet size and mix vehicle routing problem. INFOR: Information Systems and Op-
Mendoza, J., Medaglia, A., & Velasco, N. (2009). An evolutionary-based decision sup- erational Research, 54(2), 97–120.
port system for vehicle routing: The case of a public utility. Decision Support Paterson, C., Kiesmüller, G., Teunter, R., & Glazebrook, K. (2011). Inventory models
Systems, 46(3), 730–742. with lateral transshipments: A review. European Journal of Operational Research,
Meng, Q., Lee, D. H., & Cheu, R. L. (2005). Multiobjective vehicle routing and 210(2), 125–136.
scheduling problem with time window constraints in hazardous material trans- Pelletier, S., Jabali, O., & Laporte, G. (2016). Goods distribution with electric vehicles:
portation. Journal of Transportation Engineering, 131(9), 699–707. Review and research perspectives. Transportation Science, 50(1), 3–22.
Merchán, D., & Winkenbach, M. (2019). A data-driven extension of continuum Pelletier, S., Jabali, O., & Laporte, G. (2018). Charge scheduling for electric freight
approximation approaches to predict urban route distances. Networks, 73(4), vehicles. Transportation Research Part B: Methodological, 115, 246–269.
418–433. Penna, P. H. V., Subramanian, A., Ochi, L. S., Vidal, T., & Prins, C. (2019). A hybrid
Michallet, J., Prins, C., Amodeo, L., Yalaoui, F., & Vitry, G. (2014). Multi-start iter- heuristic for a broad class of vehicle routing problems with heterogeneous fleet.
ated local search for the periodic vehicle routing problem with time windows Annals of Operations Research, 273(1–2), 5–74.
and time spread constraints on services. Computers & Operations Research, 41(1), Perrier, N., Langevin, A., & Amaya, C. A. (2008). Vehicle routing for urban snow
196–207. plowing operations. Transportation Science, 42(1), 44–56.
Montoya, A., Guéret, C., Mendoza, J. E., & Villegas, J. G. (2017). The electric vehicle Pessoa, A., Sadykov, R., & Uchoa, E. (2018a). Enhanced branch-cut-and-price algo-
routing problem with nonlinear charging function. Transportation Research Part rithm for heterogeneous fleet vehicle routing problems. European Journal of Op-
B: Methodological, 103, 87–110. erational Research, 270(2), 530–543.
 T. Vidal, G. Laporte and P. Matl / European Journal of Operational Research 286 (2020) 401–416 415

Pessoa, A., Sadykov, R., Uchoa, E., & Vanderbeck, F. (2019). A generic exact solver for Soriano, A., Vidal, T., Gansterer, M., & Doerner, K. (2019). The vehicle routing prob-
vehicle routing and related problems. In IPCO. Michigan, USA: Ann Arbor. lem with arrival time diversification on a multigraph. Tech. rep.. Austria Univer-
Pessoa, A. A., Poss, M., Sadykov, R., & Vanderbeck, F. (2018b). Branch-and-cut-and- sity of Vienna.
price for the robust capacitated vehicle routing problem with knapsack uncer- Soysal, M., Bloemhof-Ruwaard, J. M., & Bektaş, T. (2015). The time-dependent
tainty. Tech. rep. Universidade Federal Fluminense. two-echelon capacitated vehicle routing problem with environmental consid-
Pillac, V., Gendreau, M., Guéret, C., & Medaglia, A. L. (2013). A review of dynamic erations. International Journal of Production Economics, 164, 366–378.
vehicle routing problems. European Journal of Operational Research, 225(1), 1–11. Spliet, R., & Dekker, R. (2016). The driver assignment vehicle routing problem. Net-
Poikonen, S., Wang, X., & Golden, B. L. (2017). The vehicle routing problem with works, 68(3), 212–223.
drones: Extended models and connections. Networks, 70(1), 34–43. Spliet, R., & Gabor, A. F. (2015). The time window assignment vehicle routing prob-
Pollaris, H., Braekers, K., Caris, A., Janssens, G. K., & Limbourg, S. (2015). Vehicle lem. Transportation Science, 49(4), 721–731.
routing problems with loading constraints: State-of-the-art and future direc- Stålhane, M., Rakke, J. G., Moe, C. R., Andersson, H., Christiansen, M., & Fager-
tions. OR Spectrum, 37(2), 297–330. holt, K. (2012). A construction and improvement heuristic for a liquefied nat-
Pollaris, H., Braekers, K., Caris, A., Janssens, G. K., & Limbourg, S. (2017). Iterated ural gas inventory routing problem. Computers & Industrial Engineering, 62(1),
local search for the capacitated vehicle routing problem with sequence-based 245–255.
pallet loading and axle weight constraints. Networks, 69(3), 304–316. Stenger, A., Schneider, M., & Goeke, D. (2013a). The prize-collecting vehicle routing
Poot, A., Kant, G., & Wagelmans, A. P. M. (2002). A savings based method for re- problem with single and multiple depots and non-linear cost. EURO Journal on
al-life vehicle routing problems. Journal of the Operational Research Society, 53(1), Transportation and Logistics, 2(1–2), 57–87.
57–68. Stenger, A., Vigo, D., Enz, S., & Schwind, M. (2013b). An adaptive variable neighbor-
Prescott-Gagnon, E., Desaulniers, G., Drexl, M., & Rousseau, L. M. (2010). European hood search algorithm for a vehicle routing problem arising in small package
driver rules in vehicle routing with time windows. Transportation Science, 44(4), shipping. Transportation Science, 47(1), 64–80.
455–473. Sungur, I., Ordóñez, F., & Dessouky, M. (2008). A robust optimization approach for
Prodhon, C., & Prins, C. (2014). A survey of recent research on location-routing prob- the capacitated vehicle routing problem with demand uncertainty. IIE Transac-
lems. European Journal of Operational Research, 238, 1–17. tions, 40(5), 509–523.
Qiu, Y., Qiao, J., & Pardalos, P. M. (2019). Optimal production, replenishment, deliv- Sungur, I., Ren, Y., Ordóñez, F., Dessouky, M., & Zhong, H. (2010). A model and algo-
ery, routing and inventory management policies for products with perishable rithm for the courier delivery problem with uncertainty. Transportation Science,
inventory. Omega, 82, 193–204. 44(2), 193–205.
Rasmussen, M. S., Justesen, T., Dohn, A., & Larsen, J. (2012). The home care crew Talarico, L., Sörensen, K., & Springael, J. (2015). Metaheuristics for the risk-con-
scheduling problem: Preference-based visit clustering and temporal dependen- strained cash-in-transit vehicle routing problem. European Journal of Operational
cies. European Journal of Operational Research, 219(3), 598–610. Research, 244(2), 457–470.
Reinhardt, L. B., Jepsen, M. K., & Pisinger, D. (2016). The edge set cost of the Tang, H., & Miller-Hooks, E. (2006). Interactive heuristic for practical vehicle rout-
vehicle routing problem with time windows. Transportation Science, 50(2), ing problem with solution shape constraints. Transportation Research Record,
694–707. 1964(1), 9–18.
Rincon-Garcia, N., Waterson, B. J., & Cherrett, T. J. (2018). Requirements from ve- Tang, J., Kong, Y., Lau, H., & Ip, A. W. H. (2010). A note on “efficient feasibil-
hicle routing software: Perspectives from literature, developers and the freight ity testing for dial-a-ride problems”. Operations Research Letters, 38(5), 405–
industry. Transport Reviews, 38(1), 117–138. 407.
Ríos-Mercado, R. S., & Fernández, E. (2009). A reactive GRASP for a commercial ter- Tarantilis, C. D., & Kiranoudis, C. T. (2001). Using the vehicle routing problem for
ritory design problem with multiple balancing requirements. Computers & Op- the transportation of hazardous materials. Operational Research. An International
erations Research, 36(3), 755–776. Journal, 1(1), 67–78.
Rosenfield, D. B., Engelstein, I., & Feigenbaum, D. (1992). An application of Tarantilis, C. D., Stavropoulou, F., & Repoussis, P. P. (2012). A template-based tabu
sizing service territories. European Journal of Operational Research, 63(2), search algorithm for the consistent vehicle routing problem. Expert Systems with
164–172. Applications, 39(4), 4233–4239.
Rossit, D. G., Vigo, D., Tohmé, F., & Frutos, M. (2019). Visual attractiveness in routing Taslimi, M., Batta, R., & Kwon, C. (2017). A comprehensive modeling framework for
problems: A review. Computers & Operations Research, 103, 13–34. HAZMAT network design, HAZMAT response team location, and equity of risk.
Rostami, B., Desaulniers, G., Errico, F., & Lodi, A. (2017). The vehicle routing problem Computers & Operations Research, 79, 119–130.
with stochastic and correlated travel times. Tech. rep. Polytechnique Montréal. Toffolo, T. A. M., Vidal, T., & Wauters, T. (2019). Heuristics for vehicle routing
Sahin, M., Çavuşlar, G., Öncan, T., Sahin, G., & Aksu, D. T. (2013). An efficient heuris- problems: Sequence or set optimization? Computers & Operations Research, 105,
tic for the multi-vehicle one-to-one pickup and delivery problem with split 118–131.
loads. Transportation Research Part C: Emerging Technologies, 27, 169–188. Toth, P., & Vigo, D. (2014). Vehicle routing: Problems, methods, and applications.
Salavati-Khoshghalb, M., Gendreau, M., Jabali, O., & Rei, W. (2019). An exact al- Society for Industrial and Applied Mathematics, Philadelphia.
gorithm to solve the vehicle routing problem with stochastic demands under Toumazis, I., & Kwon, C. (2013). Routing hazardous materials on time-dependent
an optimal restocking policy. European Journal of Operational Research, 273(1), networks using conditional value-at-risk. Transportation Research Part C: Emerg-
175–189. ing Technologies, 37, 73–92.
Schiffer, M., Laporte, G., Schneider, M., & Walther, G. (2017). The impact of synchro- Uchoa, E., Pecin, D., Pessoa, A., Poggi, M., Subramanian, A., & Vidal, T. (2017). New
nizing driver breaks and recharging operations for electric vehicles. Tech. rep. benchmark instances for the capacitated vehicle routing problem. European
Aachen RWTH. Journal of Operational Research, 257(3), 845–858.
Schiffer, M., Schneider, M., Walther, G., & Laporte, G. (2019). Vehicle routing and Vanhove, S., & Fack, V. (2012). Route planning with turn restrictions: A computa-
location-routing with intermediate stops: A review. Transportation Science, 53(2), tional experiment. Operations Research Letters, 40(5), 342–348.
319–343. Vansteenwegen, P., & Souffriau, W. (2009). Metaheuristics for tourist trip planning.
Schneider, M., & Drexl, M. (2017). A survey of the standard location-routing prob- In K. Sörensen, M. Sevaux, W. Habenicht, & M. J. Geiger (Eds.), Metaheuristics in
lem. Annals of Operations Research, 259(1–2), 389–414. the service industry. LNEMS (pp. 15–31). Berlin Heidelberg: Springer.
Schneider, M., & Löffler, M. (2019). Large composite neighborhoods for the capaci- Vidal, T. (2017). Node, edge, arc routing and turn penalties: Multiple problems –
tated location-routing problem. Transportation Science, 53(1), 301–318. One neighborhood extension. Operations Research, 65(4), 992–1010.
Schneider, M., Stenger, A., Schwahn, F., & Vigo, D. (2015). Territory-based vehi- Vidal, T., Crainic, T. G., Gendreau, M., & Prins, C. (2013). Heuristics for multi-attribute
cle routing in the presence of time-window constraints. Transportation Science, vehicle routing problems: A survey and synthesis. European Journal of Opera-
49(4), 732–751. tional Research, 231(1), 1–21.
Shah, S., Bao, F., Lu, C.-T., & Chen, I. R. (2011). Crowdsafe: Crowd sourcing of crime Vidal, T., Crainic, T. G., Gendreau, M., & Prins, C. (2014). A unified solution frame-
incidents and safe routing on mobile devices. In Proceedings of the Ninth ACM work for multi-attribute vehicle routing problems. European Journal of Opera-
SIGSPATIAL international conference on advances in geographic information systems tional Research, 234(3), 658–673.
(pp. 521–524). Vidal, T., Crainic, T. G., Gendreau, M., & Prins, C. (2015). Timing problems and algo-
Shao, J., Kulik, L., Tanin, E., & Guo, L. (2014). Travel distance versus navigation com- rithms: Time decisions for sequences of activities. Networks, 65(2), 102–128.
plexity: A study on different spatial queries on road networks. In Proceedings of Vidal, T., Martinelli, R., Pham, T. A., & Há, M. H. (2019). The capacitated arc routing
the twenty-third ACM international conference on conference on information and problem with time-dependent travel times and paths. Tech. rep. PUC-Rio.
knowledge management (pp. 1791–1794). Villegas, J. G., Prins, C., Prodhon, C., Medaglia, A. L., & Velasco, N. (2013). A
Shelbourne, C., Battarra, M., & Potts, C. N. (2017). The vehicle routing problem with matheuristic for the truck and trailer routing problem. European Journal of Op-
release and due dates. INFORMS Journal on Computing, 29(4), 705–723. erational Research, 230(2), 231–244.
Silva, M. M., Subramanian, A., Vidal, T., & Ochi, L. S. (2012). A simple and effective Wang, X., Fagerholt, K., & Wallace, S. W. (2018). Planning for charters: A stochastic
metaheuristic for the minimum latency problem. European Journal of Operational maritime fleet composition and deployment problem. Omega, 79, 54–66.
Research, 221(3), 513–520. Wong, J. F., & Beasley, J. E. (1984). Vehicle routing using fixed delivery areas. Omega,
Skiera, B., & Albers, S. (1998). COSTA: Contribution optimizing sales territory align- 12(6), 591–600.
ment. Marketing Science, 17(3), 196–213. Xie, F., Potts, C. N., & Bektaş, T. (2017). Iterated local search for workforce scheduling
Smilowitz, K., Nowak, M., & Jiang, T. (2013). Workforce management in periodic de- and routing problems. Journal of Heuristics, 23(6), 471–500.
livery operations. Transportation Science, 47(2), 214–230. Xie, W., & Ouyang, Y. (2015). Optimal layout of transshipment facility locations on
Solomon, M. M. (1987). Algorithms for the vehicle routing and scheduling problems an infinite homogeneous plane. Transportation Research Part B: Methodological,
with time window constraints. Operations Research, 35(2), 254–265. 75, 74–88.
Song, J.-H., & Savelsbergh, M. W. P. (2007). Performance measurement for inventory Yang, W.-H., Mathur, K., & Ballou, R. H. (20 0 0). Stochastic vehicle routing problem
routing. Transportation Science, 41(1), 44–54. with restocking. Transportation Science, 34(1), 99–112.
416 T. Vidal, G. Laporte and P. Matl / European Journal of Operational Research 286 (2020) 401–416

Zajac, S. (2016). The bi-objective k-dissimilar vehicle routing problem. In A. Paias, Zhong, H., Hall, R. W., & Dessouky, M. (2007). Territory planning and vehicle dis-
M. Ruthmair, & S. Voß (Eds.), Computational logistics (pp. 306–320). Berlin Hei- patching with driver learning. Transportation Science, 41(1), 74–89.
delberg: Springer. Zografos, K. (2004). A heuristic algorithm for solving hazardous materials distribu-
Zhang, Y., Baldacci, R., Sim, M., & Tang, J. (2019). Routing optimization with time tion problems. European Journal of Operational Research, 152(2), 507–519.
windows under uncertainty. Mathematical Programming, 175(1–2), 263–305.