ABDELHAMID_LAYOUT_Layout 1 1/16/15 2:37 PM Page 12

Vehicle as a Resource (VaaR)

Sherin Abdelhamid, Hossam S. Hassanein, and Glen Takahara

Abstract
Intelligent vehicles are considered key enablers for intelligent transportation sys-
tems. They are equipped with resources/components to enable services for vehicle
occupants, other vehicles on the road, and third-party recipients. In-vehicle sensors,
communication modules, and on-board units with computing and storage capabili-
ties allow the intelligent vehicle to work as a mobile service provider of sensing,
data storage, computing, cloud, data relaying, infotainment, and localization ser-
vices. In this article we introduce the concept of Vehicle as a Resource and shed
light on the services a vehicle can potentially provide on the road or parked. We
anticipate that an intelligent vehicle can be a significant service provider in a vari-
ety of situations, including emergency scenarios.

W
ith high demand for reducing the number lot with idle resources and capabilities can cooperatively
of vehicular fatalities and enhancing ITS form a powerful resource for services that can benefit a wide
applications and services, many newly range of service domains.
manufactured vehicles will be equipped Another enabler of VaaR is the availability of a multiplicity
with components that will classify them as of wireless communication technologies for communication
intelligent vehicles. Such components include sensors and between a vehicle and its surroundings. One such technology
actuators with intra-vehicle communication, and electronic specifically introduced for vehicular use is the Wireless Access
control units (ECUs) for processing and operation control. for Vehicular Environment (WAVE) technology, which is
Vehicles will be equipped with a wireless communication based on the IEEE 802.11p standard and Dedicated Short
module for supporting three types of communication: Range Communication (DSRC). Another is the Communica-
Between vehicles, known as vehicle-to-vehicle (V2V) com- tion Access for Land Mobiles (CALM) standard. CALM will
munication. support an integrated communication unit that provides many
Between vehicles and infrastructure (V2I and I2V). air interfaces that include 2G/3G cellular, infrared, millime-
Between vehicles and any neighboring object (V2X). ter-wave, mobile wireless broadband (HC-SDMA, 802.16e
In addition, an on-board unit (OBU) will be integrated in [WIMAX/WiBro] and 802.20), satellite, and DSRC. In addi-
each vehicle for interaction with drivers, displaying warnings, tion, some Zigbee communication modules are designed to
issuing alerts, offering automotive services/infotainment, and support vehicular communication. Also, Visible Light Com-
managing the communication with a vehicles surroundings. munication (VLC) is now attracting increasing interest, and
Powerful OBUs can be considered in-vehicle PCs that can many systems have been proposed for the vehicular environ-
handle computing tasks supported by abundant storage capa- ment. Having such technologies available in a vehicle will pro-
bilities. Although currently partially available in some luxury vide flexible communication with its surroundings regardless
models, the availability of such components will be expanded of the type of air interface available.
to most vehicles in the near future. Pivotal components are In this article we delineate the diversified resources a vehi-
shown in Fig. 1. cle can provide as instances of VaaR. We also elaborate on
With these components a vehicle can be considered a how these resources can be tapped into and pooled for per-
mobile resource for many services such as sensing, data forming certain tasks or providing certain services. In addi-
relaying and storage, computing, cloud, infotainment, and tion, for each resource we highlight some potential
localization. We introduce the concept of Vehicle as a applications/services that become feasible with the aid of
Resource (VaaR), which focuses on such vehicular potential. using a vehicle as a provider for this resource. We focus on
The VaaR vision is stimulated by the ubiquity of vehicles the services provided by the vehicle to other vehicles and third
(with predictable mobility patterns) and the vehicular parties, as these have the greatest potential for widely expand-
resources that are readily available. The abundance of such ed benefits. We will also discuss some prominent challenges
on-board resources distinguishes the use of a vehicle as a that face the wide adoption of VaaR.
resource from other mobile resource providers, viz. smart- The remainder of the article is organized as follows. In the
phones, which suffer from limited resources and lack of tra- next section we introduce the VaaR concept and present the
jectory prediction. We anticipate that a vehicle will be a different resources a vehicle can provide in discussing the var-
mobile provider for diversified resources currently unimag- ious instances of VaaR. Illustrative scenarios that show the
ined and will be a key enabler for the revolution of the benefit and applicability of VaaR are discussed after that.
Internet of Things (IoT). Vehicles on a road or at a parking Then we discuss challenges that may face the widespread use
of VaaR and can be considered open points for research. The
The authors are with Queens University. final section concludes the article.

12 0890-8044/15/$25.00 2015 IEEE IEEE Network January/February 2015

ABDELHAMID_LAYOUT_Layout 1 1/16/15 2:37 PM Page 13

In-vehicle sensors On-board unit Electronic control

Communication CAN serial unit
unit/box bus

Figure 1. In-vehicle components shaping the intelligent vehicle.

Vehicle as a Resource (VaaR)

VaaR-Sensing In-vehicle
sensors
According to the automotive sensors market growth in North
America, the average number of sensors per vehicle reached
70 in 2013 [1], with as many as 100 in some luxury vehicles. A
vehicle can then be considered a significant resource of senso- Safety Diagnostics Convenience Environment
ry data. We categorize sensors according to their application sensors sensors sensors monitoring
sensors
domain: safety, diagnostics, convenience, and environment
monitoring, as shown in Fig. 2. Safety sensors are the most
crucial, being targeted at decreasing accidents and driving Figure 2. Categories of in-vehicle sensors.
fatalities. Diagnostics sensors provide on-board detection of
component malfunction to avoid further breakdowns or dam-
age. In addition to on-board services, the system can include a send the processed data to database servers through the Inter-
reporting capability to support remote diagnostics. Conve- net for further analysis and publishing. Examples of such plat-
nience sensors support comfort and convenience applications forms are discussed in [3].
for drivers and passengers. Most of these sensors are deployed Currently much attention is focused on utilizing in-vehicle
inside the vehicle compartment to provide direct services for sensors for providing road condition monitoring services.
its occupants, while others are deployed to provide driving Many techniques have been proposed from using accelerome-
assistance. Finally, environment monitoring sensors provide ters to the use of cameras for capturing photos and videos
alerts/warnings about road hazards or reports about traffic, that are further analyzed to extract road features. A promi-
road, and weather conditions. nent road monitoring platform is CarMote [4].
Having a variety of sensors along with communication
capabilities shapes the concept of VaaR-Sensing. Many appli- VaaR-Data Storage and Computing
cations and platforms have been proposed to make use of Data storage and computing are tightly coupled as both ser-
vehicles as sensing resources, mainly for environment moni- vices are offered through the on-board computers in intelli-
toring as their measured data can be of benefit beyond a vehi- gent vehicles. Advanced in-vehicle computers are currently
cles compartment. In these applications, vehicles sense/ available, some almost as powerful as personal computers,
monitor the surrounding environment and store the sensed such as those in [5] with dual core processors up to 2.8 GHz
data for further relaying, either without processing or after and storage capabilities in Gigabytes. Due to advances in data
processing to search for data of interest. Such data can be storage technologies, it is anticipated that in-vehicle storage
reported to third parties through the Internet or V2X (Vehi- capacity will reach multiples of Terabytes in the future,
cle-to-Any) communications. These third parties can be data enabling the vehicle to act as a mobile data server.
centers that can offer the data for public/commercial services, With such powerful computing resources, it is foreseen that
or they can be other mobile users/drivers. Such sensing ser- computing tasks would be offloaded to vehicles. Vehicles on a
vices can expedite the adoption of public sensing in its partici- road or a parking lot can be considered a distributed system
patory and opportunistic forms. that can potentially manage computing tasks more efficiently
An example of a vehicular sensing platform is the MobEyes and cost-effectively than a centralized computing center.
platform [2], which utilizes vehicles as mobile sensors to moni- Many platforms have been proposed to utilize vehicles as a
tor surroundings, recognize objects, store data, and advertise resource for data storage for further diffusion or retrieval of
this data for potential sharing. The MobEyes platform uses the data by other agents. Vehicle-generated data or data
V2V communication. Other platforms are Internet-based, obtained from neighboring vehicles can be stored until the
where vehicles collect and process sensor data locally then vehicle reaches a dedicated data collector or kept in the vehi-

IEEE Network January/February 2015 13

ABDELHAMID_LAYOUT_Layout 1 1/16/15 2:37 PM Page 14

cle until retrieved as a reply to queries sent by data-seeking on vehicles as data relays. In DB-VDG, queries can be sent to
vehicles/agents. This communication paradigm is considered a retrieve information from areas of interest and replies can be
case of delay/disruption-tolerant networks. routed back from vehicles in these areas. Both queries and
The MobEyes platform mentioned earlier is an example of replies are delivered via intermediate vehicles that act as relays.
both VaaR-Sensing and VaaR-Data Storage. Each vehicle DB-VDG can also be considered an example of VaaR-Sensing
keeps the sensed data in local storage, and advertises its stored as the gathered data is acquired from the in-vehicle sensors.
data via meta-data. While moving, other vehicles harvest the Vehicles can be good candidates for commercial advertise-
meta-data and send queries to obtain data of interest. Another ments that utilize vehicular mobility to disseminate the ads
example is the FleaNet platform [6] that utilizes vehicles as through relays. These commercial advertisements can be pushed
information traders. In FleaNet, vehicles are used to resolve to a vehicle either through the Internet or by Adstations. Exam-
queries generated by other vehicles. While moving, vehicles ples include store special offers or restaurant menus.
receive queries and data advertisements from other vehicles VaaR-Data Relaying can also be a technique for extending
and/or information generated by roadside advertisement sta- network coverage, as presented in [10]. The authors propose a
tions (Adstations), and store this data/information locally. vehicle-to-vehicle relay (V2VR) scheme to extend a road side
When vehicles receive queries, they try to resolve them by con- access point (AP) service range and allow drive-thru vehicles to
sulting their own local storage to provide a possible match. have an extended coverage range. In this scheme, when a vehicle
Currently there is an emerging vision of utilizing intelligent approaches an AP, it selects a vehicle ahead to work as a relay of
vehicles for cloud services. With sensing, computing, and stor- the AP traffic and allow early access to the AP services. When a
age resources, and abundant power supply and communication vehicle is about to leave the direct AP coverage range, it selects
modules, a vehicle can work as a powerful cloud. As with con- another vehicle behind to work as a relay for the AP traffic and
ventional clouds, the owner of a vehicle can rent out vehicular extend the access time while maintaining high throughput.
resources on demand when these resources are not in use. In addition to relaying data to/from other vehicles and
Olariu et al., [7] introduced the term autonomous vehicular RSUs, vehicles can act as data mules and mobile sinks in
clouds (AVCs), arguing that in-vehicle resources may be wireless sensor networks to assist in their operations. VaaR-
underutilized by traditional vehicular applications as motiva- Data Relaying can also play a pivotal role in cases of emer-
tion to take vehicular networks to the clouds. They remark gency or natural disasters where infrastructure may be broken
that the benefit of on-board resources will be maximized if down. In these cases vehicles can help deliver critical data to
resources of multiple vehicles are combined. crisis/disaster-management authorities, collect environmental
VaaR-Cloud has been proposed as a mobile experimental data, and inform emergency vehicles about optimal routes.
laboratory in areas with limited facilities that hinder the use
of a remote service. Another scenario utilizes idle parked VaaR-Infotainment
vehicles in a parking lot of a company as a distributed com- Infotainment refers to the combination of information with
puting asset. Tasks can be offloaded to vehicles during the entertainment. With the on-board communication capabilities
workday in lieu of building or renting an outsource infrastruc- supporting communication with surrounding mobile agents
ture. Vehicle owners can be compensated so both the compa- (e.g. other vehicles, cellular phones, or communication-
ny and employees can benefit. Similar arguments can be enabled handheld devices), and providing the possibility for
applied to vehicles parked at an airport. Resources are ample always-on Internet access, a vehicle can be a great source of
and travelers vehicles may be left unutilized for days, result- infotainment.
ing in computing resources that, if managed properly, could Some vehicular infotainment services are solely Internet-
turn an airport parking lot into a data center. Travelers can dependent (e.g. web surfing, email access, video downloads,
share their travel plans with the airport, which can then and online gaming), and others are supported by inter-vehicle
schedule and manage the resources. Parked vehicles that are communications (e.g. exchanging information and files
part of the vehicular cloud would be plugged into a standard between vehicles). More detailed examples of these infotain-
power supply and can be provided with an Ethernet connection. ment services can be found in [11]. The entertainment part of
Vehicular clouds can provide advantages over fixed clouds the infotainment resource is usually for the benefit of the
in applications such as the mobile experimental laboratory. vehicles occupants and is not the focus of this article, which is
Another advantage of vehicular clouds is their autonomous on the information part.
formation. Neighboring vehicles can autonomously form a A vehicle can provide others with information obtained
cloud to provide instantaneous services (e.g. collecting traffic from its own sensors, other vehicles, RSUs, direct access to
information at congested intersections for traffic light man- the Internet, or Adstations. This information may include traf-
agement). A detailed comparative study of vehicular and con- fic and environmental conditions, navigational and safety
ventional clouds is presented in [8]. information, parking availability, or commercial advertise-
ments. Many of the vehicular platforms mentioned earlier in
VaaR-Data Relaying the article can be considered examples of VaaR-Infotainment
Wireless message exchange between vehicles is known as when the utilization of the other vehicular resources is ulti-
inter-vehicle communication and a network of communicating mately for the sake of providing information.
vehicles is known as a Vehicular Ad-Hoc Network (VANET). As most of the infotainment applications are real-time,
A VANET is a sub-category of wireless multi-hop networks in such applications face challenges caused mainly by the highly
which a source depends on intermediate nodes to relay mes- dynamic vehicular topologies. These challenges must be dealt
sages to a destination. In this communication paradigm, vehi- with to support the required QoS levels. The work in [12] pre-
cles can be considered a resource for relaying data to other sents some of these challenges along with proposed solutions
nodes out of the communication range of the source node. A and some issues that still need consideration.
vehicle can be used not only for relaying data to other neigh-
boring vehicles but also to/from road side units (RSUs). VaaR- Localization
As an example of VaaR-Data Relaying, Delay-Bounded Vehicles can be considered potential resources for locating
Vehicular Data Gathering (DB-VDG) [9] is a solution that sup- objects. Through their sensing and communication capabili-
ports geographical data gathering services and depends heavily ties, they can recognize and locate objects, and send this

14 IEEE Network January/February 2015

ABDELHAMID_LAYOUT_Layout 1 1/16/15 2:37 PM Page 15

VaaR-storage
information. For instance, MobEyes can be used
to enable vehicles to recognize the license plate
numbers, store them, and broadcast representa- VaaR-sensing
F
tive meta-data. Other mobile agents (e.g. police
patrol cars) can retrieve the recognized plate
numbers to locate and track lost or stolen vehi- G
cles. VaaR-info
Vehicles can be used to locate neighboring
vehicles for the sake of estimating the distance to H
these vehicles or informing them about their
positions for accuracy purposes, and so. An exam-
ple of this positioning ability is the Intel patent A
Visible Light Communications and Positioning VaaR- E
(VLCP) [13] for proposing the use of visible light localization B
communication as an inter-vehicle communica-
tion technique and introducing a scheme for
positioning neighboring vehicles. C
Vehicles can also use self-localization tech- VaaR-
niques to determine their own position, comple- relaying
menting and/or refining GPS information. These D
self-benefit services are out of this articles focus.

VaaR in Action Figure 3. Illustrative scenario showing the viability of VaaR. Vehicles G and
H had an accident and vehicles F, A, B, and C work as resource providers
In this section we discuss scenarios illustrating while being in the vicinity of the emergency situation. Vehicle E as well
some of the various resources a vehicle can pro- works as a resource after detecting falling rocks on its way.
vide on the move. The main scenario depicted in
Fig. 3 shows an emergency scenario caused by an
accident on the road. Privacy, Quality, and Redundancy Pose Major
Vehicles G and H had a head-on collision. Due to that
collision, traffic has stopped moving in the vicinity of the
Challenges in Collecting Data from and Offloading
collision. Vehicle D is a truck carrying goods that need to Tasks to Vehicular Resources
be delivered on time for shipping. The driver of the truck
needs to know how severe the situation is. He sends a The open access and resources sharing that VaaR promises
request using his vehicles OBU asking for information bring privacy to the forefront. Since a vehicles resources may
about the collision from the neighbors of the involved vehi- be accessible to many users to share, data privacy should be
cles. He receives a reply from vehicle A that it can provide guaranteed for all users sharing these same resources. Privacy
live video streaming of the situation. He agrees and starts should be maintained for vehicles owners as well. This issue
getting live video streaming from node A. Vehicles A and can be handled by virtualization and scheduling techniques for
D are out of communication range of each other. There- coordinating access to shared resources. Other techniques can
fore, they depend on vehicles B and C to work as data be used for protecting data and controlling its access to its
relays for their shared content. In this scenario, vehicle A owner only. Examples of such techniques include the use of
can be considered an example of VaaR-Infotainment, while personal data vaults, which are individually controlled data
vehicles B and C can be considered examples of VaaR- repositories. Privacy is needed as well to hide the identity of
Data Relaying. participants when they prefer to remain anonymous.
As a consequence of the collision, vehicle H had a malfunc- Pseudonymity can be considered a solution that can help in
tion in its GPS. The driver of vehicle H needs to know its hiding the actual identities for the sake of reducing the possi-
location to call for help. He can depend on the resources of bility of linking between the sent data and participants.
vehicle A to locate his vehicle. In this case, vehicle A can be The quality of information (QoI) retrieved from vehicles is
considered an example of VaaR-Localization as well. also a concern. Data/information retrieved needs to be veri-
As a direct neighbor of vehicle G, vehicle F has recorded fied and validated before making decisions and/or publishing
the collision using its on-board camera. It decides to store the to the public. Also, quality reports about cooperating vehicles
recorded video for further reporting. Therefore, vehicle F can and corresponding reputation scores should be maintained for
be considered an example of VaaR-Data Storage. future reference. Depending on the criticality of the support-
Finally, on the other side of the road vehicle E has detect- ed services, such reputation scores can be computed based on
ed falling rocks using its in-vehicle sensors. It reports this haz- either short-term or long-term participation history with the
ard using its communication module to the municipality to aid of watchdog modules. Tasks can be offloaded to those
take proper action to avoid accidents. This is considered an participants with reputation scores higher than a certain
example of VaaR-Sensing. threshold to guarantee a certain level of QoI. Feedback can
be given to participants about the quality of their reported
data and their perceived reputation levels.
Challenges and Open Issues While having a plethora of vehicles on the road increases
Although VaaR brings a wide range of benefits to various the pool of resources, it may lead to having large amounts of
applications, many challenges face its wide penetration and redundant data reported, resulting in the waste of both com-
adoption. In this section we highlight some of the prominent munication and computation resources. Such redundancy can
challenges and open issues. A summary of the challenges and be eliminated by adopting data aggregation and fusion tech-
some potential solutions can be found in Table 1. niques. Vehicular clusters and chains can be formed on the

IEEE Network January/February 2015 15

ABDELHAMID_LAYOUT_Layout 1 1/16/15 3:46 PM Page 16

Challenge Potential solutions

Applying virtualization and scheduling techniques.

Privacy Use of personal data vaults.
Hiding identities through pseudonymity.

Adopting verification and analysis techniques to build quality reports and history.
Data quality Computing reputation scores to be considered in selection refinement.
Giving feedback to participants about their data quality.

Can be eliminated by adopting data aggregation and fusion techniques while forming vehicular
Redundancy clusters and chains with representative nodes.
Can be avoided with proper participant selection.

Adopting effective resource management techniques for proper task assignment and retrieval.
Dynamic availability of
Adopting prediction techniques to anticipate the availability spatiotemporal span.
resources
Considering task sharing and handover techniques.

Adopting P2P incentive mechanisms.

Incentives for offering
Offering vehicles owners some rewards in return.
resources
Focus on the encouraging monetary incentives and their corresponding pricing models.

Proposing efficient recruitment mechanisms that consider the rewards, budget, and data quality.
Participant selection
Studying the feasibility of adapting the smartphone recruitment schemes.

Powering up parked vehicles Utilizing features of the CAN bus connecting the intra-vehicular resources.
PCs Use of scheduling and mobility prediction mechanisms.

Considering data fusion and information integration techniques to merge data/information from
Consolidation of different
different sources.
resource providers
Considering ongoing research in the area of integration of heterogeneous networks.

Table 1. VaaR challenges and potential solutions.

road while electing a leader node to take care of the intra-clus- and direct service/reward must be provided in return. Some of
ter/chain data aggregation. It is noteworthy that the success of the incentive mechanisms proposed for P2P file sharing [14]
data aggregation depends on the type of road (e.g. highway, in- can be applied for incentivizing vehicular users, especially for
city urban road, rural road) since it determines the speed of those applications that have major commonalities with the
vehicles, and hence the ability to select the cluster/chain leader P2P paradigm. In general, incentives can be of three types:
node. Also, redundancy of the reported data can be avoided by 1. Willingness to serve the public.
applying proper selection and recruitment techniques that 2. Receiving service in return.
ensure selecting participants with minimal overlapping. 3. Receiving monetary returns.
Earlier studies have shown that incentives with monetary val-
The Intermittent/Dynamic Availability of Resources ues are the most effective. Monetary incentives can be in the
Hinders Extended Usage form of pecuniary returns, vouchers, or passes. Another incen-
tive for the owners of parked vehicles can be offering them
Vehicles mobility is considered as a plus that allows vehicles free parking, while their idle resources are being accessed.
to cover wider areas compared to their static counterparts. The value of such incentives can be determined by the data
However, while pooling resources from a vehicle, the vehicle collecting/task distributing party through the use of pricing
may leave the area of interest or connectivity before report- models that can consider the level of participation and QoI,
ing/relaying the desired data, or finishing the task in hand. or through reverse auction techniques that are currently
Dynamic availability of resources may also be temporal stem- becoming popular in the rewarding models.
ming, for instance, from the high need for resources during
rush hours and their idleness during late nights. Effective Recruitment Mechanisms Should Be Deployed for
resource management techniques are needed for proper task Efficient Selection of Participants
assignment and retrieval. A vehicles cyber-physical existence
should be taken into consideration in pooling/tapping into In urban environments, offloading tasks to vehicles to utilize
vehicular resources. This is influenced by the type and speed their resources is challenging in terms of participant selection,
of roads which, in turn, affect how long a vehicle would be as there can be many potential participants in an area of inter-
present. Prediction techniques can help in anticipating the est that may reach hundreds of vehicles in a congested area.
spatiotemporal availability span, and hence taking more These participants cannot all be recruited for a task, as this
informed allocation decisions. Task sharing and handover can would result in a higher cost in terms of participants rewards.
also be potential solutions of the partial availability. Selecting the right number of potential participants in a cost-
effective manner, while providing an acceptable level of service,
Incentives Are Needed to Encourage Owners to is an open area. Such recruitment mechanisms must also con-
Offer Their Vehicular Resources sider the QoI challenge mentioned earlier. For load balancing
and robustness purposes, these mechanisms should also ensure
The VaaR concept cannot thrive without active user participa- that the same vehicles are not always chosen. A scheme that
tion, especially if users (vehicle owners) lose their willingness aims at minimizing the number of recruited participants while
to participate and share resources. Some form of incentive achieving a given level of coverage for an area of interest can

16 IEEE Network January/February 2015

ABDELHAMID_LAYOUT_Layout 1 1/16/15 3:46 PM Page 17

be found in [15]. Although the recruitment schemes proposed References

for utilizing the resources of smartphones cannot be directly [1] W. J. Fleming, New Automotive Sensors A Review, IEEE Sensors J.,
vol. 8, no. 11, 2008, pp. 190021.
applied to vehicles due to the different design considerations, a [2] U. Lee et al., Dissemination and Harvesting of Urban Data Using Vehicu-
study of potential adaptation of such schemes is worthwhile. lar Sensing Platforms, IEEE Trans. Veh. Technol., vol. 58, no. 2, 2009,
pp. 882901.
How to Power up the Resources of Parked Vehicles [3] U. Lee and M. Gerla, A Survey of Urban Vehicular Sensing Platforms,
Computer Networks, vol. 54, no. 4, 2010, pp. 52744.
for Tasking Is a Technical Challenge [4] A. Mednis, A. Elsts, and L. Selavo, Embedded Solution for Road Condi-
tion Monitoring Using Vehicular Sensor Networks, Proc. 6th IEEE Intl
Parked vehicles considered for resource utilization should be Conf. Application of Information and Communication Technologies
plugged into a power supply. However, it is not logical to (AICT12), 2012, pp. 15.
[5] SINTRONES Technology Corp. (2013) In-Vehicle Computing, available:
keep their PCs on all the time waiting for potential tasking. http://www.sintrones.com/applications/in-vehicle_computer.php (accessed
Solutions are needed to power these PCs only when required. 2013, Dec. 2).
These solutions can be categorized into on-demand and [6] U. Lee et al., FleaNet: A Virtual Market Place on Vehicular Networks,
pre-scheduled techniques. An on-demand approach [16] IEEE Trans. Veh. Technol., vol. 59, no. 1, 2010, pp. 34455.
[7] S. Olariu, M. Eltoweissy, and M. Younis, Towards Autonomous Vehicular
requires that ECUs and the in-vehicle PC are connected Clouds, ICST Trans. Mobile Commun. Applications , vol. 11, no. 79,
through the Controller Area Network (CAN) bus working as Sept. 2011, pp. 111.
the main intra-vehicle communication bus. One of the fea- [8] M. Whaiduzzaman et al., A Survey on Vehicular Cloud Computing, J.
tures of CAN-connected nodes is that they can operate in a Network and Computer Applications, vol. 40, Apr. 2014, pp. 32544.
[9] C. E. Palazzi, F. Pezzoni, and P. M. Ruiz, Delay-Bounded Data Gather-
sleep mode such that while a vehicle is stopped, these nodes ing in Urban Vehicular Sensor Networks, Pervasive and Mobile Comput-
consume a minimum amount of battery. The CAN-connected ing, vol. 8, no. 2, 2012, pp. 18093.
in-vehicle PC of a parked vehicle can be remotely powered up [10] J. Zhao et al. , Extending Drive-Thru Data Access by Vehicle-To-Vehi-
on-demand. However, experiments are needed to test the fea- cle Relay, Proc. 5th ACM Intl. workshop on VehiculAr Inter-NETwork-
ing (VANET 08), San Francisco, California, USA, Sep. 2008, pp.
sibility of the wakeup/probing techniques. Other techniques 6675.
that fit under the second category include deploying schedul- [11] M. Gerla and L. Kleinrock, Vehicular Networks and the Future of the
ing mechanisms to schedule the power up times a priori and Mobile Internet, Computer Networks , vol. 55, no. 2, 2011, pp.
sending this schedule whenever vehicles are on. Such tech- 45769.
[12] H. T. Cheng, H. Shan, and W. Zhuang, Infotainment and Road Safety
niques may also include assistance from mobility prediction Service Support in Vehicular Networking: From a Communication Perspec-
mechanisms to anticipate a vehicles daily parking times. tive, Mechanical Systems and Signal Processing, vol. 25, no. 6, 2011,
pp. 202038.
Consolidation of Different Resources Is Expected and [13] Intel Corporation White Paper, (2010) Visible Light Communications
and Positioning (VLCP) The Automotive Use Case, available:
Interoperability Should Be Guaranteed http://www.connectedvehicle.org/VisibleLight.pdf (accessed: 2013, Dec, 2).
[14] M. Salek, S. Shayandeh, and D. Kempe, You Share, I Share: Network
It is expected that vehicles as resources of services will be Effects and Economic Incentives in P2P File-Sharing Systems, Lecture
integrated with other resource providers. For example, VaaR- Notes in Computer Science, vol. 6484, 2010, pp. 35465.
[15] S. A. Hamid, G. Takahara, and H. S. Hassanein, On the Recruitment of
Sensing may be merged with other sensing resources such as Smart Vehicles for Urban Sensing, Proc. IEEE Global Communications
smartphones and sensor networks, with each paradigm being Conf. (GLOBECOM 13), Atlanta, GA, USA, Dec. 2013, pp. 5863.
assigned a part of a sensing task based on its availability and [16] M. Mahfoud, N. Al-Holou, and R. Baroody, Next Generation Vehicle
capabilities. Such consolidation of resources entails the devel- Network: Web Enabled, Proc. 3rd Intl. Conf. Information and Communi-
cation Technologies: From Theory to Applications (ICTTA 08), Umayyad
opment of communication, synchronization, and resource Palace, Damascus, Syria, Apr. 2008, pp. 17.
management techniques that ensure syntactic and semantic [17] A. A. Atayero et al., Heterogeneous Wireless Networks: A Survey of
interoperability and seamless handover among the different Interworking Architectures, Intl J. Engineering and Technology, vol. 2,
paradigms. Sensor fusion techniques can be considered to no. 1, 2012, pp. 1621.
handle the collaborative-sensing example mentioned above.
Generally speaking, information fusion/integration techniques Biographies
should be considered to handle merging information from dif- SHERIN ABDELHAMID ([email protected]) received her Ph.D. from the School
ferent sources. Work proposed for the operation of heteroge- of Computing Queens University, where she is a member of the Telecom-
neous networks (HetNets) is promising [17]. munications Research Lab. She received her B.Sc. and M.Sc. degrees in com-
puter sciences in 2005 and 2009, respectively, from Ain Shams University,
Egypt. Her research interests include recruitment and data delivery for vehicu-
Conclusions lar ad-hoc networks, routing for wireless multi-hop networks, information-centric
networks, and public sensing. Sherin has several publications in top venues
With its diversified on-board resources, an intelligent vehicle and is the winner of the best paper award of the AHSN symposium of GLOBE-
can be considered a mobile service provider that can assist in COM 2013.
a wide scope of applications and domains. In this article we H OSSAM H ASSANEIN ([email protected]) is a leading authority in the
introduced the concept of Vehicle as a Resource (VaaR) to areas of broadband, wireless and mobile networks architecture, protocols and
unveil the potential of an intelligent vehicle on the road or in performance evaluation. His record spans more than 500 publications, in
a parking lot. We showed that a vehicle can be a resource for addition to numerous keynotes and plenary talks in flagship venues. He has
received several recognitions and best paper awards at top international con-
sensing, data storage, computing, cloud, data relaying, and ferences. He is also the founder and director of the Telecommunications
infotainment, and a means for locating other objects. We also Research Lab at Queens University School of Computing. He is a senior mem-
presented demonstrating scenarios, challenges, and open ber of the IEEE, and is a former chair of the IEEE Communication Society
issues related to the adoption of VaaR. With the presented Technical Committee on Ad hoc and Sensor Networks. He is an IEEE Commu-
nications Society Distinguished Speaker.
VaaR vision, we anticipate that an intelligent vehicle will lead
the information services and the intelligent transportation sys- GLEN TAKAHARA ([email protected]) received a B.A. degree in mathe-
tems to an era of service revolution. matics from the University of British Columbia, Vancouver, B.C., in 1988, and
M.S. and Ph.D. degrees in statistics from Carnegie Mellon University in 1990,
Acknowledgment and 1994, respectively. Currently he is an associate professor in mathematics
and engineering in the Department of Mathematics and Statistics at Queens
This research is supported by a grant from the Natural Sci- University in Kingston, Ontario, Canada. His research interests are in commu-
ences and Engineering Research Council (NSERC) of Canada. nications theory, networking, applied probability, and statistics.

IEEE Network January/February 2015 17