He, Tan, Lee, and Li (2009) proposed a system for continuous
tracking in supply chain. The model used technologies such as
web services, RFID and GPS to integrate RFID events with geolocation
information and associated them to the load, so it can easily be
tracked. The GPS information is gathered by a mobile device application.
This application sends information about the location to a
central server named EPCIS Gateway. The authors presented a scenario
in which a vehicle is equipped with a mobile device that has
GPRS and GPS receivers. Thus, the mobile application can communicate
with the EPCIS Gateway to transmit location data and business
context information. This information was stored in the
database by EPCIS Gateway and available through web services so
that other applications can access and use this information for various
purposes.
The system proposed by Yang et al. (2010) presented a hybrid
cargo model for tracking and tracing. This system used GPS and
wireless sensor networks to obtain the vehicle and cargo positions.
The authors analyzed, discussed and identified the main logistics
tracking systems, also they proposed a low-cost model able to perform
the cargo tracking in a continuous way. The system architecture
was divided in three components: a Hybrid Network
Infrastructure, an Intelligent Monitoring Devices and a Central
Server. The Hybrid Network Infrastructure integrated technologies
GPS, Wi-Fi, RFID and ZigBee to perform the entities tracking, while
GSM/3G, Wi-Fi and ZigBee were used for communication between
the components. The Intelligent Monitoring Devices were embedded
devices, that had modules for motion detection, RFIDs and communication
through ZigBee networks. Furthermore, these devices had
a micro-controller and a battery. The Central Server provided services
so that different logistics applications are developed.
According to the authors, the proposed system achieves higher
availability, accuracy and lower cost than existing systems.
However, the work did not present how the location information
was modeled. So it is not possible to know how the information
was represented to perform a continuous tracking.
Papatheocharous and Gouvas (2011) presented a cargo tracking
system, named eTracer. The work also proposed to manage the
loading and unloading of products in an automated way. The system
architecture was divided in four components: Mobile
Logistics Stations Network, Fixed Logistics Stations Network,
Communication Server and Web Application Server. Mobile Logistics
Stations Network were vehicles used to cargo transport. The
Mobile Station had RFID readers that identified the entry and exit
of all objects that had RFID tags. Each vehicle also had GPS receivers
and GPRS transmitters to provide the location of objects at
any time. The Fixed Logistics Stations Networks were depot stations
that had distributed RFID readers in their inputs and internal divisions.
Thus, the fixed stations detected and identified the cargo entry or exit and transmitted information about the products
through wireless network to a local server. Subsequently, the local
server sent the information to the Communication Server, which
performed the communication between other system components.
Furthermore, the Communication Server semantically classified the
information from cargo and stored in a relational database.
Through the Web Application Server authorized users could access
statistical surveys, reports and cargo tracking.
Zhengxia and Laisheng (2010) presented a logistics monitoring
platform based on internet of things. The proposal used widespread
technologies, such as: EPC, RFID, GPS, GPRS, GSM and
WSN. The work introduced concepts about dynamic and static
EPC in order of reduce the data flow between client and server
devices. Furthermore, it could manage data regarding loads in a
simple and efficient way. The platform architecture had three layers:
Data Acquisition, Data Transport and Background Data
Processing. The Data Acquisition layer was in charge of getting information
about loads, through EPC and GPS technologies. According
to the EPC functions, every load can receive a unique EPC number.
However, the authors developed a static and dynamic EPC model,
which reduced the data flow and reused EPC numbers. The Data
Transport layer did the communication between components using
GPRS. The Background Data Processing received and processed the
loads information, storing in database. This layer provided information
about load history, statistics analysis and products reports.
Even more, it had alerts that inform when a product expired or it
was delayed.
Musa, Gunasekaran, Yusuf, and Abdelazim (2014) presented a
study for RFID technology usage in supply chain processes, in
which they identified the intra- and inter-entreprise location and
visibility of resources and products. With that, the model provided
a collaboration and integration all through the supply chain
network.That was possible due to the adoption of the EPCglobal
network, which is an international model based on RFID. The main
contribution of the work was the continuous and seamless visibility
of fixed and mobile resources from smart enterprises. For the
prototype implementation, the model employed an embedded
microsystem, integrating RFID, GPRS, GPS and environmental sensors
for logistics’ applications.
Table 1 summarizes the comparison of related works. All considered
systems employed the GPS and RFID technologies. The
monitoring of cargo trucks through GPS was mainly used because
it is a widespread technology that has a wide coverage area. The
RFID was used to perform the monitoring of goods, because it is
considered by the industry as a standard for items and goods identification.
Related works used some mobile technologies to acquire
the device position.
He et al. (2009), Papatheocharous and Gouvas (2011) and
Zhengxia and Laisheng (2010) used off-the-shelf mobile devices,
while Yang et al. (2010) and Musa et al. (2014) use a hardware
built specifically for this purpose. Nowadays, off-the-shelf mobile
devices, such as smartphones and tablets, have high processing
power and storage. In addition, they have several resources, such
as SMS messaging, temperature sensors, humidity, accelerometer
and NFC. Another advantage is the fact that they provide high-level
platforms for software development, easing the implementation
and integration with other devices. However, off-the-shelf mobile
devices limit the system flexibility, because it is restricted to the
mobile device hardware.
The method of the tracking is also something important to consider.
Yang et al. (2010) and Musa et al. (2014) performed tracking
in real-time and continuously, through own built hardwares.
However, it was unclear how continuous monitoring was used,
since these works did not present how the information from the
various sources of location were aggregated and represented.
When it comes to delivery management, both eTracer
(Papatheocharous & Gouvas, 2011) and Musa et al. (2014) performed
it automatically. eTracer used sensors located at the access
points in every delivery spot to consider if a load had to be dispatched
there. One drawback of eTracer is a false positive, in cases
where the deliveryman did not remove the load from the carrier. In
this case, the system would consider that the delivery was successful,
but erroneously. Musa et al. (2014) used the Smart Tag hardware
in order to do continuous tracking. However, this approach
had some drawbacks regarding its implementation costs. This is
because every tag incorporates several sensors, making the final
hardware cost higher if compared to other approaches.
Furthermore, every goods needed an assigned smart tag.
All related works manually performed travel management, i.e.
systems require the user interaction to initialize and finalize travels.
We believe that in logistics systems it is important to shorten
the interaction of the user to optimize time and reduce problems
caused by forgetfulness people.
Another relevant point for tracking and tracing systems is
related to the disconnection support. It is a known fact that network
might not be available in some locations. In this case the system
has to provide a way to tolerate communication’s failures, in
order to guarantee that the information will not be lost in ‘shadow
zones’ (areas without GPRS coverage that impede the device to
send data to the server). Among the studied works, just Yang
et al. (2010) did that support.
Regarding sending of alerts, just the work of Zhengxia and
Laisheng (2010) provided it automatically. The alerts are sent via
SMS for the monitoring users. That is an important issue, because
it reduces the response time and allows a quicker decision-making
without depending of an user. In their proposal, several events can
be alerted, such as overdue validity, long stayed goods, and damaged
goods. However, as the alerts were sent by SMS, the driver
may not notice the arrival of a SMS message on the mobile device.
This would cause a delay in decision-making and possibly losses
during logistics flow. Thus, an improvement of SMS alerts with
audible alerts would be interesting.