Page 84 - ITU Journal Future and evolving technologies Volume 2 (2021), Issue 5 – Internet of Everything

ITU Journal on Future and Evolving Technologies, Volume 2 (2021), Issue 5

3.1 Architecture Layer appropriate
mechanisms both for the provisioning, management,
The physical entities that collectively carry out the ac‑ monitoring and optimization of the Infrastructure Layer
tual logistics processes are denoted as the Infrastructure
as well as for the effective coordination of resources and
layer. These refer to a variety of concepts, including cargo
their actions towards business objectives. A key compo‑
(parcel, palette, container, etc.), transportation means
nent of the Coordination Layer is thus the Social Internet
(vehicles, trucks, trains, etc.), back‑end ICT systems and of Things.
services, as well as infrastructure components, like hubs,
The Intelligence Layer provides the necessary logical in‑
parking places, ports and other. The entities of the infras‑
ference mechanisms for knowledge extraction and for‑
tructure are complemented by the datasources, including
malisation, learning and reasoning, as well as cognitive
any source of data such as Enterprise Resource Planning,
behaviour of the underlying entities. To achieve this, the
Warehouse Management System, Intelligent Transporta‑
COG‑LO intelligence Layer consists of multiple analytics
tion Systems, Traf ic Information Systems, along with op‑
technologies able to merge and aggregate data from dif‑
erational and con iguration data that are essential for the
ferent logistics entities and CLOs, to identify patterns, and
operation of COG‑LO.
to propose operation improvements. Secondly, predic‑
tive analytics with event processing enables foreseeing
the impact of state changes of one or more operations and
determining the corresponding effects on multiple stake‑
holders. COG‑LO Intelligence Layer is coupled with op‑
timization algorithms and heuristics for enabling CLOs
adaptation to operational changes from the external en‑
vironment in near‑real time. In particular, COG‑LO cou‑
ples analytics and optimization for considering the effect
of optimization control measures to the performance of
operations in an environment with continuous external
variations.
Finally, the Application Layer consists of the Cognitive
Advisor (CA) and the Tweeting CLOs. The CA provides the
logistics operator with visual decision support for rout‑
ing optimisation. The CA interacts with the MSB to visu‑
alise the formalisation, reasoning and cognitive outputs
of the Intelligence and Coordination Layers. The Tweet‑
ing CLO is the generic prototype consisting of the appro‑
priate APIs that receive the messages from the logistics
Fig. 2 – COG‑LO Architecture objects and vehicles and transmit them to the CLOs in the
same network and to the CA via the MSB. Fostering inter‑
In order to provide for effective interaction, an Inter‑ operability, semantic data integration and operational ef‑
fectiveness, COG‑LO components rely on semantic ontolo‑
action Layer has been created. It includes a Message
gies for grounding data being collected, processed and
and Service Bus (MSB) which acts as mediation middle‑
disseminated, as well as establishing a common under‑
ware between the components of the COG‑LO ecosystem.
standing between the collaborating entities.
The MSB comprises a message‑oriented system provid‑
ing both asynchronous and point‑to‑point message ex‑
3.2 Data model
change between the system entities, circulation of events,
and interaction between the CLOs. Furthermore, the MSB
The COG‑LO integrates data from various sources that
provides the integration of the infrastructure objects and
correspond to stakeholders with different roles in the lo‑
data sources, by means of appropriate connectors. To this
gistics domain. Τhe provenance of data and the use cases
end, a fundamental duty of the MSB is the transforma‑
supported by this data exhibit great variety in their na‑
tion of the Platform Independent Model (PIM) of the un‑
ture and cover a broad range of logistics services. Thus,
derlying components operational behaviour to the COG‑
the main challenges faced were (i) the design of a broad
LO Platform‑Speci ic Model (PSM). The MSB incorporates
homogenized data model, which serves all purposes re‑
also the functionality for the orchestration of COG‑LO
quired, and at the same time (ii) the delimitation of the
components and operations as regards the execution of design to the context of the project.
work lows [21]. In addition, the MSB is the main COG‑LO
The context of COG‑LO demonstrates a dual nature,
system entity for the enforcement of mechanisms for data namely the physical and the digital one. In the physical
security, privacy and trust.
context, objects perform normal logistics actions, while
in the virtual context they act as virtualized entities with
intelligence capabilities. In the physical context, logistics
objects are parcels, containers, trucks, ships, trains,
72 © International Telecommunication Union, 2021