Page 82 - 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

Despite the enormous bene its of the aforementioned sys- relationships, querying its friends and the friends of its
tems, there are still many limitations. Most dynamic rout‑ friends in a distributed manner. This procedure guaran‑
ing algorithms are unable to analyze and distinguish the tees an ef icient and scalable discovery of CLOs and ser‑
nature of external events affecting the supply chain, ren‑ vices following the same principles that characterize the
dering therefore these systems unable to perform peri‑ social networks for humans. The following types of rela‑
odic re‑optimizations [8]. In consequence, new models tionships are indicative:
are needed that move away from the concepts on which
traditional (functional) logistics are based. • Ownership Object Relationship (OOR): created be‑
tween objects that belong to the same owner.
2.1 Digital twin – Cognitive Logistic Object • Co‑location Object Relationship (CLOR): created be‑
tween stationary devices located in the same place
The concept of Digital Twins (DT) has been introduced by
(also called Co‑Geolocation CGLOR).
M. Grieves [9] as a digital representation of a physical en‑
tity, for which a speci ic behaviour can be modelled. After • Parental Object Relationship (POR): created between
performing appropriate simulations and calculations on objects of the same model, producer and production
this behaviour an action may be triggered on that entity. batch.
In the basis of the DTs, various scenarios have been intro‑
duced at the manufacturing, logistics sector [10], health • Co‑Work Object Relationship (CWOR): created be‑
tween objects that meet each other at the owner’s
and other sectors [11]. As the DT becomes a strong tech‑
nological trend [12], with companies tending to invest on workplace (e.g., two trucks parked at a depot).
digital transformation, the transition to a DT modelling • Social Object Relationship (SOR): created as a conse‑
approach in logistics can offer new collaborative models quence of frequent meetings between objects.
and optimization procedures in the domain. The applica‑
tion of the DT paradigm in logistics has been introduced • Transactional Object Relationship (TOR): estab‑
by the concept of a Cognitive Logistics Object [13]. A Cog‑ lished between devices that interact with each other
nitive Logistics Object (CLO) is a virtualized object (simi‑ frequently [15].
lar to a DT) or system that participates in the logistics pro‑
• Time Plan Object Relationship(TPOR): created be‑
cess. It exhibits properties like autonomy, context aware‑
tween CLOs that have coincident or overlapping
ness, responsiveness and learning ability. The CLO rep‑
resents different actors such as cargo, truck, traf ic light, schedules.
supporting system, etc., with each one having different ca‑
The social graph generated by the SIoT in the COG‑LO
pabilities. In particular, a CLO is an autonomous object,
context is an undirected graph of CLOs connected with
reactive to changes in the environment and its context. It
the aforementioned relationships. It is similar to a social
is able to learn, collaborate, decide on next actions, create
graph generated by friendships between humans with
social networks and solve local problems. Thanks to the
common characteristics. The navigability problem of a
virtualization of objects, communication between hetero‑
social network has been widely addressed by Milgram,
geneous systems becomes possible, and each CLO action
and the small world phenomenon [16] has been at the
takes into account various variables such as business pri‑
centre of social science research for decades. According to
orities, environmental conditions, traf ic conditions, load
Milgram’s hypothesis, even if a social graph is very large
information etc.. Furthermore, each virtualized entity im‑ and two nodes are very distant from each other, it is possi‑
plements the functionalities required for managing the ble, starting from one, to reach the other by sur ing the net
entity’s communications. A CLO exhibits social behaviour, in less than 6 hops, thanks to the existence of short paths
which means that the digital counterpart of the logistics between pairs of nodes. The SIoT is responsible for es‑
object implements a series of services offered by SIoT tablishing relationships on the basis of local information,
to establish relationships with other CLOs, dynamically
therefore creating the necessary conditions for social nav‑
exchange information and thus optimize logistics oper‑
igation of the CLOs’ graph. To achieve this, advanced stor‑
ations. The virtualization‑based approach is quite com‑
age techniques are adopted to facilitate navigability and
mon in the IoT domain [14] as it promotes interoperabil‑
ensure that the connections between the nodes of a graph
ity and extends an object’s physical and digital character‑ on a logical level are equally accessible on the data plane
istics.
[17]. To meet this challenge, SIoT exploits the metadata
2.2 Social Internet of Things of the social nodes to ef iciently index every single data,
connection or path in the graph. SIoT can support an ML‑
The SIoT paradigm brings social network concepts to the based optimizer capable of pruning the social graph in
IoT context. According to this paradigm, each object is order to generate a smaller subgraph, which represents
characterized by a social behavior and therefore its digi‑ some elements of Milgram’s small world, where nodes
tal twin is capable of creating social relations on the basis have high correlation based on their local information.
of common elements and af inity. In the resulting social According to the principles explained above, each logis‑
network, any object looks for desired services by using its tics object is associated with a CLO, which is the digital

70 © International Telecommunication Union, 2021