ArticlePDF Available

Abstract and Figures

The maritime industry expects several improvements to efficiently manage the operation processes by introducing Industry 4.0 enabling technologies. Seaports are the most critical point in the maritime logistics chain because of its multimodal and complex nature. Consequently, coordinated communication among any seaport stakeholders is vital to improving their operations. Currently, Electronic Data Interchange (EDI) and Port Community Systems (PCS), as primary enablers of digital seaports, have demonstrated their limitations to interchange information on time, accurately, efficiently, and securely, causing high operation costs, low resource management, and low performance. For these reasons, this contribution presents the Seaport Data Space (SDS) based on the Industrial Data Space (IDS) reference architecture model to enable a secure data sharing space and promote an intelligent transport multimodal terminal. Each seaport stakeholders implements the IDS connector to take part in the SDS and share their data. On top of SDS, a Big Data architecture is integrated to manage the massive data shared in the SDS and extract useful information to improve the decision-making. The architecture has been evaluated by enabling a port authority and a container terminal to share its data with a shipping company. As a result, several Key Performance Indicators (KPIs) have been developed by using the Big Data architecture functionalities. The KPIs have been shown in a dashboard to allow easy interpretability of results for planning vessel operations. The SDS environment may improve the communication between stakeholders by reducing the transaction costs, enhancing the quality of information, and exhibiting effectiveness.
Content may be subject to copyright.
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2963283, IEEE Access
Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000.
Digital Object Identifier 10.1109/ACCESS.2017.DOI
Seaport Data Space for Improving
Logistic Maritime Operations
DAVID SARABIA-JÁCOME1, CARLOS E. PALAU1,(SENIOR MEMBER, IEEE), MANUEL
ESTEVE1, AND FERNANDO BORONAT2,(SENIOR MEMBER, IEEE)
1Communication Department, Universitat Politècnica de València, 46022, Valencia, Spain
2Communication Department, Universitat Politècnica de València, Campus Gandia, 46730, Gandia, Spain
Corresponding author: David Sarabia-Jácome (e-mail: dasaja@teleco.upv.es).
This work was supported in part by the European Union’s Horizon 2020 Research and Innovation Programme through the Pixel Port
Project under Grant 769355, and in part by the Secretaría Nacional de Educación Superior, Ciencia, Tecnología e Innovación
(SENESCYT), Ecuador.
ABSTRACT The maritime industry expects several improvements to efficiently manage the operation
processes by introducing Industry 4.0 enabling technologies. Seaports are the most critical point in
the maritime logistics chain because of its multimodal and complex nature. Consequently, coordinated
communication among any seaport stakeholders is vital to improving their operations. Currently, Electronic
Data Interchange (EDI) and Port Community Systems (PCS), as primary enablers of digital seaports, have
demonstrated their limitations to interchange information on time, accurately, efficiently, and securely,
causing high operation costs, low resource management, and low performance. For these reasons, this
contribution presents the Seaport Data Space (SDS) based on the Industrial Data Space (IDS) reference
architecture model to enable a secure data sharing space and promote an intelligent transport multimodal
terminal. Each seaport stakeholders implements the IDS connector to take part in the SDS and share their
data. On top of SDS, a Big Data architecture is integrated to manage the massive data shared in the SDS and
extract useful information to improve the decision-making. The architecture has been evaluated by enabling
a port authority and a container terminal to share its data with a shipping company. As a result, several Key
Performance Indicators (KPIs) have been developed by using the Big Data architecture functionalities. The
KPIs have been shown in a dashboard to allow easy interpretability of results for planning vessel operations.
The SDS environment may improve the communication between stakeholders by reducing the transaction
costs, enhancing the quality of information, and exhibiting effectiveness.
INDEX TERMS Analytics, Big Data, Industry 4.0, Industrial Data Spaces, Internet of Things, Maritime,
Seaport, Intelligent Transport.
I. INTRODUCTION
The rapid growth of new technologies is leading the industry
to the fourth industrial revolution, named Industry 4.0 [1].
This concept refers to the digitalization and optimization of
industrial processes through the use of emerging technology
enablers such as the Internet of Things (IoT), Cloud Comput-
ing, Big Data, or Artificial Intelligence [2] [3]. Although the
concept of Industry 4.0 has been present for some years, only
about 48% of manufacturing companies declared that they
are ready to face technological changes supported by such
building blocks [4]. The Industry 4.0 technologies adoption
gap is caused by the existing barriers encountered during the
enabling of industrial environments 4.0 [4].
The maritime industry is one of the transportation and
logistics industries with the most significant economic im-
pact on world trade. Maritime seaports support about 80%
of the world trade [5]. Each year the traffic that seaports
support increases by 1.4% [6]. Seaports must be able to adapt
to this constant growth in an efficient manner, minimizing
unproductive operations. Industry 4.0 enablers can transform
the seaports into smart seaports capable of optimizing their
processes to support the expected growth of traffic in the
coming years.
Seaports are complex intermodal terminals where several
stakeholders are involved. The synergy between them is of
vital importance for the efficient management of resources
and optimization of stakeholders’ processes. The coordinated
interaction between stakeholders might bring several bene-
VOLUME 4, 2016 1
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2963283, IEEE Access
D. Sarabia-Jácome et al.: Seaport Data Space for Improving Logistic Maritime Operations
fits, such as reliability, timeliness, safety, lower transaction
costs, and lower operational costs [7]. However, enabling
such coordinated interaction is challenging because each
of the participants involved in the distribution chain uses
heterogeneous systems. Heterogeneous data sharing is one
of the challenges with greater difficulty than the industrial
environments 4.0 have to face [3]. Currently, the stakeholders
use systems based on Electronic Data Interchange (EDI) to
exchange information under the same data format, but this
approach has shown drawbacks such as incorrect, double, and
out of time information exchange [7]. Another problem asso-
ciated with enabling coordinated communication is privacy
and security. The stakeholders are reluctant to share their data
to improve the maritime processes since the data is one of
the most critical assets [8]. Thus, seaports require a secure-
by-design environment, overcoming the limitations of cur-
rent information exchange systems to become an intermodal
intelligent transport terminal.
The Industrial Data Space (IDS) initiative emerges as a
reference architecture model to solve the problems of het-
erogeneous data sharing, considering the data sovereignty,
privacy, and traceability [9]. This model developed by the
Industrial Data Space Association (IDSA) is in the process of
being standardized by the German Institute for Standardiza-
tion (DIN). The main objective of IDS is to enable a trusted
virtual data space to support the secure exchange and linkage
of data in business ecosystems. IDS architecture has been
used in several industrial cases successfully. IDS architecture
is a novelty as it has not been implemented in the maritime
industry yet.
This work presents the Seaport Data Space (SDS) based
on the IDS architecture to solve the problem of data in-
teroperability and associated interoperation among stake-
holders in a seaport to lead to the promotion of the smart
seaport concept. SDS enables a secure virtual environment
for sharing data in a Seaport environment. Additionally, this
work presents a Big Data architecture to provide scalabil-
ity and reliability to support the massive data shared in
the SDS. The SDS was evaluated using three stakeholders:
(i) a port authority; (ii) a container terminal operator, and
(iii) a shipping company. Each stakeholder implemented an
IDS connector based on the Fiware IoT platform [10] to
interconnect to each other in the SDS. The port authority
shared data related to the vessel position in real-time, while
the container terminal operator historical load/unload berth
operations. The shipping company implemented the Big Data
architecture to manage the massive shared data and exploit
it to extract useful information. The Big Data architecture
used the flow-based system Apache NiFi [11] for pulling
data from the IDS connector and pushing them to the Big
Data modules. The Big Data modules were implemented
using Big Data open-source frameworks and systems such as
Apache Spark [12], Apache Kafka [13], and among others.
The Big Data architecture facilitated the development of
relevant Key Performance Indicators (KPIs) about vessels’
fuel consumption, time at berth and anchorage, and about
container terminal occupancy. These KPIs might be useful
to improve operational planning of a shipping company fleet,
and consequently, seaport operations.
In summary, the main contributions and novelties of the
proposed work are:
A SDS where seaport stakeholders can share and track
data to overcome the information exchanging issues
with ownership, interoperability, privacy, and security
guarantees.
A Big Data Architecture which is integrated with the
IDS architecture to handle the massive data shared and
extract useful information to improve making decisions.
Several KPIs that are extracted from the massive data
shared in SDS to improve planning operations.
The remainder of the paper is structured as follows. Sec-
tion 2 reviews the current literature concerning this field of
research. Section 3 presents the SDS architecture and the
Big Data architecture overview, as well as implementation
and the integration process details. Section 4 presents the
Big Data analytics results and KPIs by using the Big Data
Architecture in the SDS scenario. Finally, Section 5 presents
conclusions and future work.
II. RELATED WORKS AND MOTIVATION
The fundamental pillars of smart seaports are the automation
of operations and seaport equipment, and the interconnection
of the participants involved in the seaport logistics chain [14]
[15]. Cyber-physical systems (CPSs) are being used to enable
the automation of seaport equipment. These systems are able
to connect physical devices with the virtual world. Currently,
the Reference Architectural Model Industry (RAMI) 4.0 and
the Industrial Internet Reference Architecture (IIRA) leads
the implementation of the CPSs [16]. Meanwhile, IoT allows
the interconnection of any seaport equipment to the Internet.
IoT as a smart seaport enabler is being used within important
European seaports such as the Seaport of Valencia, Hamburg,
Rotterdam, among others, demonstrating its effectiveness
[17] [18] [19] [20].
Nowadays, the maritime environment employs EDI-based
systems for the information exchange between subsystems
in charge of container tracking, rail management, and in-
land navigation, and between partners in the supply chain
[21] [22]. These systems allow vertical cooperation between
stakeholders. Moreover, Port Community Systems (PCS)
have been created to reduce the complexity of the information
interchange between the stakeholders in the seaport opera-
tions [23]. The PCS are systems that centralize the vessels’
information and the goods they transport so that the stake-
holders can better control and coordinate the movements of
goods [24]. Also, the Port Collaboration for Decision Making
(PortCDM) platform proposed by the Sea Traffic Manage-
ment (STM) aims to serve as an integral point of transport
information systems to encourage cooperation among them
and allows the intelligent management of maritime traffic
[25]. The PortCDM provides information about the cargo
2VOLUME 4, 2016
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2963283, IEEE Access
D. Sarabia-Jácome et al.: Seaport Data Space for Improving Logistic Maritime Operations
FIGURE 1. Roles interaction in the IDS Architecture.
arrival and delay, and loading-unloading process in the op-
erations terminal to facilitate making decisions. The current
information exchange systems based on EDI and PCS are not
sufficient to solve problems in cooperative communication.
Mainly, these systems do not exchange information on time,
accurately, and efficiently [7]. The information exchange is
vital to improve the quality of transport.
Unlike the information interchange model, data sharing
involves both vertical and horizontal collaboration between
companies. A data market might be created to encourage
collaboration among competitors to achieve a common goal
using the data-sharing approach. The IDS architecture pro-
poses a secure environment to ease the data sharing between
companies involved in the production and distribution of a
product [9]. The IDS reference architecture was developed to
meet the industrial needs of trust, security, data sovereignty,
data ecosystem, standardized interoperability, value-added
applications, and a data market. The IDS reference archi-
tecture is composed of five layers: business, functional, in-
formation, process, and system layer. The business layer
defines the specific roles to enable data exchange, Fig.1. The
functional layer describes the characteristics of trust, security,
data ecosystem, interoperability, value-added applications,
and data market. The process layer specifies the interactions
between the components of the architecture. These compo-
nents are grouped into sub-processes that are responsible
for accessing the data space, exchanging data, publishing,
and using applications. The information layer specifies the
information model to facilitate compatibility and interoper-
ability. The system layer describes the specific roles of the
business layer to cover the functional requirements. This
layer defines the connector, the broker, the identity provider,
and the application store [26].
Since the IDS architecture does not define the interfaces
to be used nor provide details for implementation, the inter-
action between the academia and industry provides relevant
information through implementation cases. For example,
some research has provided relevant information about com-
ponents implementation [27], security implementation [28],
and the ontology-based information model [29]. On the other
hand, the industry has implemented the IDS architecture
successfully in logistics cases: to optimize the loading and
unloading times of trucks [30] and to predict railway-tracks
maintenance [31]. In the case of the maritime industrial
sector, the SINTEF Ocean institute analyzed the use of the
IDS architecture to support the Maritime Data Space (MDS)
[32]. They stated that the obstacles to enable the MDS
are the connection of vessels with the IDS, the shipping
system complexity, and the international nature of shipping.
Even though IDSA encourages the industry to use the IDS
architecture in industrial environments, the implementation
of the IDS architecture model in the Seaport case has not
been implemented yet.
Recently, the Boost 4.0 project was released to design and
implement Big Data middleware for IDS support. The main
project motivation is to fill the gap between IDS architecture
and Big Data management [33]. The project is planning
to publish its results by the end of 2021. Also, few Big
Data architectures were proposed in the current literature
for the maritime industry [34] [35]. The primary approach
used by these architectures was to employ the Lambda
processing architecture, which has proven to be efficient
in meeting the requirements of scalability, efficiency, and
high availability [36]. However, these architectures did not
consider IoT requirements for Big Data management or the
use of the Big Data life cycle model for their designs. The
International Telecommunication Union (ITU) has released a
bunch of recommendations (Y. 2066 [37] and Y4114 [38]) to
be considered in the design of the Big Data architecture for
IoT. Also, the Big Data life-cycle model (BDLM) proposed
by Demchenko et al. [39] provides essential advantages to
the data re-usability at any life cycle stage and the massive
reduction of the data at an initial stage. Big Data architecture
is fundamental to extract relevant information from shared
data to improve seaport operations.
There are several models, considered as state of the art,
used to estimate some vessel operations process. For exam-
ple, these models are intended to estimate fuel consumption
and pollution generated by vessels [5] [40] [41]. However,
the problem appears when the models are applied to large
datasets without the support of adequate processing infras-
tructure. The models need to be adapted so that they can be
efficiently exploited by the resources used by the Big Data
architecture [42]. The lack of know-how to implement these
algorithms in a Big Data architecture is limiting the efficient
exploitation of the shared data to improve the operation in the
seaport.
Unlike related works, the main motivation of this work is
facilitating the coordinated communication between stake-
holders in a multimodal seaport terminal thought the IDS
reference architecture. Also, this work fills the gap between
IDS reference architecture and Big Data management by
providing a Big Data architecture implementation details
and know-how. Finally, this work proposes vessel operations
algorithms based on Big Data techniques to improve seaport
operations.
VOLUME 4, 2016 3
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2963283, IEEE Access
D. Sarabia-Jácome et al.: Seaport Data Space for Improving Logistic Maritime Operations
FIGURE 2. Valencia SDS case overview.
III. VALENCIA SEAPORT DATA SPACE
This section presents the design and implementation of the
SDS, the Big Data architecture implementation details, and
the Big Data architecture and SDS integration. The proposal
is applied to the Valencia-Spain Seaport.
A. REQUIREMENTS
Valencia seaport is considered one of the most important
ports on the Mediterranean coast. This seaport supports more
than 4.7 million Twenty-foot Equivalent Units (TEUs) per
year [43]. Recently, traffic has shown a growth of 1.77
million TEUs, which has affected the seaport operations effi-
ciency [15]. Valencia seaport requires strategies that would
allow it to optimize seaport operations and exploit its re-
sources efficiently.
The transformation of Valencia seaport into a smart seaport
requires solving the problems that appear in the data sharing
process between stakeholders to improve decision making.
The stakeholders involved in the maritime port’s opera-
tions are the port authority, terminal operators, shipping com-
panies, truck companies, railway operators, seaport equip-
ment maintenance companies, and cold container mainte-
nance companies. This work studies the integration of the
port authority, a container terminal operator, and a shipping
company to demonstrate the feasibility of the SDS. Fig. 2
shows an overview of the Valencia Seaport case.
The container terminal operator is responsible for loading
and unloading containers from vessels, trains, and trucks. In
this case, the container terminal operator has a system based
on Structured Query Language (SQL) to record the opera-
tions of loading and unloading performances. The shared data
are structured using the JavaScript Object Notation (JSON)
data format.
The port authority controls the arrivals and departures of
vessels, trains, and trucks to and from the seaport. In this
case, the port authority keeps track of the vessels that are near
the port through the Automatic Identification System (AIS).
The AIS provides information about the vessel’s navigation
state.
FIGURE 3. SDS architecture components.
B. SDS ARCHITECTURE OVERVIEW
The main objective of this case is to design a secure Big
Data sharing environment among seaport stakeholders based
on the IDS reference architecture. The SDS architecture
overview presents the details about the SDS architecture
components, systems adapters, data models, and the sharing
process. Fig. 3 shows the high-level SDS architecture.
1) SDS Architecture Components
The SDS architecture is composed of IDS connectors, an
identity provider, and an IDS broker.
The IDS connectors share data, ensure data sovereignty,
and keep the interoperability between systems [9]. This con-
nector uses a publish/subscribe mechanism to share data,
a proxy Policy Enforcement Point (PEP) to ensure data
sovereignty, and an information model to keep the same
data model and format. The IDS connector functionalities
are implemented using the Fiware IoT platform Generic
Enablers (GEs) Orion Context Broker, and Wilma. Each
stakeholder implements an IDS connector in their techno-
logical infrastructure to connect to the SDS and share data.
Fig.4 shows the structure of the IDS connector for this case.
The GE Orion Context Broker provides a publish/subscribe
mechanism to receive entities context updates. To do so, the
Orion Context Broker uses the NGSI9 and NGSI10 interfaces
to send information about the context data and send context
data. On the other hand, the GE Wilma provides PEP-Proxy
functionalities to keep control of the data.
The identity provider keeps the information about the IDS
connectors of the SDS and validates the connectors’ identity
[9]. These functionalities are implemented using the GE
Identity Manager-Keyrock (IdM). The IdM uses the Oauth2
protocol to allow connectors authentication. Also, the IdM
keeps a record of its service through logs.
The IDS broker keeps the information about the data
sources, data models, and usage policies [9]. These func-
tionalities are implemented using the GE Policy Decision
Point/Policy Administration Point (PDP/PAP) AuthZForce
GE. AuthZForce uses the eXtensible Access Control Markup
Language (XACML) to allow the definition of fine-grained
policies.
4VOLUME 4, 2016
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2963283, IEEE Access
D. Sarabia-Jácome et al.: Seaport Data Space for Improving Logistic Maritime Operations
FIGURE 4. IDS Connector.
FIGURE 5. JSON schema of vesselObserved entity.
2) Data Models
Data models abstract the stakeholders’ systems to organize
the data into the Fiware IoT platform. Vessels are modeled
through the entity vesselObserved in the case of the AIS
system. This entity represents a vessel with its characteris-
tics. These characteristics are the maritime mobile service
identity (MMSI), position (latitude and longitude), course
over ground (COG), speed over ground (SOG), rate of turn
(ROT), observation date, and operation mode. The identity
created in the Fiware based IDS connector is updated with
every AIS vessel message. Fig. 5 shows a vesselObserved
JSON example.
Meanwhile, the seaport berths are modeled through the
entity berth in the case of the container terminal operating
system. this entity represents a container terminal berth.
These characteristics are the initial and final operational
dates.
3) Systems Adapters
The systems adapters interconnected the stakeholders’ Ap-
plication Programming Interfaces (APIs). This connection is
developed using the Node-RED platform [44]. This platform
facilitates the development of data flows through its web user
FIGURE 6. AIS system adapter Node-RED flow.
interface. The flows use nodes that are capable of making
data transformations.
The AIS system flow queries the AIS HTTP server, con-
verts the AIS message format (NMEA standard) to JSON
format, assemblies the JSON based on the data model entity,
adds the Fiware JSON headers and sends the data to the IDS
connector. Fig. 6 shows the AIS system Node-RED flow.
4) Sharing Process
The process of data sharing between Fiware-based IDS con-
nectors exploits the federation functionality of the Orion
Context Broker. The federation push mode allows the send-
ing of context notifications between two Orion Context Bro-
kers. After enabling the Orion Context Broker federation,
the sharing process requires a subscription notification that
contains the entities’ id and the uniform resource locator
(URL) of the other IDS connector that is going to receive
the entities’ data.
C. BIG DATA ARCHITECTURE OVERVIEW
The architecture is based on the Lambda processing architec-
ture [36] and the BDLM [39]. Also, the Big Data architecture
design considers the ITU recommendations Y. 2066 [37]
and Y4114 [38]. The Big Data architecture is composed of
several modules that can be adapted depending on the needs
of each SDS member. The different Big Data architecture
modules are implemented using open source platforms for
Big Data management. In the case studied, the Big Data
architecture is implemented in the shipping company tech-
nological infrastructure to exploit the data, Fig.7. Next, the
functionalities, technologies, and platforms used for the im-
plementation are described:
Integration module: is in charge of facilitating the
connection between the IDS connector and the Big
Data architecture. This module exploits the pull/push
mechanisms employed by the IDS connector to collect
the data. This module is implemented using the flow
automation system, Apache NiFi [11]. The selection
of Apache NiFi responds to its ability to design flows
visually through its user interface and as a highly scal-
able, configurable, and secure tool. Apache NiFi pro-
vides several processors capable of performing specific
operations over the data flow. The primary operations
to be carried out in this module are the connection
with the IDS connector and the conversion of the data
format JSON to the Parquet format. The Parquet format
is a high-performance format [45]. The historical data
repository receives the data resulting from the convert-
VOLUME 4, 2016 5
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2963283, IEEE Access
D. Sarabia-Jácome et al.: Seaport Data Space for Improving Logistic Maritime Operations
FIGURE 7. Big Data Architecture Implementation.
ing format tasks. In the case of real-time data, they are
sent to the data processing platform in real-time through
Apache Kafka [13].
Historical Database: stores all the data that come from
the IDS connector. This module is implemented using
the Hadoop Distributed File System (HDFS). HDFS
is widely used in the current literature in Big Data
ecosystems because it provides high availability, relia-
bility, and scalability [46]. The HDFS cluster consists
of a Name Node and several Data Nodes. The Data
Nodes are intended for data storage in 64 megabytes
data blocks, while the Name Node manages the location
of the files and their replicas in the Data Nodes.
Processing module: provides batch and real-time data
processing. The data processing modules are imple-
mented using the Apache Spark framework [12]. The
selection of Apache Spark responds to its ability to ex-
ecute jobs both in batch mode and in real-time through
its APIs. Apache Spark temporarily stores the results of
its operations in memory, so it has shown better perfor-
mance than Hadoop, which stores its operations on disk
[47]. The Spark Streaming API processes real-time data.
This API allows the execution of applications almost in
real-time. The applications are focused on performing
operations over data in a time sliding window.
Batch Repository: stores the batch processing re-
sults. This repository is implemented using the NoSQL
Apache HBase database system [48]. This database
system provides scalability and high availability.
Real-time Repository: is in charge of storing the real-
time processing results. This module is implemented
using the Postgres relational database system and its
PostGIS extension [49]. PostGIS is an efficient system
to store geospatial data. The selection of PostGIS re-
sponds to the data reduction in the initial processing
stages, which reduces the scaling problems overtime.
Query Manager module: manages queries that are
used to generate descriptive analytics. Mainly, the mod-
ule exploits the platforms’ API and functions to generate
data queries. This module is implemented using the
tools provided by the GeoMesa framework and the
SparkSQL API [50]. GeoMesa allows the analysis of
geospatial data through a set of tools that are integrated
with processing frameworks such as Apache Spark and
with a database system like HBase. GeoMesa provides
a Spatio-temporal indexation to store data of point, line,
polygon type in HBase. While SparkSQL allows struc-
turing data in DataFrames for analysis using a language
similar to SQL. Also, SparkSQL presents functionalities
to perform a descriptive analysis of data (descriptive
statistics) and to perform a data cleaning, aggregations,
and filtering. These tools are used to extract useful
information from the data.
Data Visualization module: displays the results of Big
Data Analytics to users. This module implements a
graphical user interface (GUI) for the deployment of
KPIs and diagrams. The GUI goal is to help opera-
tors infer the information extracted from the data. The
GUI is a web application implemented using Bootstrap,
NodeJS, Socketio, ChartJS, and Leaflet. The web appli-
cation uses a backend and frontend structure to present
the information to the user efficiently.
D. BIG DATA AND IDS ARCHITECTURE INTEGRATION
The integration of Big Data architecture and IDS architecture
is developed by implementing a dataflow in Apache NiFi
[11]. This dataflow is in charge of the data extraction from
the IDS connector, the data transformation, and loading
data to Big Data platforms. The dataflow is composed of
4 processors: ListenHTTP, PutParquet, PublishKafka, and
6VOLUME 4, 2016
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2963283, IEEE Access
D. Sarabia-Jácome et al.: Seaport Data Space for Improving Logistic Maritime Operations
FIGURE 8. Apache NiFi DataFlow for integration.
LogAttribute. Fig. 8 shows the dataflow and the processors
used for its implementation.
The ListenHTTP processor is in charge of receiving the
Context Notification from the IDS connector. The Listen-
HTTP processor implements an HTTP server. This server
listens for POST requests on a specific port. The port con-
figured is port 8080 in this case. The IDS connector sends
POST requests with context notifications in JSON format to
the HTTP server. The ListenHTTP processor redirects these
notifications to the PutParquet and PublishKafka processors.
The PutParquet processor is in charge of receiving the
Context Notification from the ListenHTTP processor, con-
verting to Parquet format, and storing it into the HDFS
cluster. The PutParquet processor requires the data schema
in Avro format to translate successfully to Parquet. The Avro
format is a JSON format that describes the data types and
protocols used in the definition of the data model. Also, the
PutParquet processor requires the information of the HDFS
cluster to store the converted data. The processor needs
access to the configuration files of the HDFS cluster (core-
site.xml and hdfs-site.xml) to know what the Node Name
and Data Nodes servers are, and the configuration of the
replication blocks. Another essential configuration parameter
is the file tree path, where the data are loaded. Finally, the
PutParquet processor loads the converted data into the HDFS
file tree path.
The PublishKafka processor is in charge of receiving
the Context Notification from the ListenHTTP processor
and publishing it into the Kafka broker. The PublishKafka
processor requires the Kafka broker URL and the topic as
configuration parameters. The Apache Spark cluster receives
the data by subscribing to the same topic to processing on-
the-fly.
Finally, the LogAttribute processor completes the
dataflow. This processor allows registering the status of
each transaction of both the PutParquet processor and Pub-
lishKafka. This process facilitates the easy identification of
errors that may occur during the dataflow operation.
IV. RESULTS
The results show the SDS feasibility and the use of Big
Data architecture to extract useful information in planning
the shipping company operations. For this, two datasets were
used in the experimental evaluation. Table 1 describes the
datasets used in this experiment. The data shared between the
IDS connectors of the SDS are exploited to extract relevant
KPIs for the planning of a shipping company operations.
The KPIs are about the vessel’s average occupation time in
the containers loading and unloading process in the seaport
container terminal, the container terminal occupancy, the
waiting time of the shipping company vessels in the seaport
anchorage zone and the vessel’s fuel consumption estimation
during its waiting time in the seaport anchorage zone. Several
applications were developed in Apache Spark to load, pro-
cess, and analyze the shared data for generating these KPIs.
TABLE 1. Seaport Datasets details
Dataset Size Period
AIS 10 GB 2016/01/01 - 2016/03/31
Terminal Operation 520 MB 2014/01/01 - 2019/01/31
A. VESSELS AVERAGE TIME OCCUPANCY
The vessels’ occupation time in the container terminal is
calculated using the container terminal operations dataset.
The container terminal IDS connector shares the data about
berths load/unload processes to the shipping company IDS
connector. Subsequently, the HDFS repository stores the data
shared in the IDS environment by following the data flow
defined in the Big Data and IDS architecture integration
subsection.
The loaded data from HDFS is structured in a DataFrame
using a SparkSQL function. The features related to the berth’
id, unloading process start timestamp, and loading process
finish timestamp are selected from the first DataFrame to
calculate the occupation time. Vessels occupancy on berths
is performed by the difference between the unloading and
loading timestamps. Next, the application calculates the av-
erage time at berth throughout an aggregation function by
evaluating the new DataFrame in a week as frequency.
The vessels’ maximum and minimum time occupancy in
berth provides more information to make decisions. Since the
dataset is a time-series data, the time-series decomposition
into season and trend components is necessary to assess
whether the maximum and minimum values vary over time.
For this, the DataFrame obtained in the previous phase is
decomposed using the Python Statsmodels library. Fig. 9
shows the decomposition into components of the DataFrame.
The figure shows an incremental trend in time and a repeated
season pattern every two months. As a result, vessels’ av-
erage time is not the same in short periods (season), and
it varies over time (trend). In the same way, the maximum
and minimum values also vary over time, so the maximum
and minimum calculation is made using the maximum and
minimum average values per week. The average, maximum,
VOLUME 4, 2016 7
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2963283, IEEE Access
D. Sarabia-Jácome et al.: Seaport Data Space for Improving Logistic Maritime Operations
FIGURE 9. Terminal containers time-series decomposition.
and minimum occupation time values are used as KPIs to
estimate the time that the shipping company fleet is going to
be berthing in the seaport.
B. CONTAINER TERMINAL OCCUPANCY WEEKLY
The container terminal occupancy gives information about
how many vessels the seaport terminal can support during
the week. Similar to previous KPI, this is calculated using
the dataset of the container terminal operations shared from
the container terminal IDS to the shipping company IDS.
Unlike the previous KPI, the features related to the vessel’
MMSI and the starting timestamp are selected from the first
DataFrame to calculate berth’s occupancy. In this case, the
application aggregates the data by days, weeks, and months,
and applies the count function on them. In this way, the
application extracts information about the number of vessels
per day, week, and month that are berthing in the container
terminal. The result is summarized in a box-whisker diagram
to represent the days and months most busy. Fig. 10 shows
that Saturday and August are the day of the week and the
month of the year most busy.
The information obtained allows the shipping company to
plan its operations in the days and months with less working
load so that its vessels stay the least amount of time possible
in the seaport.
C. AVERAGE TIME WAITING FOR A FREE TERMINAL
The vessels’ anchorage time reduces the shipping company
efficiency and produces a higher operational cost. The ap-
plication estimates this time by using the AIS dataset. The
port authority IDS connector shares AIS data (only messages
from the company’s fleet) to the shipping company IDS
connector. Next, the shared data are stored in the HDFS
repository and sent to the Apache Spark platform following
the data flow defined in the Big Data and IDS architecture in-
tegration subsection. Fig. 11 shows the application data flow
developed using Apache Spark Streaming and SparkSQL.
In this case, the application executes the SparkSQL func-
tionalities on the real-time data through the Apache Spark
FIGURE 10. Box-whisker container terminal occupancy by day and month.
Structures Streaming engine. This engine takes advantage
of micro-batch processing to execute highly scalable, fast,
and fault-tolerant queries. The AIS messages are filtered in
real-time to select only the AIS messages whose vessels’
positions are near the seaport. The application applies a filter
based on a polygon of 4 geospatial points (latitude, longi-
tude). Next, the filtered data get onto a 60-minutes-sliding
window. This data is structured in a DataFrame by using the
Structured Streaming engine. Then, the application splits the
DataFrame into two DataFrames based on vessels’ operation
mode (anchorage and not anchorage). The new DataFrame
contains the vessels’ MMSI identification and timestamp.
Next, the application searches the vessels that have changed
the operation mode from the anchorage state. The application
employs the MMSI identifier in the searching process. The
join DataFrame function allows making a comparison of the
DataFrames and selects the rows where the MMSI is the
same in both DataFrames. Next, the application creates a
new DataFrame with the selected rows. This new DataFrame
contains the MMSI and the start and finish timestamp of
the anchorage mode. The application updates the anchorage
DataFrame by deleting the row with the MMSI founded in the
searching process. Finally, the application calculates the time
at anchor by mean of the difference between the timestamps
(finish-start) and loads the results into a table in the PostGIS
database. The web application queries the table and shows
the average, the minimum, and maximum time at anchor per
vessel as KPI. In this way, the shipping company can order to
their vessels to reduce the speed before arriving at the seaport
to save fuel.
8VOLUME 4, 2016
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2963283, IEEE Access
D. Sarabia-Jácome et al.: Seaport Data Space for Improving Logistic Maritime Operations
FIGURE 11. The algorithm used for calculating vessels berthing time.
FIGURE 12. The algorithm used for estimating vessel fuel consumption.
D. VESSELS AVERAGE FUEL CONSUMPTION IN THE
WAITING PERIOD
The fuel consumption estimation provides information about
the tons of fuel occupied by the vessels when they are waiting
for a berth. As the previous KPI, this is calculated using the
AIS message dataset shared from the port authority IDS to
the shipping company IDS. Fig. 12 shows the algorithm used
based on the study presented in [40] and [5].
The fuel consumption estimation depends directly on the
power used during the period evaluated. Vessels’ power infor-
mation is not available on the AIS messages, so it is estimated
based on the motor’s load factor during anchorage operating
mode. The literature estimates that the load factor is 5% for
the main engine (ME) and 50% for the auxiliary engine (AE)
[41]. The total power is the result of the sum ME and AE
powers during the anchorage operation mode. These results
are stored on a table inside the HBase database. Next, the
application multiplies the total power by the waiting time at
the anchorage zone (previously calculated in the subsection
C) by the base-specific consumption fuel and by a factor of
1.1, according to [40]. The base-specific fuel consumption
has been estimated at 195g/kWh for vessels built since 2001
and at 205 g/kWh for vessels built between 1984 and 2000
[5]. Finally, the application stores the results in a table
in the Postgres database. The web application queries the
table and shows an average, minimum, and maximum fuel
consumption during the vessels’ waiting time at the seaport
anchorage zone.
The web GUI dashboard groups the KPIs developed for
a straightforward user interpretation. Fig. 13 shows the GUI
with the developed KPIs. Also, the GUI has a map to show
the vessel’s position near the Valencia seaport.
V. CONCLUSION
In this paper, we have proposed the use of the IDS reference
architecture to overcome the current limitations on seaport
systems data sharing and facilitate the cooperative commu-
nication interoperability and interaction between stakehold-
ers. Data sovereignty is the main advantage of using IDS
architecture in industry 4.0, and specifically in transport and
logistics. Since the IDS architecture does not consider the
Big Data management, it has been proposed the integration
of a Big Data architecture in our SDS environment proposal.
The integration module facilitates the connection to the IDS
connector to extract, clean, and load data into the Big Data
platforms. The integration module functionalities were im-
plemented using Apache NiFi. Apache NiFi proved to be
useful on the integration due to its high capacity for designing
data flows. The rest of the Big Data architecture modules
provide features to store, process, and analyze the data shared
in IDS. These functionalities were implemented using current
Big Data open-source platforms and frameworks such as
Hadoop, HDFS, Apache Spark, Apache Kafka, and HBase.
The chosen platforms guaranteed efficient management of
the data shared in the IDS architecture and provided a scal-
able and high available environment to develop and execute
applications for massive data processing.
The feasibility of our proposal was validated by using
several datasets related to vessel positions and terminal
operations. The use of the Big Data architecture in SDS
allowed the extraction of valuable information for the opera-
tions planning of a shipping company. The information was
transformed into KPIs for a better interpretation of the data
analysis results. The KPIs were compiled on a dashboard to
improve the decision-making. Also, we adapted some state
of art vessel operation models to be used in the Big Data
Architecture.
The SDS improves the coordination between stakeholders
by lower transaction costs. Also, the SDS allows the re-
use of information by multiple parties and improving the
VOLUME 4, 2016 9
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2963283, IEEE Access
D. Sarabia-Jácome et al.: Seaport Data Space for Improving Logistic Maritime Operations
FIGURE 13. Shipping Company Web GUI dashboard.
quality of information. Finally, the data shared through SDS
in time enhanced the vessel transit time and saved cost in
the seaport operations. Although this paper has evaluated the
proposed architecture in the maritime application domain, it
is extensible and flexible to any industrial sector.
Moreover, the proposed Big Data architecture covers the
IoT requirements proposed by the ITU-T so that it can be ex-
tended to application domains and cases involving industrial
IoT devices. As future work, there will be further testing of
the Big Data architecture in other application domains and
cases to demonstrate its extensibility and adaptability. Also,
more stakeholders and their dataset will be added to the SDS.
REFERENCES
[1] A. V. Bogoviz, “Industry 4.0 as a new vector of growth and development
of knowledge economy,” in Industry 4.0: Industrial Revolution of the
21st Century. Springer International Publishing, jul 2018, pp. 85–91.
[Online]. Available: https://doi.org/10.1007%2F978-3-319-94310-7_8
[2] M. Hermann, T. Pentek, and B. Otto, “Design principles for industrie
4.0 scenarios,” in 2016 49th Hawaii International Conference on
System Sciences (HICSS). IEEE, jan 2016. [Online]. Available:
https://doi.org/10.1109%2Fhicss.2016.488
[3] Y. Lu, “Industry 4.0: A survey on technologies, applications and open
research issues,” Journal of Industrial Information Integration, vol. 6, pp.
1–10, jun 2017. [Online]. Available: https://doi.org/10.1016%2Fj.jii.2017.
04.005
[4] C. Baur and D. Wee, “Manufacturing’s nextact,” McKinsey Quarterly, Jun,
2015.
[5] Ø. Buhaug, J. J. Corbett, Ø. Endresen, V. Eyring, J. Faber, S. Hanayama,
D. S. Lee, D. Lee, H. Lindstad, A. Markowska et al., “Second imo ghg
study 2009,” International Maritime Organization (IMO) London, UK,
vol. 20, 2009.
[6] “Publications - eurostats: Maritime transport of goods quarterly data.”
[Online]. Available: http://ec.europa.eu/eurostat/statistics-explained/
index.php
[7] A. Gharehgozli, H. de Vries, and S. Decrauw, “The role of standardisation
in european intermodal transportation,” Maritime Business Review,
vol. 4, no. 2, pp. 151–168, jun 2019. [Online]. Available: https:
//doi.org/10.1108%2Fmabr-09-2018- 0038
[8] T. K. Sung, “Industry 4.0: a korea perspective,” Technological forecasting
and social change, vol. 132, pp. 40–45, 2018.
[9] “International data spaces association.” [Online]. Available: https:
//www.internationaldataspaces.org/
[10] FI-WARE Consortium, “FIWARE: THE OPEN SOURCE PLATFORM
FOR OUR SMART DIGITAL FUTURE.” [Online]. Available: https:
//www.fiware.org/
[11] “Apache NiFi.” [Online]. Available: https://nifi.apache.org/
[12] “Apache Spark - Lightning-Fast Cluster Computing.” [Online]. Available:
http://spark.apache.org/
[13] “Apache kafka. a distributed streaming platform.” [Online]. Available:
https://kafka.apache.org/
[14] C. Liu, H. Jula, K. Vukadinovic, and P. Ioannou, “Comparing
different technologies for containers movement in marine container
terminals,” in ITSC2000. 2000 IEEE Intelligent Transportation Systems.
Proceedings (Cat. No.00TH8493). IEEE. [Online]. Available: https:
//doi.org/10.1109%2Fitsc.2000.881118
[15] D. Sarabia-Jacome, I. Lacalle, C. E. Palau, and M. Esteve, “Enabling
industrial data space architecture for seaport scenario,” in 2019 IEEE 5th
World Forum on Internet of Things (WF-IoT). IEEE, apr 2019, pp. 101–
106.
[16] N. Jesse, “Internet of things and big data – the disruption of
the value chain and the rise of new software ecosystems,” IFAC-
PapersOnLine, vol. 49, no. 29, pp. 275–282, 2016. [Online]. Available:
https://doi.org/10.1016%2Fj.ifacol.2016.11.079
[17] I. Schirmer, P. Drews, S. Saxe, U. Baldauf, and J. Tesse,
“Extending Enterprise Architectures for Adopting the Internet
of Things – Lessons Learned from the smartPORT Projects
in Hamburg,” in Business Information Systems. Springer
International Publishing, 2016, pp. 169–180. [Online]. Available:
https://doi.org/10.1007%2F978-3-319-39426-8_14
[18] P. Fernández, J. Santana, S. Ortega, A. Trujillo, J. Suárez, C. Domínguez,
J. Santana, and A. Sánchez, “SmartPort: A platform for sensor data
monitoring in a seaport based on FIWARE,” Sensors, vol. 16, no. 3, p. 417,
mar 2016. [Online]. Available: https://doi.org/10.3390%2Fs16030417
[19] A. Belfkih, C. Duvallet, and B. Sadeg, “The Internet of Things for Smart
Ports: Application to the Port of Le Havre,” in International Conference on
Intelligent Platform for Smart Port (IPaSPort 2017), 2017.
[20] A. Belsa, D. Sarabia-Jacome, C. E. Palau, and M. Esteve, “Flow-based
programming interoperability solution for IoT platform applications,” in
10 VOLUME 4, 2016
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2963283, IEEE Access
D. Sarabia-Jácome et al.: Seaport Data Space for Improving Logistic Maritime Operations
2018 IEEE International Conference on Cloud Engineering (IC2E), IEEE.
IEEE, 2018, pp. 304–309.
[21] L. Heilig and S. Voß, “Information systems in seaports: a categorization
and overview,” Information Technology and Management, vol. 18, no. 3,
pp. 179–201, nov 2016. [Online]. Available: https://doi.org/10.1007%
2Fs10799-016-0269-1
[22] J. Muñuzuri, L. Onieva, P. Cortés, and J. Guadix, “Using IoT data and
applications to improve port-based intermodal supply chains,” Computers
& Industrial Engineering, p. 105668, jan 2019. [Online]. Available:
https://doi.org/10.1016%2Fj.cie.2019.01.042
[23] V. Carlan, C. Sys, and T. Vanelslander, “How port community systems can
contribute to port competitiveness: Developing a cost–benefit framework,”
Research in Transportation Business & Management, vol. 19, pp. 51–64,
2016.
[24] M.-L. Baron and H. Mathieu, “PCS interoperability in Europe: a market
for PCS operators?” The International Journal of Logistics Management,
vol. 24, no. 1, pp. 117–129, may 2013.
[25] M. Lind, T. Andersen, M. Bergmann, R. T. Watson, S. Haraldson,
M. Karlsson, M. Michaelides, J. Gimenez, R. Ward, N. Bjørn-Andersen
et al., “The maturity level framework for PortCDM,” 2018.
[26] B. Otto, M. ten Hompel, and S. Wrobel, “International data spaces,” in
Digital Transformation. Springer, 2019, pp. 109–128.
[27] Á. Alonso, A. Pozo, J. Cantera, F. de la Vega, and J. Hierro, “Industrial
data space architecture implementation using FIWARE,” Sensors, vol. 18,
no. 7, p. 2226, 2018.
[28] G. S. Brost, M. Huber, M. Weiß, M. Protsenko, J. Schütte, and S. Wessel,
“An ecosystem and IoT device architecture for building trust in the
industrial data space,” in Proceedings of the 4th ACM Workshop on Cyber-
Physical System Security - CPSS '18. ACM Press, 2018, pp. 39–50.
[29] J. Pullmann, N. Petersen, C. Mader, S. Lohmann, and Z. Kemeny,
“Ontology-based information modelling in the industrial data space,” in
2017 22nd IEEE International Conference on Emerging Technologies and
Factory Automation (ETFA). IEEE, sep 2017, pp. 1–8.
[30] “International data spaces association.” [On-
line]. Available: https://www.internationaldataspaces.org/
thyssenkrupp-implements-first-use- case-for- industrial-data- space
[31] “International data spaces association.” [Online]. Available: https:
//www.internationaldataspaces.org/success-stories/{\#}advaneo/
[32] Ø. J. Rødseth and A. J. Berre, “From digital twin to maritime data space:
Transparent ownership and use of ship information.
[33] “Boost40.” [Online]. Available: https://boost40.eu/
[34] G. Vouros, C. Doulkeridis, G. Santipantakis, A. Vlachou, N. Pelekis,
H. Georgiou, Y. Theodoridis, K. Patroumpas, E. Alevizos, A. Artikis
et al., “Big data analytics for time critical maritime and aerial mobility
forecasting,” Advances in Database Technology-EDBT, vol. 2018, pp.
612–623, 2018.
[35] H. Wang, O. L. Osen, G. Li, W. Li, H.-N. Dai, and W. Zeng, “Big data
and industrial Internet of things for the maritime industry in northwestern
Norway,” in TENCON 2015-2015 IEEE Region 10 Conference. IEEE,
2015, pp. 1–5.
[36] N. Marz and J. Warren, Big data: principles and best practices of scalable
realtime data systems. Shelter Island, NY : Manning, cop. 2015, 2015.
[Online]. Available: http://nathanmarz.com/about/
[37] ITU-T, “Next-Generation Networks - Common requirements of the
Internet of things. T-REC-Y.2066,” 2014. [Online]. Available: https:
//www.itu.int/rec/T- REC-Y.2066-201406-I
[38] ITU-T, “Internet of things and smart cities - Specific requirements and
capabilities of the Internet of things for big data. T-REC-Y.4114,” 2017.
[Online]. Available: https://www.itu.int/rec/T- REC-Y.4114/en
[39] Y. Demchenko, C. de Laat, and P. Membrey, “Defining architecture com-
ponents of the big data ecosystem,” in 2014 International Conference on
Collaboration Technologies and Systems (CTS), IEEE. IEEE, may 2014,
pp. 104–112.
[40] L. Schrooten, I. De Vlieger, L. I. Panis, K. Styns, and R. Torfs, “Inventory
and forecasting of maritime emissions in the Belgian sea territory, an
activity-based emission model,” Atmospheric Environment, vol. 42, no. 4,
pp. 667–676, 2008.
[41] P. D. Meyer, F. Maes, and A. Volckaert, “Emissions from international
shipping in the Belgian part of the North Sea and the Belgian seaports,”
Atmospheric Environment, vol. 42, no. 1, pp. 196–206, jan 2008.
[Online]. Available: https://doi.org/10.1016%2Fj.atmosenv.2007.06.059
[42] X. Sun, Z. Tian, R. Malekian, and Z. Li, “Estimation of vessel emissions
inventory in Qingdao Port based on big data analysis,” Symmetry, vol. 10,
no. 10, p. 452, 2018.
[43] “Valencia port report 2016.” [Online]. Avail-
able: https://www.valenciaport.com/wp-content/uploads/
Boletin-EstadisticoDiciembre-2016.pdf
[44] “Node-red low-code programming for event-driven applications.
[Online]. Available: https://nodered.org/
[45] J. Kestelyn, “Benchmarking Apache Parquet: The Allstate Experience,”
2016. [Online]. Available: https://blog.cloudera.com/blog/2016/04/
benchmarking-apache-parquet-the- allstate-experience/
[46] “Hdfs architecture guide.” [Online]. Available: https://hadoop.apache.org/
docs/r1.2.1/hdfs{_}design.html
[47] M. Zaharia, M. Chowdhury, M. J. Franklin, S. Shenker, and I. Stoica,
“Spark: Cluster computing with working sets.” HotCloud, 2010.
[48] “Apache hbase.” [Online]. Available: https://hbase.apache.org/
[49] “Postgis spatial database.” [Online]. Available: https://postgis.net/
[50] “Geomesa.” [Online]. Available: https://www.geomesa.org/
DAVID SARABIA-JÁCOME received the M.Sc.
degree in communications technologies, systems
and networks from the Universitat Politécnica de
Valéncia,Spain,in 2016,where he is currently pur-
suing the Ph.D. degree with the Escuela Técnica
Superior de Ingenieros de Telecomunicación. His
research activities and interests cover Internet of
Things,Big Data,Cloud Computing,Fog Comput-
ing, and virtualization.
CARLOS E. PALAU (SM’17) received the M.Sc.
and Ph.D. (Dr.-Ing.) degrees in telecommunication
engineering from the Universitat Politècnica de
València in 1993 and 1997, respectively. He is
currently a Full Professor with the Escuela Téc-
nica Superior de Ingenieros de Telecomunicación,
Universitat Politècnica de València. He has over
20 years of experience in the ICT research area
in the field of networking. He has collaborated
extensively in the research and development of
multimedia streaming, security, networking, and wireless communications
for government agencies, defense, and European Commission as a Main
Researcher of EU-FP6, EU-FP7, and EU-H2020 Programs. He has authored
or co-authored over 120 research papers. He is a TPC Member of several
IEEE, ACM, and IFIP conferences.
MANUEL ESTEVE received the M.Sc. degree
in computer engineering and the Ph.D. (Dr.-Ing.)
degree in telecommunication engineering from the
Universitat Politècnica de València (UPVLC) in
1989 and 1994, respectively. He is currently a Full
Professor with the Escuela Técnica Superior de
Ingenieros de Telecomunicación, UPVLC, where
he is also the Leader of the Distributed Real-Time
Systems Research Group. He has over 25 years of
experience in the ICT research area in the area of
networking. He is managing several research and development projects at
regional, national, and international levels. He has collaborated extensively
in the research and development of projects for the government agencies and
defense, and EU-FP6, EU-FP7, and EU-H2020 Programs as the Chairman
of the agreement between Spanish MoD and UPVLC. He has authored or
co-authored over 100 research papers.
VOLUME 4, 2016 11
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2019.2963283, IEEE Access
D. Sarabia-Jácome et al.: Seaport Data Space for Improving Logistic Maritime Operations
FERNANDO BORONAT (M’93–SM’11) was
born in Gandia, Spain. He received the M.E.and
Ph.D. degrees in telecommunication engineer-
ing from the Universitat Politècnica de Valèn-
cia(UPV), Spain, in 1994 and 2004, respectively.
After working for several Spanish telecommuni-
cation companies, he moved back to UPV in 1996,
where he is currently an Associate Professor with
the Communications Department. He is also the
Head of the Immersive Interactive Media R&D
Group, UPV, Gandia Campus. He has authored two books, several book
chapters, an IETF RFC, and more than 100 research papers. He is an Editor
of Media Sync: Handbook on Multimedia Synchronization(Springer,2018).
He is involved in several IPCs of national and international journals and
conferences. His main topics of interest are communication networks, mul-
timedia systems, multimedia protocols, and media synchronization. He is a
member of the ACM.
12 VOLUME 4, 2016
... Advances in real-time information gathering and exchange systems and the analysis provided by Big Data have enabled abilities to be developed for the smart management of logistics. This allows, e.g., the calculation of the best transport route depending on the traffic, thus minimizing transportation times and costs (Liu et al., 2019bSarabia-Jacome et al., 2020;Strandhagen et al., 2017), reducing carbon emissions, and improving customer satisfaction (Su and Fan, 2020). ...
... Also, the topics identified in SciMAT that form each of the Clusters in Graphext show that the latter are interrelated. For example, smart manufacturing is related to sustainability and the circular economy (Bag and Pretorius, 2020;Nascimento et al., 2019), logistics (Chekurov et al., 2018;Khajavi et al., 2018;Verboeket and Krikke, 2019), and Big Data (Cavalcante et al., 2019;Fu and Chien, 2019;González Rodríguez et al., 2020;Liu et al., 2019a); smart logistics is related to Big Data Sarabia-Jacome et al., 2020), and resilience to security and Big Data Lohmer et al., 2020). ...
Article
Full-text available
The purpose of this work is to offer a grouping of the aspects that the literature addresses on Industry 4.0 and its relationship with the supply chain, to contribute an innovative literature review and bibliometric analysis technique, and to detect gaps in the research and how these can be covered in the future. The Industry 4.0 concept originated in manufacturing but has evolved over recent years and extended beyond the limits of the focal company to embrace the entire supply chain. This study applies a variety of techniques sequentially. First, the initial phases of a Systematic Literature Review are applied to identify, select, and evaluate documents for analysis. Subsequently, Artificial Intelligence/Natural Language Processing techniques are applied to identify research topics and group articles. Lastly, bibliometric analysis techniques are applied using an innovative tool that enables dynamic data association. A total of 41 research topics are identified that produce a grouping of 663 articles in 8 clusters. The obtained results are used for an analysis of the extant literature, the detection of gaps, and proposals to guide future research in the Industry 4.0 and supply chain sphere.
... It uses the FIWARE ecosystem to store and notify the alerts related to the management of containers and the state of the sea. For the seaport of Valencia, Spain, Sarabia-Jácome et al. [42] implemented a seaport data space also based on the FIWARE ecosystem to share data and facilitate the interaction between stakeholders. This architecture is integrated with a large data platform implemented with Apache Spark to analyze data on terminal operations and vessel positions. ...
Article
Major advances in telecommunications and the Internet of Things have given rise to numerous smart city scenarios in which smart services are provided. What was once a dream for the future has now become reality. However, the need to provide these smart services quickly, efficiently, in an interoperable manner and in real time is a cutting-edge technological challenge. Although some software architectures offer solutions in this area, these are often limited in terms of reusability and maintenance by independent modules —involving the need for system downtime when maintaining or evolving, as well as by a lack of standards in terms of the interoperability of their interface. In this paper, we propose a fully reusable microservice architecture, standardized through the use of the Web of Things paradigm, and with high efficiency in real-time data processing, supported by complex event processing techniques. To illustrate the proposal, we present a fully reusable implementation of the microservices necessary for the deployment of the architecture in the field of air quality monitoring and alerting in smart ports. The performance evaluation of this architecture shows excellent results.
... It should be mentioned that Key Performance Indicators (KPIs) are widely applied in shipping marketing and companies' performance evaluation [67]. KPIs have been developed using Big Data architecture functionalities [68], showing a dashboard to allow easy interpretability of results for planning vessel operations. Marlow and Casaca [13] proposed new port measurement indicators that, besides considering quantitative aspects, also focus on qualitative issues, measure lean port performance and sustain the subsequent development of agile ports. ...
Article
Full-text available
Currently, seaports are actively searching for methods and ways to improve their operational efficiency. Digitalization is considered as one of the main directions of current ports’ development. Ports’ digitalization levels are varied and may depend on different factors, including port size, traditions, turnover and handled cargo type, etc. Ports often face decision-making challenges related to assessment of their digitization level and choice of development directions. The article aims to develop a methodology to evaluate ports’ digitalization level. A marketing research tool was used to collect the data needed for the analysis. A mathematical model allowing simulations is proposed and a case study of 30 ports located in the Baltic, North and Mediterranean Seas regions is explored. Based on conducted calculations, a ranking of analysed ports considering their digitalization level was created. The ports were compared within groups of small, medium-sized and large ports. It was estimated that the digitalization level in small and medium-sized ports is about 30% lower than the level of large seaports. The research results may be of interest to seaports striving to assess their level of digitalization and choose the best digital improvement solutions.
... The Industry 4.0 has already affected several domains connected to maritime logistics [6]. The concept is driven by digitalization and optimization of industrial processes that use the new emerging technology [7,8]. ...
Chapter
Full-text available
Safe and efficient vessel traffic flow is the overall aim of maritime transportation. Each of the different components of the transportation system contributes to this aim. The current situation is characterized by substantial changes in the maritime world. They are caused by he introduction of new technologies and the implementation of e-Navigation concept. This concept bases on sustainable digitalization and ambitions and may once allow for unmanned and autonomous shipping. Basing on participant observation, this paper analyzes the present situation from the perspective of shore-based vessel traffic management elements and suggests an approach to address rising challenges by research.
... PCSs are defined as centralized port information and data hubs that integrate and distribute data from different sources [36]. The PCS is a system that centralizes the vessels information and the goods they transport so that the stakeholders can better control and coordinate the movements of goods [37]. The main goals when developing a PCS are, most of all, increasing data quality, paperwork reduction, facilitating data connection between various stakeholders, improving and supporting operations across the entire transport and logistics chain [38]. ...
Article
Full-text available
This paper researches the role of the port authorities in port governance, and the role of the port authorities in Port Community System implementation. The authors provide the theoretical frameworks of seaports, port authorities, port governance, and Port Community Systems. The literature review was carried out using the Web of Science database and additional relevant sources. The authors concluded that although different port governance models exist (regarding the type of port authority), there is no evidence as to which governance model is universally preferable, as it is specific to each seaport. In addition, the research has shown that port authorities play a very important role in the implementation of a Port Community System, increasing the sustainability of seaport operations. Its implementation enables the port authorities to evolve into real digital hubs and neutral data managers, which ultimately leads to the optimization of seaport processes and more efficient use of transport infrastructure.
... The crucial managerial problem for such compound systems as multimodal transport terminals is the proper data management aiming at obtaining, processing, and analyzing information about the system parameters (including characteristics of demand, technological processes, and the resulting efficiency indicators). To ensure the quality and adequacy of the data on transport terminals' functioning, the advanced data management systems are being developed [2], data envelopment analysis is being implemented [3], and complex network data models are being constructed [4]. For the purposes of comparative analysis and evaluation of the multimodal terminals' sustainability level, surveys are held [5][6][7] and the Analytical Hierarchy Process method is used [8]. ...
Article
Full-text available
Multimodal transport terminals are the elements of transport systems that ensure the interaction of enterprises of various modes of transport. The effective functioning of transport terminals significantly affects the efficiency of the material flow servicing in a supply chain and the sustainability of the whole transport system. The paper proposes an approach to determine the optimal parameters of production resources in multimodal transport terminals, based on numerical computer simulations of technological operations in a transport terminal for the given parameters of incoming and outgoing material flows. The practical use of the proposed approach is shown on the example of the Amur Harbor cargo area of the Dnipro River Port.
Chapter
This paper aims at devising an IoT-based system to collaboratively calculate quantitative composite indicators among Industrial actors. Requirements have been considered to build a reliable, trusted, anonymised environment within which involved stakeholders can safely share information on the indicators while keeping sovereignty over their data. Technologies such as enhanced Context Brokers, the usage of Industrial Data Spaces agents and the introduction of distributed publish-subscribe data exchange schemes lay the foundations of the proposed solution. This work entails the design of the system altogether with a proposal of technologies for its deployment. Drawing from previous public-funded projects’ results and highly relying on open-source technologies, the authors have come up with a solution that aims to be proven in real environments in the near future.
Chapter
In the transport chain there are involved several public and private companies including ports, terminals, hauliers companies, shipping lines, freight forwarders, customs, etc. The transport activity requires the exchange of data and documentation between these companies. However, each of them has its own system and it is not easy to share data. This data can be used to provide better services to their clients. This chapter presents the demonstration of the INTER-IoT components in a real environment in the port and logistic domain. Three of the main actors in the transport chain are sharing data in the port of Valencia. With the data provided, we defined three different scenarios focused on access control and traffic, dynamic lighting, and wind gusts detection. With this pilot we have shown the benefits of sharing data in the port and logistic sector using INTER-IoT.
Chapter
This article gives insight into management of water infrastructure with the application of space technology. The article also outlines the involvement of South Africa in space-related matters. The three pillars of space technology, namely, satellite communication; satellite-aided position, timing, and navigation; and Earth observation, are outlined. Preparation of the workforce for the swift changes in the space technology sector is discussed. The barriers to business in the space technology sector are outlined. A brief discussion of Africa’s difficulties as a global space technology player and the resistance of some African states are provided. The number of satellites for African states, years launched, the country of ownership, and progress and performance made by African states within the space industry to date are presented. Within this context, an outline of the application and benefits of space technology to water infrastructure is given.
Conference Paper
Full-text available
The concept of the digital twin is gaining momentum in the maritime industry, but it brings with it new questions about data ownership and governance. Given the wide range of different data that can be stored in the digital twin, it is not clear if all these data can be stored in the same data base or even use the same governance model. Within the Industrie 4.0 framework, a concept called "Industrial Data Space" (IDS) has been developed to cater to these problems for land-based industry. With the advent of "Shipping 4.0" as the Industrie 4.0 of the seas, we propose a similar concept as IDS for the maritime world, called "Maritime Data Space" (MDS). This will be an extension of IDS but will be adapted to the specific nature of shipping and, hence, a modified digital trust framework, lower connectivity for ships at sea and the highly heterogenous control and physical systems architecture on ships. However, MDS will still be a part of the IDS ecosystem and gain benefit from the further developments in IDS and its applications.
Article
Full-text available
Exhaust emissions from vessels have increasingly attracted attention in the continuously growing marine transport world trade market. The International Maritime Organization (IMO) has introduced a number of measures designed to reduce exhaust emissions from global shipping. As one of the busiest ports in the world, Qingdao port has been studied to propose possible support to the development of efficient emission reduction. In this study, a large amount data of emissions inventory in Qingdao port was used to predict its annual exhaust emissions, and hence, to help understand maritime pollution in Qingdao port. Bigdata analysis methodology was employed to perform accurate predictions on vessel emissions. The analysis results show that the emissions were dominated by container ships, oil tankers, and bulk cargo ships. The comparison between Qingdao port and other ports in emission control areas demonstrates the necessity of control measures for exhaust emissions. The adoption of shore power and efficient cargo handling seems to be a potential solution to reduce exhaust emissions. The findings of this study are meaningful for maritime safety administration to understand the current emission situation in Qingdao port, propose corresponding control measures, and perform pollution prevention.
Article
Full-text available
We are in front of a new digital revolution that will transform the way we understand and use services and infrastructures. One of the key factors of this revolution is related to the evolution of the Internet of Things (IoT). Connected sensors will be installed in cities and homes affecting the daily life of people and providing them new ways of performing their daily activities. However, this revolution will also affect business and industry bringing the IoT to the production processes in what is called Industry 4.0. Sensor-enabled manufacturing equipment will allow real time communication, smart diagnosis and autonomous decision making. In this scope, the Industrial Data Spaces (IDS) Association has created a Reference Architecture model that aims to provide a common frame for designing and deploying Industry IoT infrastructures. In this paper, we present an implementation of such Reference Architecture based on FIWARE open source software components (Generic Enablers). We validate the proposed architecture by deploying and testing it in a real industry use case that tries to improve the maintenance and operation of milling machines. We conclude that the FIWARE-based IDS implementation fits the requirements of the IDS Reference Architecture providing open source software suitable to any Industry 4.0 environment.
Conference Paper
The evolution of seaports to smart ports is the most important revolution they must face in order to meet the high requirements of efficiency, economy, and security. Achieving this evolution is a task full of challenges. One of these challenges is the interoperability of the stakeholders’ information systems. Interoperability has been widely studied but there is still some reluctance to adopt the solutions because the stakeholders are afraid of sharing their data. Recently, the Industrial Data Space (IDS) reference architecture proposes several recommendations and specifications to solve this issue based on data sovereignty concept. This architecture has not been studied nor implemented for the seaport use case yet. For this reason, this paper presents a seaport data space based on IDS architecture to share information in a secure and interoperable manner. Along with this, a big data architecture is added to exploit the shared data. The seaport data space is implemented using the FIWARE IoT platform. The seaport data space is tested by sharing the data of the port authority to the port terminal and analyzing the large shared data through descriptive analysis. The results show that by using the seaport data space, the shared data allow improving the decision making in the planning operations of the seaport.
Article
Purpose This paper aims to study the function of standardisation in intermodal transport. It identifies where standardisation helps to improve intermodal transport, who is active in intermodal transport standardisation, what types of standards are needed and what the decision-making process and implementation of standards should be like to positively influence the performance of intermodal transportation. Design/methodology/approach An empirical study is designed to carry out this research project. The empirical study starts with a review of the organisations that are responsible for standardisation and intermodal transport, together with the standards that they have developed, and are developing, so far. It continues with analysing the topics where standardisation helps to improve the performance of intermodal transport. The analysis is based on 12 interviews, followed by desk research, to validate the respondents’ statements. Findings The results show that intermodal transportation should be distinguished in continental and maritime transport, which require different standards. In maritime transport, the hardware aspects of the system are highly standardised. However, further standardisation of information exchange offers potential to improve the quality of transport. For continental transport, challenges appear in the heterogeneity of infrastructure and loading units used in Europe. For both systems, openness and consensus are main requirements for the development of successful standards. Originality/value Standards facilitate interoperability, quality and safety of intermodal transportation, which leads to better performance. This has drawn little attention in the literature. This study addresses this gap and focuses on Europe.
Article
The complexity of port terminal operations and intermodal transshipments has pushed the introduction of information and communication technologies (ICT) to assist the multiple stakeholders involved in the multiple decision-making processes. More recently the development of systems based on the Internet of Things (IoT) represents a further step in the automation of data acquisition and processing. We present here an IoT system designed to optimize, manage and monitor container transport operations along an intermodal corridor, combining rail scheduling and inland vessel navigation. The system has a modular design, enabling the interaction with external systems, independently of their nature, through the cloud-based FIWARE platform. We provide a description of the three inter-related subsystems (container tracking, rail management and inland navigation), and a description of the expected supply chain benefits derived from the implementation of the system, both in terms of shippers and of terminal integration in the logistics flow. This system is currently operative at the Port of Seville, the main transshipment node in the Madrid – Seville – Canary Islands intermodal corridor in Spain.
Chapter
The purpose of the work is to study the possibilities of formation of Industry 4.0 as a new vector of growth and development of knowledge economy by the example of modern Russia and to develop recommendations for their practical implementation. For this, the method of dynamics (horizontal and trend) analysis of time rows and correlation analysis are used. The indicator of development of knowledge economy is the corresponding index that is prepared by the World Bank. The vectors of growth and development of knowledge economy are share of innovations-active companies in the structure of entrepreneurship, number of developed completely new leading production technologies, and share of high-tech spheres in economy (as the indicator of development of the sphere of science and education) according to the Federal State Statistics Service. The author shows that knowledge economy, which was developing dynamically at the initial stage of its formation, has slowed down. The existing growth vectors—innovational entrepreneurship, high-tech spheres of economy, and the sphere of science and education—have depleted their potential and cannot ensure its further development. It is necessary to look for such vectors, of which the most perspective is Industry 4.0, as formation of Industry 4.0 leads to growth of knowledge economy: innovational development, increase of the values of indicators of socio-economic development of economic system, and increase of the role of intellectual component of economy—the sphere of science and education.
Conference Paper
The most recent and prominent advances in industrial computing include the growing interconnectivity of cyber-physical devices, as well as the increasing variety of complex applications exchanging data across company domains. In this context, the data becomes a valuable business asset and a trade good. The Industrial Data Space is a platform designed for the industry, allowing organizations the efficient data exchange and trade. The possibilities such platforms enable inevitably come along with new security risks regarding the establishment of trust, communication security, data usage control, or the integrity of participating systems. We define the key security requirements for the operation of such platforms in untrusted environments and present an overall security architecture for the whole ecosystem including the secure design and implementation of an architecture for the participating cyber-physical devices. On these devices, we allow for the controlled and isolated execution of services for application-specific gathering, processing and exchanging of data between organizations.
Conference Paper
Applications and Services are located at the top of the IoT deployments and represent a big portion of the IoT stack. Each Platform has several IoT Services and these use to be domain-oriented and very heterogeneous. This heterogeneity hinders the interoperability between services and applications from different IoT Platforms. Flow Based Programming is a paradigm that allows the interconnection of services and the creation of an execution flow between them. There are several solutions that use this paradigm within the IoT platforms, but none is mainly focused in the aim of connect services from different IoT platforms. To address this, a methodology is described to develop functionalities to access to the services, and an architecture is provided with different components to offer a solution to this interoperability problem. This solution offers advantages in the registry, cataloging and discovery of services, and in the creation and management of composite IoT Services. Finally, this paper show a use case related with transport and logistic to validate this approach