AIS data analytics for adaptive rotating shift in vessel traffic service
| Date | 08 March 2020 |
| Pages | 749-767 |
| DOI | https://doi.org/10.1108/IMDS-01-2019-0056 |
| Published date | 08 March 2020 |
| Author | Gangyan Xu,Chun-Hsien Chen,Fan Li,Xuan Qiu |
| Subject Matter | Information & knowledge management,Information systems,Data management systems,Knowledge management,Knowledge sharing,Management science & operations,Supply chain management,Supply chain information systems,Logistics,Quality management/systems |
AIS data analytics for adaptive
rotating shift in vessel
traffic service
Gangyan Xu
School of Architecture, Harbin Institute of Technology, Shenzhen, China
Chun-Hsien Chen and Fan Li
School of Mechanical and Aerospace Engineering, Nanyang Technological University
Singapore, and
Xuan Qiu
Department of Industrial Engineering and Decision Analytics,
Hong Kong University of Science and Technology, Hong Kong
Abstract
Purpose –Considering the varied and dynamic workload of vessel traffic service (VTS) operators, design an
adaptive rotating shift solution to prevent them from getting tired while ensuring continuous high-quality
services and finally guarantee a benign maritime traffic environment.
Design/methodology/approach –The problem of rotating shift in VTS and its influencing factors are
analyzed first, then the framework of automatic identification system (AIS) data analytics is proposed, as well
as the data model to extract spatial–temporal information. Besides, K-means-based anomaly detection method
is adjusted to generate anomaly-free data, with whichthe traffic trend analysis and prediction are made. Based
on this knowledge, strategies and methods for adaptive rotating shift design are worked out.
Findings –In VTS, vessel number and speed are identified as two most crucial factors influencing operators’
workload. Basedon the two factors, the proposed data model is verified to be effective on reducing data size and
improving data processing efficiency. Besides, the K-means-based anomaly detection method could provide
stable results, and the work shift pattern planning algorithm could efficiently generate acceptable solutions
based on maritime traffic information.
Originality/value –This is a pioneer work on utilizing maritime traffic data to facilitate the operation
management in VTS,which provides a new direction to improve their daily management. Besides, a systematic
data-driven solution for adaptive rotating shift is proposed, including knowledge discovery method and
decision-making algorithm for adaptive rotating shift design. The technical framework is flexible and can be
extended for managing other activities in VTS or adapted in diverse fields.
Keywords Vessel traffic service, Data-driven application, Rotating shift management, Workload balancing
Paper type Research paper
1. Introduction
Vessel traffic service (VTS) refers to the integrated shore-side service implemented by
competent authorities that provides various navigational support for vessels and extensive
traffic management within designated maritime area (Praetorius, 2014). Along with the rapid
development of global trade and the prosperity of maritime transportation, VTS becomes
more and more important in ensuring maritime safety, promoting traffic fluency, and
protecting the environment (Praetorius and Hollnagel, 2014). It has attracted much attention
from governments, academia, and industrial sectors, where lots of efforts have been made to
AIS for
adaptive
rotating shift in
VTS
749
This research was supported by the Singapore Maritime Institute Research Project (SMI-2014-MA-06),
National Natural Science Foundation of China (71804034), Research Foundation of STIC
(JCYJ20180306171958907), and CCF-Tencent Open Research Fund. The authors would like to thank
all participants who had participated in this study.
The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/0263-5577.htm
Received 31 January 2019
Revised 27 August 2019
24 October 2019
5 November 2019
28 November 2019
Accepted 5 January 2020
Industrial Management & Data
Systems
Vol. 120 No. 4, 2020
pp. 749-767
© Emerald Publishing Limited
0263-5577
DOI 10.1108/IMDS-01-2019-0056
improve its service level, such as enacting standards and regulations (IMO, 1997), building hi-
tech VTS centers (Bloisi et al., 2017), and adopting intelligent decision-making (Kao et al.,
2007). Furthermore, as one of the most important determinate success factors of VTS (Xu
et al., 2017), special attention has been paid on improving the performance of VTS operators
(VTSOs), from increasing their situational awareness (Wiersma, 2010), making effective
performance evaluation (Wiersma and Mastenbroek, 1998), to decreasing fatigue levels (Kum
et al., 2007;Li et al., 2019).
Nevertheless, through investigations at one VTS center, it is found that the VTSO
management lags far behind the development of technologies, which has already become a
major obstruction to further improve the service level of VTS. In current practices, VTSOs
are working at separate operator consoles to manage distinct subregions of the VTSservice
area. Through analyzing the information from several screens, they are responsible for
incoming/outgoing communications with vessels, monitoring traffic status, detecting
potential risks , recording key informatio n, and interacting with the V TS team (Devoe et al.,
1979). Moreover, these tasks are required to be conducted in real time within a highly
dynamic environment, which are complex and intensive in nature that makeVTSOs easily
tired, thus is har mful for their performance ( Praetoriusand Hollnagel, 2014;Xu et al., 2017;
Li et al., 2020). Despite rotating shift being adopted to decrease their fatigue levels,
traditional strategies still dominate its design, and the dynamic of maritime traffic and
differences among work shifts/subregions are seldom considered. It is common that some
VTSOs work with higher workload due to more complex traffic conditions in their areas of
responsibility while the others with lower workload. Under the same schedule of rotating
shift, such imbalanced workload would not only cause complains from VTSOs but also
bring them higher fatigue levels, thus higher possibility of making errors, which may
threaten the safety and fluency of maritime traffic. Therefore, it is urgently needed to
design adaptive rotating shift that considers the dynamics and differences of maritime
traffic to balance the workload of VTSOs, improve their performance, and finally guarantee
a benign maritime trafficenvironment.
Data could be of great help to realize adaptive rotating shift. With the rapid development
of Internet of Things (IoT), it is possible to collect extensive data to facilitate the operation and
management activities (Xu et al., 2015;Li et al., 2017), which has been widely recognized in
diverse fields (Zhong et al., 2017;Xu et al., 2018,2019;Tsang et al., 2018). As a typical IoT
application, AIS (automatic identification system) provides extensive vessel kinematics data
in real-time manner, which makes it possible to monitor and track vessel movement
trajectories. Besides, many efforts have been made based on AIS data, such as maritime
traffic prediction (Xiao et al., 2017), emissions analysis (Traverso et al., 2017), collision
avoidance (Zhang et al., 2015), and anomaly detection (Chen et al., 2014).
However, scarce efforts were made on VTSO rotating shift, and major challenges still
exist. First, managing rotating shift in VTS is extremely complicated that involves numbers
of interrelated factors, including policies, operation tasks, traffic conditions, and human
factors. It is a challenge to identify key parameters and major concerns of designing rotating
shift. Second, since AIS data only contain massive discrete vessel kinematic records, efforts
are needed on processing them into knowledge to facilitate adaptive rotating shift. Third, AIS
data contain records collected in abnormal situations, which may hinder the understanding of
maritime situations and would further affect the accuracy of adaptive rotating shift design.
Considering these challenges, this paper proposes an AIS data analytics solution for
adaptiverotating shift. Specifically,through extensive fieldinvestigations, this paperanalyzes
the problem of rotating shift in VTS and proposes the framework of AIS data analytics.
Meanwhile, maritime traffic data cube is introduced. It could transform AIS data into well-
structuredspatial–temporal dataand facilitate the knowledgediscovery processes. Besides,K-
means-basedanomaly detection method is adjustedto identify and remove theabnormal data
IMDS
120,4
750
To continue reading
Request your trial