Artificial Intelligence for Maritime Transport
Maritime Industry is one industry that heavily relies on humans for decision-making, work processing, and operations. It changes traditional business practices using digital technology such as computer software, database, and web pages. The application of using Artificial Intelligence (AI) and machine learning for navigating traffic could facilitate and improve productivity.
Maritime industries have conducted several initiatives to explore new digital technologies. Shipowners would prefer to minimize human-related activities if the financial operating cost is rational and AI can attain better productivity. AI enables the transformation to paperless and automated procedures. However, the maritime industries struggle when facing big data for analysis and implementation to enhance work productivity. One of the common issues is the lack of understanding of how AI can impact the business and the lack of cooperation and standards among partners. We will unearth the potential of AI and machine learning for maritime applications.
Applications in Maritime Industry
According to research, there are four main clusters of AI in maritime research involving big data, as seen in Figure 1. They are namely: digital transformation, application of big data from Automatic Identification System (AIS), energy efficiency, and predictive analytics. The digital transformation, application of big data from AIS, and energy efficiency are interlinked. The identified research for this cluster is on autonomous ships, big data, AI, cybersecurity, the Internet of Things (IoT), and virtual reality.
Figure 1: Main clusters in Big data and AI Applications in Maritime 
Digitalization  focuses on operations automation, business process automation, and information processing. For example, the digitalization in maritime transport such as using e-navigation systems to better organization of data and communication between ships and shore.
On the other hand, Port Community Systems is an inter-organizational system that electronically integrates heterogeneous compositions of the public and private sectors, technologies, systems, processes, and standards within the port community. Before any investment, it is essential to evaluate the port community systems. A study has shown that Multi-Criteria Decision-Making Analysis  could improve the decision-making process when assessing, selecting, or purchasing in maritime logistics, port, green shipping and port, safety, and security problems.
In addition, the concept of the Internet of Vessels (IoV)  was proposed to innovate maritime transport. The IoV is a network of intelligent, interconnected vessels and shore facilities with a series of digital entities. It integrated all key technologies, such as sensing, automation, telecommunications, information, computers, and intelligent control, into a single platform. It was found that the use of eco ship with patented fuel saving and emission reduction systems and container technology, big data solutions for ship information management, and automation of systems are the three most effective solutions. However, the three main challenges to big data analytics for maritime organizations are the lack of understanding in improving business using analytics, lack of executive sponsorships, and lack of skills.
Moreover, maritime surveillance supports widespread maritime policies such as maritime security, illegal bunkering, tracking of marine oil transportation, and search and rescue. Some studies in this sub-cluster focused on anomaly detection. There are typically three anomaly detection approaches: data mining, statistical, and machine learning . Bayesian networks (BN) for anomaly detection in vessel tracking  were used. For example, the Maritime and Port Authority of Singapore (MPA) and IBM recently collaborated on Sense-making Analytics For maritime Event Recognition (SAFER). They use the machine learning approach to predict vessel arrival time and potential traffic hot spots for unusual behaviors of vessels and illegal bunkering. But the challenges include the unforeseen weather condition that hinder the accuracy of the prediction.
Furthermore, Speed Optimisation was used to increase the energy efficiency of the vessels moving freight and passengers between specified ports. Liner companies  usually consider slow steaming as the best practice. The proposed approach is effective in reducing vessel energy consumption and CO2 emissions. Apart from speed optimization, energy efficiency can be achieved from intelligent shipping route planning. An improved hybrid model for intelligent shipping route planning via Genetic Algorithms  demonstrated the possibilities of optimizing energy consumption in shipping route planning.
Another application is in visual surveillance systems . Deep learning-based Convolutional Neural Networks (CNNs) models improve the accuracy of visual recognition and verification of maritime vessels. AI was applied for ship image recognition systems. For example, deep learning was used to recognize surrounding vessels for Mitsui OSK Lines (MOL) in the FOCUS EYE Project (see Figure 2). The system recognizes the vessels accurately using high-resolution cameras and a graphic processing unit (GPU). The image recognition technology could also be used to monitor shipping lanes and security operations. However, deep learning algorithms need large sets of labeled data to do their job accurately. So what happens when there is a lack of sufficient labeled training data? It can fail to recognize the presence of the vessel and resulting in collision or untimely human intervention.
It often lacks a mechanism for learning abstractions through explicit, verbal definitions. It can fail spectacularly,
Figure 2: Deep learning for recognizing surrounding vessels for Mitsui OSK Lines (MOL)
(cite from: https://www.mol.co.jp/en/pr/2019/19057.html)
Artificial Intelligence (AI) is an essential tool for future technological development in the maritime industry. The maritime transport sectors are emerging with the rapid advancement of AI. Many work processes have yet to adopt AI and machine learning. As the maritime industry navigates many complexities of sustainable shipping and digitalization, a good talent pool of data engineers and scientists is required to tackle pressing engineering and operation issues. Besides extensive computation capability, supervised machine learning essential to AI still needs many human efforts to label the training data correctly. We can employ an engineer to do data tagging and ensure they have done it properly despite the tedious works. If this problem can be solved, the complexity and advances in machine learning algorithms can cause another “transparency” problem. People need to understand how the decision is made and how fairly the decision is made without bias. It implies that the engineer needs to be both multidisciplinary and proficient in coding and legality of AI.
 Z.H. Munim, M. Dushenko, V. J. Jimenez and M. H. Shakilb, "Big data and artificial intelligence in the maritime industry: a bibliometric review and future research directions," Maritime Policy & Management, vol. 47, no. 5, pp. 577–597, 2020.
 E. Tijan, M. Jović, S. Aksentijević, A. Pucihar, “Digital transformation in the maritime transport sector,” Technological Forecasting and Social Change, vol.170, 120879, 2021.
 L. Heilig and V. Stefan, "Information Systems in Seaports: A Categorization and Overview," Information Technology and Management, vol.18, no. 3, pp. 179–201, 2017.
 P. T. Woo and L. Z. Yang, "Multi-Criteria Decision Making in Maritime Studies and Logistics: Applications and Cases, Springer, 2018
 Z. Tian, F. Liu, Z. Li, R. Malekian and Y. Xie, "The Development of Key Technologies in Applications of Vessels Connected to the Internet," Symmetry, vol. 9, no. 10, p. 211, 2017.
 A. Sidibé and G. Shu, "Study of Automatic Anomalous Behaviour Detection Techniques for Maritime Vessels," The Journal of Navigation, vol. 70, no. 4, pp. 847–858, 2017.
 D. Handayani and W. Sediono, "Anomaly Detection in Vessel Tracking: A Bayesian Networks Approach," International Journal of Maritime Engineering (RINA Transactions Part A), Vol. 157, no. A3, pp. 145–152, 2015.
 J. S. L. Lam and X. Zhang, "Innovative solutions for enhancing customer value in liner shipping," Transport policy, vol. 82, pp. 88–95, 2019.
 S. Kuhlemann and K. Tierney, "A genetic algorithm for finding realistic sea routes considering the weather," Journal of Heuristics, vol. 26, pp. 801–825, 2020.
 B. Solmaz, E. Gundogdu, V. Yucesoy, A. Koç and A. A. Alatan, "Fine-grained Recognition of Maritime Vessels and Land Vehicles by Deep Feature Embedding," IET Computer Vision, vol.12, no.8, pp. 1121–1132, 2018.
|Cheng Siong Chin is the Associate Professor (Reader) at Newcastle University. He received his Ph.D. in Applied Control Engineering at Research Robotics Centre in Nanyang Technological University (NTU) in 2008 and M.Sc. (Distinction) in Advanced Control and Systems Engineering from The University of Manchester in 2001. He worked in the consumer electronics industry for a few years before moving into academia. He currently holds 3 U.S. Patents in electronics test systems and components. His research interests include intelligent systems modeling and designing complex systems under uncertain environments such as marine electric vehicles, energy storage systems, and acoustic systems.
He is an author/co-author of over 100 peer-reviewed papers on journals and conference proceedings. He was an elected Vice-Chairman for the IEEE Oceanic Engineering Society Singapore Section, Associate Editor for IEEE Access Journal, and IEEE Transportation Electrification Community (TEC) eNewsletter. He has served as General Chair, Session Chair, and Technical Committee in various international conferences.
M. Venkateshkumar is an Assistant Professor (Sr. Gr) in the Department of Electrical and Electronics Engineering, Amrita School of Engineering, Amrita Vishwa Vidyapeetham, Chennai Campus. He has over 12 years of teaching and research experience and has specialized in the area of power systems engineering.
About the Newsletter
TEC Call for Articles - 2022 Second Quarter Theme: Electrified Mass Transit
The TEC eNewsletter is now being indexed by Google Scholar.