Keywords

1 Introduction

Ocean shipping is one of the most economical, secure and environmental-friendly transportation options for delivering large quantities of goods and cargos across different countries and in fact continents [1]. As such, approximately 90% of the world’s international trading is completed by using ocean transportation, which covers items of all kinds including important natural resources (e.g. iron ore, coal, oil and gas), semi-manufactured products (e.g. car engines), electronic equipment (e.g. TVs, Fridges, PCs), as well as commodity goods and foods [2, 3]. According to the annual statistical report provided by the United Nations, there were 90,917 commercial vessels in service in the world by Jan 2016, with a combined tonnage of 1.8 billion dwt, implying an average tonnage of around 20 thousand dwt per ship [4].

In order to monitor and manage ships and vessels offshore continuously and properly, fleet management systems have been widely used in the maritime industry. Fleet management is the management of a company’s transportation fleet (including cars, vans, trucks, airplanes, helicopters, and ships), and aims to remove or minimize potential risks associated with transportation usage, coordination and maintenance, when increasing productivity and reducing overall transportation and staff costs [5]. Maritime fleet management systems generally include a set of functions, like operation monitoring, maintenance planning and management, route management, fuel consumption management, and crew management [5].

Traditionally, maritime fleet management systems focus on the whole vessel/ship. With the emergence of the Internet of Things and new smart technologies (e.g. sensors, cloud computing, big data tools), there is an increasing trend of developing a smarter generation of fleet management systems, which can be used to monitor the performance, status and behaviour of not just the whole ship (in a macro level) but crucial components and equipment inside the ship (in a micro level), such as engines, water treatment equipment, and propellers. However, developing, implementing and operating such smart fleet equipment management systems in the maritime industry will not be straightforward and can encounter a wide range of socio-technical challenges, which have not been sufficiently explored in the current literature. This paper attempts to fill this knowledge gap by drawing on the perspectives of Computer Sciences, Information Systems, Electronic and Electrical Engineering, and Management Sciences. The paper is structured as follows: the next section presents the technical features of smart fleet equipment management systems (SFEM) used in the maritime sector, followed by a discussion of the methodology that involves a desktop study based on the process of a critical literature review. We then present and discuss a range of identified socio-technical challenges associated with SFEM, with practical suggestions and conclusions drawn.

2 Technical Features of Smart Fleet Equipment Management System (SFEM)

Generally speaking, SFEM systems contain three types of functions targeting on key components of the ship [5, 6]: (1) real-time monitoring and diagnosis: to continuously monitor the operational performance of the component, diagnose signs of problems, and identify changes associated with its internal and external conditions; (2) self-control and optimization: based on the results of monitoring and self-diagnosis, the system can make automatic adjustments on the operational settings of the component in relation to environment changes and needs, and so optimize its performance while reducing cost, fuel consumption and pollution; (3) maintenance and repairing: to perform predictive data analysis, generate proactive warnings on repairing and maintenance needs, and support maintenance planning and decision marking of onshore staff. In order to achieve these intended functions with high smartness, SFEMs typically combines the usage of wireless sensors, microprocessors, automatic control systems, cloud computing data storage, software applications, and big data analytics with enhanced user interface [6].

More specifically, and taking the propulsion system as an example, a marine power plant generally consists of a main engine and several auxiliary engines [3]. The main engine generates the power needed for propeller operation and vessel navigation, and auxiliary engines are used to generate additional energy to support the main engine and ensure that there is sufficient power to support all other systems (e.g. heating system, freshwater treatment system) of the ship to perform properly. When an SFEM system is implemented, each engine will be embedded with one or more sensors to collect real-time data about, e.g. the engine’s temperature, fuel consumption, power rate, and running condition. There could be many other types of sensors to be deployed into different kinds of ship components for collecting the suitable types of data needed, such as propeller shaft rpm sensor, propeller shaft power sensor, propeller shaft torque sensor, freshwater consumption sensor, ship speed over ground (i.e. GPS speed) sensor, and relative wind speed/direction sensor [4]. In addition, SFEM systems will also rely on telematic and GPS technology to track the offshore location of the vessel (and in fact also related components of the vessel [7].

All these data will then need to be transmitted to the onshore control center by using a stable, reliable and cost-effective channel, namely satellite communication system based on DVB-RCS standard (an acronym for Digital Video Broadcasting - Return Channel via Satellite). By using geostationary satellites, DVB-RCS systems can offer large offshore coverage all over the world, wide band and flexible bandwidth allocation [9]. Together with GSM (global system for mobile)/VOIP (voice over Internet protocol) technologies, crew members onboard can even enjoy the services of web browsing, e-mail, e-banking, and Internet voice call [8].

The real-time data collected from various sensors will be initially processed by a local data server and automatic control devices onboard, to allow the system to perform self-control and self-adjustment on the settings of related ship components. But these data will eventually be transmitted through DVB-RCS technologies to onshore data centers for proper storage, processing and analysis. As a large amount of data will be generated from different components and different ships on a daily basis, these data will need to be centrally stored in onshore data centers by using cloud computing technologies. Cloud computing is a technology that provides a sharing resource pool to store large datasets, while also allowing on-demand access to this shared pool of computing resources (networks, servers, data storage, applications and services) via a network [11, 12].

Further to internal data, ship companies may also retrieve and store external environment data (e.g. water-column data, bathymetry, ocean acoustics, bioluminescence, and tides) from the Naval Oceanographic Office (NAVOCEANO in short) [10] in their cloud data center. These internal and external data with large volume and various formats will be integrated together and analyzed by using machine learning, artificial intelligence and big data analytics tools embedded in SFEM systems, which can thus assist the crew and onshore staff in complex decision making and other knowledge processing tasks. Some very specific analytical functions can be designed and developed in SFEM systems, depending on the actual needs of shipping companies, including: (1) voyage performance analysis and management, such as speed optimization, autopilot improvements, trim and draft optimization; (2) hull condition analysis and management, such as propeller condition management, engine maintenance onboard, auxiliary engine maintenance, boiler maintenance; (3) energy consumption analysis and management; (4) prognostic performance analysis and management, such as fault prediction, trending analysis, alarm prediction for various crucial components [13, 14].

3 Methodology

Although the technical features of SFEM can offer very attractive benefits to shipping companies, implementing, operating and using such systems will not be easy in practical terms. The study presented in this paper aims to identify and explore potential socio-technical challenges associated with the implementation and exploitation of SFEM systems by conducting a desktop study based on the process of a critical literature review. This critical review followed the approach proposed by Saunders et al. [15], and relied on surveying and using secondary sources. Literature search for this critical review consisted of two phases.

At the first phase of the critical review, the researchers attempted to retrieve academic articles that are directly related to issues, problems or challenges regarding the development and implementation and usage of SFEM from a number of reputable journals and databases (see Table 1 for details). A set of predefined keywords (e.g. challenges, smart fleet management system, Internet of Things, fleet equipment management, intelligent fleet management systems, vessel, ships, maritime industry) as shown in Table 2 was used in this round of article search. Although this endeavor returned some technical studies on SFEM (e.g. [8, 11, 13]), we were not able to retrieve any relevant articles or in-depth studies on the socio-technical phenomenon under investigation.

Table 1. Journals and databases searched
Full size table
Table 2. Keywords used for journal and database search at stage one
Full size table

As a result, a broader and more extensive literature review was conducted at the second stage. Instead of looking for studies directly about SFEM, this second attempt focused on all related technologies (e.g. sensors, cloud computing, software applications, and big data analytics) covered in fleet management and the difficulties of implementing such kind of technologies (Table 3). Although the same set of academic journals and databases were searched, but an alternative set of search keywords were defined and used (e.g. including new terms like sensors, cloud computing, and big data) to reflect the broader coverage of this round of literature search.

Table 3. Keywords used for journal and database search at stage two
Full size table

With much effort, the researchers successfully retrieved and reviewed a good number of valuable literatures. These retrieved articles and materials were systematically analyzed and synthesized, and then used as raw materials to construct arguments for the identification and exploration of socio-technical challenges associated with the development, implementation and usage of SFEM. These challenges are presented and discussed in the next section.

4 Challenges Associated with the Development, Implementation and Usage of SFEM Systems

The results of the critical literature review process identified a range of interrelated socio-technical SFEM challenges, respectively related to system integration, cloud computing, data quality and big data analytics, user resistance, and operational aspects, as detailed below.

System Integration Challenges.

Crucial ship components like main engines, auxiliary engines, freshwater treatment equipment, and propellers are generally designed, manufactured and maintained by very different and specialized vendors [16, 17]. As a consequence, a separated SFEM system may often be developed for each of these components and equipment by different vendors, using diverse database models and interfaces, different data formats and possibly totally incompatible architectures [18, 19]. Therefore, integrating these separated SFEM systems together with legacy systems (which can in turn have very different IT architectures and formats) can be a very difficult and costly task [17, 19]. In light of this discussion, and in order to protect their own benefits, component/system vendors may often set up strict system access and data sharing policies, which can present further barriers for cross-system integration [20]. This is particularly true in a cloud environment, where client companies often have limited control on their data stored by third-party cloud vendors and have insufficient freedom and right to customize a cloud application and integrate it with other systems [19, 21, 22]. Such system integration issues will inevitably lead to system fragmentation in the shipping company, through the creation of technological islands, which are very often totally isolated and non-communicable and so leading to poor data quality and potentially misleading data analytical predictions and results [20, 23].

Lack of Control in the Cloud.

Traditionally, IT resources (including data, module applications, and database servers) are internally hosted and maintained by user organizations. Accompanied with the emergence of cloud technologies, there is an increasing trend for companies to migrate their internal IT applications and databases into the cloud [24, 25], which allows companies to use these applications and data as on-demand services through a web browser, without physically installing software programs or storing the data in local servers [19, 26]. Such cloud service model enables less hardware investments, as well as less fees and internal hazard for system maintenance and upgrade [11, 12], and is highly suitable for the deployment of SFEM solutions that involve large amount of real-time datasets collected by sensors [25, 26]. However, and in contrast to these attractive benefits, the adoption of cloud technologies can also raise many challenges, uncertainties and problems to shipping companies if they do not plan carefully [12, 19, 27]. The root of these problems lies in the fact that client companies often have less transparency, freedom and rights to control their data and applications in the cloud environment [12, 28]. Further to causing potential system integration problems as discussed above, lack of control and rights can often make client companies suffer from what so called “vendor lock-in” events [22, 28]. A vendor lock-in event refers to the scenario that user companies are not able to change their cloud providers even in the case of service dissatisfaction, due to high cost of moving data from one cloud provider to another and potentially legal traps set up by the original vendor, e.g. cloud providers may claim that they (rather than the client company) are the owner of the data stored [19, 22, 29]. For smart fleet equipment management systems, when cloud vendors are often also shipping component/equipment providers, changing them will be extremely difficult, costly and time-consuming [27].

Challenges on Data Quality and Big Data Analytics.

Maintaining high data quality is crucial for the success of any IT applications [30, 31]. However, our extensive review of the literature identified that data quality of SFEM systems can potentially be influenced by a range of factors [21, 29, 32]. Most notably, real-time data of ship components are collected by using a wide range of wireless sensors, but the quality and sensitivity of sensors used by different providers can vary and so may lead to inaccurate measurements [21, 33, 34]. Further to the usage of sensors, some significant operational data (e.g. actual time and place of component repairing) will need to be input by crew members and/or onshore staff, but human errors can occur during the data entry process, especially when there is a lack of suitable training and there is a high level of user resistance [23, 35, 36]. Moreover, the issue of system integration as discussed above can lead to incomplete, redundant, and potentially very fragmented datasets [23, 37]. Poor data quality will inevitably affect the analytical results (such as proactive fault detection, repairing needs prediction and route planning) generated by big data tools of SFEM systems [12, 23, 37]. On the other hand, the practicality, suitability and feasibility of these big data analytical results and predictions can be threatened by real-world conditions, legal restrictions and operational limitations [32, 38]. For instance, when the SFEM system identifies the need of component repairing and makes advanced warning, the vessel may still not be able to reach onshore repairing centers on time [30, 31, 39]. This is not just caused by distance-related reasons, but may also be attributed to the fact that different countries will have specific legislative regulations, custom control systems, and official documentation requirements, which can potentially delay the voyage and so make the vessel miss the best time for component repairing [32, 33, 38].

Sources of User Resistance.

The implementation and usage of SFEM systems can have a lot of impacts on both crew members and onshore supporting staff [30, 36]. For crew members, they will be assigned with new responsibilities and tasks for monitoring the performance of crucial shipping components, by using SFEM and new operational processes [40, 41]. This will require crew members to learn a new set of skills in order to use and operate the system effectively [23, 35, 42]. However, crew members may not always be willing to accept such new duties and changes [35, 43]. And the learning process could also be difficult to crew members with higher age and lower education level [23, 35, 44]. Similarly, onshore supporting staff will be assigned with more power and freedom to make important maintenance, repairing and route planning decisions, by considering the analytical results and predictions generated by big data functions of SFEM systems [42, 44, 45]. They may also be required to perform further data analysis on the system, in order to explore the datasets more deeply and generate more in-depth insights [43, 45]. Nevertheless, onshore supporting staff may not always be equipped with the right level of skills, experience and knowledge to play such decision-making and data analysis roles [35, 46]. All of these issues can become potential sources of user resistance, which will not just reduce users’ intention/willingness to use the new SFEM system but will often also increase the possibility of human errors and so leading to poor data quality as discussed earlier on [23, 46].

5 Further Discussion and Recommendations

It clearly emerged from our above analysis and discussion that the identified SFEM challenges seem to be interwoven and closely related with each other (e.g. lack of control in the cloud environment can raise problems of system integration, which together with user resistance can lead to poor data quality and affect the results of big data analytics). In order to overcome these interrelated socio-technical challenges, shipping companies need to establish very holistic and thorough system development and implementation strategies, by considering the following recommendations:

Shipping companies need to be very careful when selecting their SFEM system vendors. In particular, they have to make sure that the selected solution can offer high compatibility, and so all legacy and new systems can be integrated easily and seamless. Shipping companies also need to pay specific attention to service agreement with cloud vendors, and need to have a clear understanding about their rights, data ownership, and level of freedom before signing the contract.

In order to minimize potential user resistance, a suitable change management program should be designed and implemented to make sure that all staff concerned are properly trained, know the benefits of the new SFEM systems, and understand the importance of their roles and cooperation. It will also be beneficial to inform, consult and involve users as early as possible in order to improve their level of acceptance toward using the system.

In order to increase the practicality of maintenance and repairing options suggested by big data tools, more flexible plans and collaborative relationships will be needed to be set up with component/system vendors and other shipping companies. For instance, when a component of a vessel breaks down during voyage, it is difficult for onshore support centers to provide immediate help over long distance. But the crew can seek for help from nearby vessels, which may not be from the same shipping company but have the same system vendor and so may be able to offer the needed spare components and even offshore technical support. Such new type of collaborative relationship may also lead to the establishment of new service and business models, and can benefit all parties involved.

6 Conclusions

This paper reviews the technical features of the next generation of vessel fleet management systems, namely smart fleet equipment systems or SFEMs. It also provides a critical analysis and discussion on potential socio-technical challenges associated with the development, implementation and usage of SFEM systems in the maritime industry, together with recommendations. Some important conclusions have been drawn from the study. Specifically, although SFEM relies heavily on advanced information technologies, its success can be affected by a wide range of regional, legal, organizational, operational, analytical, user and vendor-related factors and challenges as identified, explored and discussed in this paper. More importantly, because these factors and challenges will not exist independently and can have very sophisticated interrelationships, they may be very difficult to manage and contain. Overall, it can be concluded that researchers and practitioners should not simply focus on the technical aspects of SFEM, more attention should be paid to socio-technical factors and their interrelationships that can lead to substantial usage problems and even severe technical failures.