Journal Search Engine
Search Advanced Search Adode Reader(link)
Download PDF Export Citaion korean bibliography PMC previewer
ISSN : 1598-7248 (Print)
ISSN : 2234-6473 (Online)
Industrial Engineering & Management Systems Vol.19 No.2 pp.426-441
DOI : https://doi.org/10.7232/iems.2020.19.2.426

A Study on Ship Assignment Strategies of Shipping Companies Accounting for Shipper Time Value

Kenji Tanaka, Heng Wang, Takaaki Kawanaka*, Jing Zhang
Technology Management for Innovation, Graduate School of Engineering, The University of Tokyo, Tokyo, Japan
NS Solutions Corporation, Tokyo, Japan
Institute for Innovation in International Engineering Education, Graduate School of Engineering, The University of Tokyo, Tokyo, Japan
Physics and Information Science, School of Mathematics, Zhejiang Ocean University, Zhejiang, China
*Corresponding Author, E-mail: kawanaka@cce.t.u-tokyo.ac.jp
December 5, 2019 May 2, 2020 May 8, 2020

ABSTRACT


The three major shipping companies consolidated their liner business. However, the global share of the newly consolidated company was a low 6.5%, making it only the 6th largest in the world. Thus, the Japanese shipping industry still remain in a disadvantageous position in terms of competing with more efficient large-scale shipping companies. Hence, Japanese shipping companies require a strategy not based upon price competition in order to survive. In this study, we propose a method for formulating vessel assignment plans accounting for inter-company competition. The flow of research is as follows: 1) Estimates time value accounting for shipper demand fluctuation risk, 2) Evaluates vessel assignment plans in terms of shipping company container transport volume and profit, 3) As an example, focuses on a regional dominance strategy for Japanese shipping company vessel assignment, conducts a case study through simulation, and then evaluates it. From the simulation results, a regional dominance strategy based on the vessel assignment plan formulation method proposed in this study is thought to be superior in the long term as a strategy for The Alliance in Southeast Asia.



초록


    1. INTRODUCTION

    Since the collapse of Lehman Brothers in 2009, there has been an excess of transportation capacity in the shipping industry resulting from increasing uncertainties in the global economy. The cause of this has been a socalled state of “slow trade,” in which the growth rate of global GDP has exceeded the expansion rate of world trade. The shipping industry has grown as a result of trade using liners, and when expansion in trade becomes stag- nant, the industry itself also becomes sluggish. A result of this excess transportation capacity has been a drop-off in container freight. Figure 1 shows the freight per container TEU on main routes (Japan Maritime Center, 2016). Comparing 2015 freight to 2010 freight, freight dropped off in all routes. The drop-off has been particularly pronounced on the US to Asia and Asia to Europe routes.

    Owing to this recession in shipping, the three major Japanese shipping companies ⎯ NYK Line, Mitsui O.S.K. and Kawasaki Kisen Kaisha ⎯ have had persistently sluggish performance. Then, in July 2017, they consolidated their liner business to establish Ocean Network Express Holdings, Ltd. It started service in April 2018 (Nihon Keizai Shimbun, 2017). The purpose of this consolidation was to reduce costs through efficient management. However, the global share of the newly consolidated company was a low 6.5%, making it only the 6th largest in the world. Thus, the Japanese shipping industry still remained in a disadvantageous position in terms of competing with more efficient large-scale shipping companies. Hence, Japanese shipping companies required a strategy not based upon price competition in order to survive.

    In this study, we propose a method for formulating vessel assignment plans accounting for inter-company competition. Thus, focusing on the possibility of a regional dominance strategy for Japanese shipping companies to increase the transport frequency of their liner service and the number of ports of call, we conduct a simulation- based case study of strategies adopted by these companies for container transport within the Asian region.

    The features of this study are the following.

    • 1) Proposes a method to formulate vessel assignment plans while accounting for inter-company competition.

    • 2) Estimates time value accounting for shipper demand fluctuation risk

    • 3) Evaluates vessel assignment plans in terms of shipping company container transport volume and profit.

    • 4) As an example, focuses on a regional dominance strategy for Japanese shipping company vessel assignment, conducts a case study through simulation, and then evaluates it.

    2. PREVIOUS STUDY

    2.1 Study of Shipping Company Strategies

    Studies of shipping company strategies may be categorized into macro methods from an economic approach and micro methods from a numerical approach. Alix et al. (1999) used an economic approach to study the shipping company CP Ships, which was evaluated by comparing the formation of conferences (alliances) between shipping companies and M&A (Alix et al., 1999). Yanagisawa et al. discuss shipping conferences from the viewpoint of economics and service, and further covers the direction of conferences (Yanagisawa et al., 2003). Ding and Liang conducted a study of the decision-making methods when shipping companies choose partners for conferences using the fuzzy multiple-criteria decision-making (MCDM) method (Ding and Liang, 2005). Yang et al. evaluates shipping company conference strategies using the game theory concept of “core.” (Yang et al., 2011) Panayides and Wiedmer discuss the active conditions when there is a cooperative relationship between shipping companies from changes in shipping conferences (Panayides and Wiedmer, 2011). Imai and Nishimura analyze the container mega-ship viability by considering competitive circumstances (Imai et al., 2006). Lee et al. survey the extant research in the field of ocean container transport and discuss a wide range of issues including strategic planning, tactical planning and operations management issues, which are categorized into six research areas (Lee and Song, 2017). The advantage of economic approaches such as those mentioned above is that they offer decision-making methods for individual shipping companies accounting for inter-company competition, while their disadvantage is that they do not account for vessel assignment plans involving the performance and operating schedules of specific ships in use.

    Kimura used a numerical approach to develop a shipping company strategy that optimized liner networks between Japan and China by returning to the problem of routing (Kimura, 2008). Takizawa selected hub ports within Japan using a container transport simulation accounting for shipper sacrifice models (Takizawa, 2009). Qi and Song minimized fuel losses in liner networks based upon restrictions to keep liner service at a constant level of quality or higher (Qi and Song, 2012). Agarwal et al. assess how shipping lines have adapted their liner service schedules (in terms of commercial speed, number of vessels deployed per loop, etc.) to deal with increased bunker costs (Agarwal and Ergun, 2008). Dulebenets formulate a vessel-scheduling problem as a mixed integer nonlinear programming model, where the total liner shipping route service cost is minimized (Dulebenets, 2018a, 2019). Wang and Meng computed a robust operating schedule through nonlinear programming, wherein an overall network was little affected by delays in container ships caused by unforeseen circumstances (Wang and Meng, 2012). Several scholars subsequently study the issue of optimization with empty container repositioning for a liner shipping company (Kimura, 2016;Akyüz and Lee, 2016;Song and Dong, 2012;Brouer et al., 2011;Feng and Chang, 2008;Meng and Wang, 2011;Shintani et al., 2007). Dynamics of liner shipping network and port connectivity in supply chain systems are discussed (Lam and Yap, 2011;Liu et al., 2016). Dulebenets evaluates the effect of introducing both emission constraints and cargo transit time requirements on design of vessel schedules (Dulebenets, 2018c). And he proposes the model to calculate carbon dioxide emission costs in sea and at ports of call for the green vessel scheduling problem (Dulebenets, 2018b). The advantage of numerical approaches such as those above is the design of specific vessel assignment plans; while the disadvantage is that only a company’s own vessel assignment plan is taken into account, and not those of competitors.

    2.2 Study of Demand Fluctuation Risk

    Tezuka et al. (2006) used the demand fluctuation risk approach to quantitatively evaluate the risk of producer supply plans through Monte Carlo simulation accounting for surplus stock and opportunity loss (Tezuka et al., 2006). Kurokawa turned to the topic of the Toyota Group’s production system, and discussed ways of dealing with demand fluctuation risk and construction methods for efficient supply chains (Kurokawa, 2007). Ueno et al. expressed opportunity loss risk from depleted inventory as average value-at-risk (AVaR), and proposed formulations and solutions of multiperiod production planning problems using game theory (Ueno et al., 2015). Chen and Zeng (2010) propose the optimization of container shipping network and its operations under changing cargo demand and freight rates and formulate as a mixed integer non‐linear programming problem (Chen and Zeng, 2010). Wang et al. proposes a tangible methodology to deal with the liner ship fleet deployment problem aiming at minimizing the total cost while maintaining a service level under uncertain container demand (Wang et al., 2013). Jeon and Yeo analyze the optimal timing of container ship orders by considering the uncertain shipping environment by System Dynamics (Jeon and Yeo, 2017). The advantage of the above studies is that they quantify demand fluctuation risk, while the disadvantage is the validity of methods cannot be verified as arguments were made using data generated from random numbers, not reality-based data.

    2.3 Study of Container Shipper Time Value

    An example of a study that uses an approach that involves the evaluation of container shipper time value, which is used when modeling container shipper service selection, is the study by Iyama et al. (2012), who computed the time value of cargo per main routes by analyzing data on the logistics of containers on international waters (Iyama et al., 2012),. Seta and Matsukura computed time value from actual disaggregated data. The benefit of these studies is that they quantitatively compute time value, while the drawback is that they only take container transport costs into account and not shipper demand fluctuation risk (Seta and Matsukura, 2014).

    Among existing studies that have dealt with shipping company strategies, there have been no studies simultaneously dealing with inter-company competition and the design of detailed vessel assignment plans. In addition, no studies have been established that evaluate shipping company vessel assignment plans by modeling shipper service selection while accounting for shipper demand fluctuation risk.

    3. STUDY METHOD

    Figure 2 shows the overall process flow for this study.

    STEP 1. Input data

    • ・Shipper data: Product demand, costs

    • ・Geographical data: Distances between ports

    • ・Demand data: Regional OD

    • ・Cost data: Fuel prices, port factors, cargo handling prices

    • ・Shipping company data: Freight, shipping network

    STEP 2. Estimation of time value accounting for shipper demand fluctuation risk

    Time value is the value used in sacrifice models, and represents tradeoffs in container transport times and amounts of money. We shall explain an estimation method for time value in detail in Section 5.

    STEP 3. Formulation of shipping company vessel assignment plan

    The vessel assignment plans referred to here indicate mainly liner operating schedules, and ones in which liner operating networks are provided. In Section 5, we shall describe a vessel assignment plan design method.

    STEP 4. Container transport simulation

    Simulations performed in consideration of the diverse needs of chippers, and irregularities in ship performance and waiting times. Vessel assignment plans are evaluated according to shipping company container transport volume and profit. In Section 6, we shall explain container transport simulation in detail, and in Section 7, we shall describe a case study using it.

    STEP 5. Discussion

    Discussion of simulation results. We shall describe this in Section 8.

    4. ESTIMATION OF TIME VALUE ACCOUNTING FOR SHIPPER DEMAND FLUCTUATION RISK

    Here, we estimate time value using a sacrifice model.

    4.1 Sacrifice Model

    A sacrifice model is when a shipper selects a liner service, selecting the one with the smallest amount of sacrifice in terms of required time and money. Equation (1) is the formulation of the amount of sacrifice accounting for container transport on land and at sea.

    S r = L C r + S C r + α C L T r
    (1)

    • Sr : Total amount of sacrifice when using service r [yen/TEU]

    • LCr : Cost of transport on land to reach a representative port when using service r [yen/TEU]

    • SCr : Cost of transport at sea when using service r [yen/TEU]

    • α : Shipper time value [yen/ (TEU∙days)]

    • LTr : Total lead time when using service r [days]

    Here, a representative port refers to the largest port in a given region, and a regional port refers to any other port. Shipper time value is a conceptual time value, and thus can only be estimated, but in this study, we estimate time value accounting for shipper demand fluctuation risk. This is one of the novelties of this study.

    4.2 Demand Forecast

    In this study, we use exponential smoothing as a demand forecast method. The three reasons for this are that forecasts can be made using just demand time series data, the forecasts are rarely very wrong, and this method is in wide actual use because it is simple and intuitively understood. Equation (2) is a prediction formula using exponential smoothing.

    d ^ t + 1 = d ^ t + β ( d t d ^ t )
    (2)

    • dt : Actual amount of demand at a point in time t [units]

    • d t : Predicted amount of demand at a point in time t [units]

    • β : Smoothing parameter

    4.3 Shipper Inventory Management

    In general, shippers dislike opportunity loss from shortfalls in inventory. Thus, shippers try to avoid shortfalls by maintaining safety stock. Safety stock is determined using Equation (3).

    I s a f e = K s a f e × σ * L T
    (3)

    • Isafe : Safety stock [units]

    • Ksafe : Safety factor

    • σ : Demand standard deviation

    • LT : Lead time [days]

    4.4 Ordering System

    In this study, fixed interval and fixed quantity ordering is adopted as the ordering method. This is so it is possible to make orders reflecting demand forecast results. Equation (4) shows a formula for determining order quantity Qt at point in time t.

    Q t = { i = 1 T d ^ t + i + I s a f e I t , i = 1 T d ^ t + i + I s a f e > I t 0 , e l s e
    (4)

    • Qt : Order quantity at a point in time t [units]

    • d ^ t : Predicted demand quantity at a point in time t [units]

    • It : Inventory quantity at a point in time t [units]

    • Isafe : Safety stock [units]

    • T : Order interval [days]

    4.5 Sales Volume

    We model shipper inventory and shipping circumstances based on the above demand forecast method, inventory management method and ordering method, and then determine shipper profit. Equation (5) shows a formula for determining the sales volume Ut of shipper products at a point in time t. In this formula, if demand quantity is falling below the sum of inventory quantity and arrived order quantity, then demand can be satisfied, but falling below means that only the quantity of products on hand can be sold. Equation (6) shows a formula for determining the amount of opportunity loss OLt at a point in time t. In addition, Equation (7) shows a formula for determining inventory quantity It at a point in time t.

    U t = { d t , I t 1 + Q t L T d t > 0 I t 1 + Q t L T ,   e l s e
    (5)

    I t = I t 1 U t + Q t L T
    (6)

    O L t = d t U t
    (7)

    • Ut : Sales volume at a point in time t [units]

    • dt : Demand quantity at a point in time t [units]

    • It : Inventory quantity at a point in time t [units]

    • Qt : Order quantity at a point in time t [units]

    • OLt : Opportunity loss quantity at a point in time t [units]

    • LT : Lead time [days]

    4.6 Shipper Expected Profit

    Shipper expected profit is determined by deducting costs from sales proceeds.

    O P ( L T ) = t = 1 p e r i o d { p U t ( c G U t + c Q Q t + c I I t + c O L O L t ) } c G I p e r i o d
    (8)

    • OP(LT) : Shipper expected profit with lead time LT [yen]

    • Ut : Sales volume at a point in time t [units]

    • Qt : Order quantity at a point in time t [units]

    • It : Inventory quantity at a point in time t [units]

    • OLt : Opportunity loss quantity at a point in time t [units]

    • p : Selling price [yen/unit]

    • cG : Cost of goods purchased [yen/unit]

    • cQ : Transport cost factor [yen/unit]

    • cI : Storage cost factor [yen/unit]

    • cOL : Opportunity loss factor [yen/unit]

    • period : Compuation period [days]

    4.7 Time Value Estimation

    In this study, time value is the value showing how much money was lost when one TEU container was delayed in transportation by one day. Thus, the shipper expected profit computed in the preceding section needs to be converted into shipper expected profit from sales proceeds per TEU. Equation (9) shows a formula for determining shipper expected profit OPC (LT) per TEU. Furthermore, as value is computed in units of TEU containers, then care should be taken as the sales volume units are different from those in the preceding section.

    OPC ( LT ) = O P ( L T ) t = 1 p e r i o d U t
    (9)

    • OPC(LT) : Profit per TEU with lead time LT [yen/TEU]

    • OP(LT) : Shipper gross profit with lead time LT [yen]

    • Ut : Sales volume at a point in time t [TEU]

    • period : Shipper profit computation period [days]

    We created a scatter diagram with expected profit per TEU determined as above on the vertical axis and lead time on the horizontal axis. We then used a linear approximation with this scatter diagram.

    The slope of the line in Figure 3 indicates how much shipper expected profit per TEU decreases when lead time is prolonged by one day. This is equivalent to the definition of shipper time value in the sacrifice model. Therefore, in this study, shipper time value assumes a linear slope estimated with the above method.

    5. SHIPPING COMPANY VESSEL ASSIGNMENT PLAN FORMULATION METHOD

    In this section, we design a vessel assignment plan for when a regional dominance strategy is taken to increase liner service transport frequency and the number of ports of call.

    Once weekly is basic for transport frequency in present liner businesses. This is because it has become easier for shippers to make plans as the day of the week that a container ship arrives at a port is fixed. However, transporting goods once per week is not necessarily best for a shipper. If fluctuations in the demand for transported goods are large, then high frequency transport of goods is preferred in order to avoid demand fluctuation risk.

    In addition, it is believed that more shippers will use a service the greater the number of ports of call. When a shipper wants to transport a container from a departure port to an arrival port, it may be good to use a liner directly linking the two ports if available. On the other hand, if there are no liners linking the two ports, then the shipper can only use transport at sea after transporting a container over land to another port. Land transport is relatively expensive compared to sea transport, which increases the burden for the shipper. Against this backdrop, shippers often prefer sea transport. Consequently, increasing the number of ports of call is a strategy adopted by more shippers.

    On the other hand, increasing transport frequency and the number of ports of call, despite leading to improved service, also leads to increased costs for the shipping company. Hence, when designing vessel assignment plans, transport frequency and the number of ports of call must not be increased arbitrarily, but with consideration given to the costs versus effects of a strategy. Here, we use a shipping company’s transport volume and profit when providing a designed service as evaluation indices of a vessel assignment plan.

    5.1 Listing of Vessel Assignment Plan Candidates

    We consider five transport frequencies: once every two weeks, once every week, twice every week, three times every week, and seven times every week (daily service). Regarding the number of ports of call, we consider how many other ports could be added to an existing vessel assignment schedule. We assumed the target route section of a vessel assignment plan to be Port A and Port B. Equation (1) shows a formula for computing the number of days at sea DAB from Port A to Port B.

    D A B = L A B v
    (10)

    • DAB : Number of days at sea from Port A to Port B [days]

    • LAB : Distance between Port A and Port B [nautical mile]

    • v : Container ship speed [kt]

    We compare the number of days at sea DAB computed from Equation (10) and the number of days at sea D A B ^ from Port A to Port B in an existing vessel assignment schedule, and if D A B ^ is larger than DAB, determine the surplus number of days from D A B ^ DAB in the existing vessel assignment schedule, and determine whether it is possible to make a port of call at a regional port around Port B. Furthermore, we assume that cargo handling time is not considered.

    As in the above, we prepare vessel assignment plan candidates by combining transport frequency and number of ports of call.

    5.2 Profit Per Shipping Company Service

    Below we show a formula for computing shipping company profit.

    P r = S C r C i = 1 a l l r T R i i = 1 a l l r ( C f i + C p i + C h i + C t i )
    (11)

    • Pr : Shipping company profit with service r [yen]

    • SCr : Service r container freight [yen/TEU]

    • TRi : Ships i container transport volume [TEU]

    • Cfi : Ships i fuel cost [yen]

    • Cpi : Ships i port cost [yen]

    • Chi : Ships i cargo handling cost [yen]

    • Cti : Ships i charter cost [yen]

    • allr : Number of ships comprising service r [ships]

    Sub-sections 6.2.1 to 6.2.5 show computation methods for the fuel costs, port costs, cargo handling costs and charter costs of ships i.

    5.2.1 Fuel Costs

    Equation (12) shows a formula for determining the fuel costs Cfi of ships i. We assume C heavy oil for fuel, and cite a numerical formula and a factor for outwardbound container ships described by Suzuki of fuel consumption volume. In addition, Equation (13) shows a formula for determining ship fuel consumption volume FOi, and Equation (14) shows a formula for determining the displacement when fully laden, DSPi, of ships i.

    C f i = K f × F O i × L i
    (12)

    F O i = 6.87 × 10 5 × { D S P i ( 1 0.35 × L F i ) × D W i } × D S P i 1 3 × V i 2
    (13)

    D S P i = 1.37 × D W i + 1660
    (14)

    • Cfi : Ships i fuel costs [yen]

    • Kf : Fuel prices [yen/kg]

    • FOi : Ships i fuel consumption volume [kg/km]

    • Li : Ships I operating distance [km]

    • DSPi : Ships i displacement when fully laden [tons]

    • LFi : Ships i slot utilization

    • DWi : Ships i loading weight [tons]

    • Vi : Ships i speed [km/h]

    5.2.2 Port Costs

    Port costs are those required when a ship enters a port. They specifically account for ships’ harbor charges and port facility use fees. According to the majority of harbor fare structures, these costs are proportional to ships’ gross tonnage. Equation (15) shows a formula for determining the port costs CPi of ships i.

    C p i = K p × G T i × N p i
    (15)

    • Cpi : Ships i port costs [yen]

    • Kp : Port factor [yen/ton]

    • GTi : Ships i gross tonnage [tons]

    • Npi : Ships i arrival frequency [times]

    5.2.3 Cargo Handling Costs

    Cargo costs are those required when unloading containers from a ship. Equation (16) shows a formula for determining the cargo handling costs Chi of ships i.

    C h i = K h × N h i
    (16)

    • Chi : Ships i cargo handling costs [yen]

    • Kh : Cargo handling factor [yen/TEU]

    • Nhi : Ships i cargo handling container quantity [TEU]

    5.2.4 Charter Costs

    Charter costs are costs including all ship crew costs, repair costs, lubricating oil costs, depreciation costs (including loan interest), and insurance premiums. According to Kimura, there is a correlation between ship charter costs and loading capacity (Kimura, 2008). In this study, we cite an approximation formula through regression analysis of charter costs and loading capacity by Kimura (Kimura, 2008). Equation (17) shows a formula for determining charter costs Cti of ships i.

    C t i = ( 1490 × T E U s i + 206600 ) × d a y i
    (17)

    • Cti : Ships i charter costs [yen]

    • TEUsi : Ships i loading capacity [TEU]

    • dayi : Ships i number of days use [days]

    5.3 Shipping Company Gross Profit

    Equation (18) shows a formula for determining shipping company gross profit TPs.

    T P s = r = 1 a l l s P r
    (18)

    • TPs : Shipping company gross profit [yen]

    • Pr : Shipping company profit with service r [yen]

    • alls : Number of shipping company liners in service [routes]

    5.4 Shpping Company Total Transport Volume

    Equation (19) shows a formula for determining shipping company total container transport volume TTRS. In addition, we also determine shipping company shares based on transport volume from the total container transport volume of each shipping company.

    T T R s = r = 1 a l l s T R r
    (19)

    • TTRs : Shipping company total container transport volume [TEU]

    • TRr : Service r container transport volume [TEU]

    • alls : Number of shipping company liners in service [routes]

    We use shipping company total transport and gross profit as evaluation indices of vessel assignment plans, and evaluate designed services.

    6. CONTAINER TRANSPORT SIMULATION

    Figure 4 shows the overall process flow for container transport simulation.

    STEP 1. input data

    We use the following data as input data:

    • ・Shipper data

    • ・Geographical data

    • ・Demand data

    • ・Cost data

    • ・Freight data

    • ・Operating data

    STEP 2. Container shipper generation

    We focus upon generation quantity and generation location as container shipper generation elements. Generation quantity randomly generates container shippers in a state with two elements (container transport departure port and destination port) from Equation (20). A representative port refers to the largest port in a given region, and a regional port refers to any other port

    N d = O D i j 365 × s e a s o n k × p ( x )
    (20)

    • Nd : Container shipper generation quantity on day d

    • ODij : Annual container demand quantity from representative port i to representative port j

    • seasonk : Business fluctuation in month k

    • p(x) : Probability density function

    Generation quantity is set per representative port, and must account for regional ports regarding generation location. Therefore, in this simulator, the departure port or arrival port is randomly selected by equal probability from among representative ports and regional ports in a region. For example, with ODij, if there are two regional ports in the region of representative port i, and three regional ports in the region of representative port j, then a departure port is randomly selected by equal probability from a total of three ports (representative ports and regional ports) in the region of representative port i, and an arrival port is randomly selected by equal probability from a total of four ports (representative ports and regional ports) in the region of representative port j. This operation is repeated ODij times, thereby generating a container shipper from representative port i to representative port j.

    STEP 3. Selection of container shipper service

    Regarding container transport routes, four are considered: ① transport on sea from a departure port to an arrival port; ② transport on land from a departure port to a representative port in the sector thereof, followed by transport on sea to an arrival port; ③ transport on sea from a departure port to a representative port in the sector of an arrival port, followed by transport on land to an arrival port; and ④ transport on land from a departure port to a representative port in the sector thereof, followed by transport on sea to a representative port in the sector of an arrival port, and transport on land to an arrival port. Figure 5 shows an organization of these four routes. We consider these four transport routes per ship, and assume the service selected by a shipper to be the combination of all ships and all transport routes with the smallest amount of sacrifice for the shipper. Furthermore, we refer to the starting port of on sea transport as the port of loading, and the ending port of on sea transport as the port of unloading.

    STEP 4. Any space in selected container ship?

    We determine whether or not there is space on the container ship selected in STEP 3.

    STEP 5. Container ship space reservation

    Next, the ship’s space is reserved for the service selected by the shipper. The loading and unloading ports are determined according to the shipper’s transport route. Since the use period of the ship is known by the shipper, it is necessary to check whether empty space is available on the ship for that particular period. If it is not available, then service selection and search for other service candidates is reconducted. If empty space is available, the service to be used is selected and a reservation made. Subsequently, the ship used, the loading and unloading ports, and the loading and unloading dates are all set for the container object.

    STEP 6. Container flow process

    Ships travel between ports according to their own schedule. When a ship is stopped at a port, container unloading and loading takes place. The ship can use the port of loading, port of unloading, loading date, and unloading date held by a container object; if the ship used the port of unloading, and the unloading date satisfies certain conditions, the container is unloaded, and if the ship used the port of loading, and loading date satisfies certain conditions, the container is loaded. In addition, after container unloading and loading has ended, ship transport container quantities are recorded, ship fuel costs, port costs, and cargo handling costs are calculated, and results recorded

    STEP 7. Output data

    Finally, we obtain the following data as output data.

    • ・Shipping company transport volume

    • ・Shipping company profit

    7. CASE STUDY

    7.1 Case Study Summary

    A case study, mainly of Japanese shipping companies concerning container transport within the Asian region, is conducted. Currently, container transport within the region falls under two competitive conferences: The Alliance and Ocean Alliance. Japanese shipping companies are involved in The Alliance. Because this study is focused on the Asian region, 2M is not covered here. In this study, the present competition between the two conferences is simulated first. Thereafter, a vessel assignment plan for The Alliance is created assuming a regional dominance strategy through the proposed method, following which the competition between the two conferences is simulated again. The results of these two simulations are used to verify the effectiveness of the vessel assignment plan formulated according to the proposed method.

    7.2 Input Data Settings

    Geographical data, consisting of 12 representative ports and 26 regional ports in Asia.

    Demand data, estimated from movement of goods data within the Asian region obtained from the Japan Maritime Center (Japan Maritime Center, 2016).

    Cost data, determined from published literature, such as the Japan Long Course Ferry Service Association (Japan Long Course Ferry Service Association, 2016).

    Freight data, obtained by referring to the home pages of shipping companies. The freight of The Alliance and Ocean Alliance is determined by referring to the NYK Line and COSCO, respectively.

    Operating data, determined from 2015 data from Ocean Commerce Ltd. (Ocean Commerce Ltd., 2015). The operating data from each conference is summarized in Table 1.

    A simulation is conducted using the above input data for a period of one year.

    7.3 Shipper Time Value Estimation

    Here, in order to perform shipper time value estimation, we used data on printers from Company A for shipper demand data. Company A imports printers from overseas via sea transport, and then sells them according to demand. Demand data lasts for one year in one-day units. Figure 6 shows the printer demand quantity of Company A. In addition, Table 2 shows printer quantity and cost data.

    From the above data, we estimate time value accounting for container shipper demand fluctuation risk. Figure 7 is a scatter diagram showing the relationship of lead time and expected profit per TEU. When linear approximation is used with this data, the slope of the line is -30,000 [yen/ (TEU・days)]. Based on the results, shipper time value is established at 30,000 [yen/ (TEU・ days)].

    7.4 Results of the Simulation Recteating the Case of Present Status

    Using the data from sub-sections 7.2 and 7.3, we recreated present conditions (2015). Table 3 shows the results. From Table 3, we find that The Alliance dominate in Japan and Southeast Asia, and the two conferences are competitive in China and Taiwan.

    The total container cargo handling volume in all ports is 6.57 million TEU for The Alliance and 4.20 million TEU for Ocean Alliance. When we calculate shares based on the container cargo handling volume of each conference from the results, they are 61.0% for The Alliance and 39.0% for Ocean Alliance. In addition, revenue from The Alliance is 200 billion yen, representing 32.4 billion in income and expenditures.

    7.5 Vessel Assignment Plan Formulation

    We apply the design method for liner vessel assignment plans proposed in this study to The Alliance, which is the conference to which Japanese shipping companies belong, and create a vessel assignment plan for The Alliance assuming a regional dominance strategy. Furthermore, we perform a simulation, and evaluate the liner vessel assignment plans of each conference. For shipper value time, we use 30,000 [yen/ (TEU・ days)] similar to the creation of the case of recreating present status.

    Next, of the combinations of transport frequency and number of ports of call for a target route section of the vessel assignment plan, we determine the combination with the largest profit for The Alliance accounting for competition with Ocean Alliance. Table 4 shows a comparison of a present vessel assignment plan and the proposed vessel assignment plan.

    As an example of how Table 4 was prepared, we explain a Japan-Thailand vessel assignment plan. From data on the annual volume of container movement of goods, the average transport volume per day to two ports in Thailand (Laem Chabang Port, Bangkok Port) is 40 TEU, and the average transport volume per day to Japan from the two ports is 100 TEU. Transport frequency in the present service by The Alliance involves the use of four container ships with a loading capacity of 3,000 TEU, making stops at each port with a frequency of once per week. Therefore, the total loading capacity of container ships used does not change as much as possible. We change the transport frequency and loading capacity of a single container ship, and conceive transport frequency for a total of seven routes. Table 5 shows combinations of transport frequency and container ship performance. Furthermore, those with a transport frequency of once per week represent present combinations, while others represent newly conceived combinations.

    Regarding the number of ports of call, the present service by The Alliance operates on a Tokyo → Nagoya → Busan → Laem Chabang → Ho Chi Minh City → Manila → Tokyo route, without calling on regional ports in Thailand. The distance between Laem Chabang Port and Port of Ho Chi Minh City is 612 nautical miles, so the number of days at sea is two days with the presently imagined container ship performance. On the present route, calls are made at Port of Ho Chi Minh City four days after calling at Port Kelang. Therefore, we assume a two-day margin in the Laem Chabang → Ho Chi Minh City section, and determine that it is possible to make an additional call at one regional port in Vietnam. That is, regarding the number of ports of call, it is believed there are two possible routes: making a call only at Port of Ho Chi Minh City, and making a call at Port of Ho Chi Minh City and one regional port.

    Based on the above conditions, we calculate how much transport container volume changes when the transport frequency and number of ports of call are changed. Table 6 shows the container transport volume per week when the transport frequency and number of ports of call are changed. Furthermore, the present vessel assignment plan consists of a transport frequency of once per week, and six ports of call.

    In Table 6, both transport frequencies of three and seven times per week have the same value (1660) when the total number of ports of call is seven. This is because, even if the transport frequency increased, an additional share cannot be obtained. In this case, the total transport volume per week between Japan and Thailand is as follows:

    ( 100 + 40 ) × 2 ports × 7 days = 1960

    The Alliance has a market share of about 85% at the transport frequency of three times per week, amounting to a transport volume of 1660. Owing to the vessel assignment plan for Ocean Alliance, even if the transport frequency is increased from three to seven times per week, an additional share cannot be obtained.

    Based upon the Table 5 container ship fleet data and Table 6 container transport volume, we determine shipping company profit for one year of operation.

    Table 7 shows the changes in shipping company profit for each vessel assignment plan with the present vessel assignment plan as a reference. According to Table 7, when a vessel assignment plan is assumed with a transport frequency of three times per week and seven total number of ports of call, the shipping company profit is at maximum, so this is the Japan-Thailand vessel assignment plan taken. The Japan-Indonesia, Japan-Malaysia, and Japan-Vietnam sections in Table 4 are derived similar to this Japan-Thailand section.

    7.6 Proposed Vessel Assignment Plan Simulation Results

    We then made a container transport simulation using the vessel assignment plan for The Alliance formulated in 7.5. Furthermore, we used the same input data as in the case recreating the present status for vessel assignment plans not of The Alliance. In addition, the simulation period was one year.

    Table 8 shows the container cargo handling volume at each port derived from the simulation, and the shares of each conference. From Table 8, The Alliance dominate in Japan and Southeast Asia, and the two conferences are competitive in China and Taiwan.

    Figure 8 shows the tallying of the container cargo handling volume of each conference divided between representative ports and regional ports. From Figure 8, we find that The Alliance dominate representative ports, and the two conferences are competitive in regional ports. In addition, the container cargo handling volume in all ports is 6.84 million TEU for The Alliance and 3.97 million for Ocean Alliance. When we calculate shares based on the container cargo handling volume of each conference from the results, they are 63.3% for The Alliance and 36.7% for Ocean Alliance.

    Figure 9 shows the tallying of the container transport volume of each conference per region. From Figure 9, we find that The Alliance dominate in transport between East Asia and Southeast Asia, and within the Southeast Asia region, and are competitive in transport within the East Asia region.

    Next, we describe the profit of each conference. Figure 10 shows the profit structure of The Alliance, and Figure 11 shows the profit structure of Ocean Alliance. The revenue of The Alliance is 209 billion yen with 33.4 billion yen in income and expenditures. On the other hand, the revenue of Ocean alliance is 116 billion yen with 20.7 billion yen in income and expenditures.

    8. DISCUSSION

    The simulation results for the present status and proposed method cases are compared and discussed.

    Figure 12 shows the tallying of recreated container cargo handling volume for the present status and the proposed method; the results are divided by representative and regional ports and per conference. It can be seen that container cargo handling volume decreases in representative ports for both The Alliance and Ocean Alliance. On the other hand, container cargo handling volume increases for The Alliance in regional ports. This is because, as a result of The Alliance increasing the number of calls to regional ports, cargo currently carried from regional to representative ports on land, and then taken by sea, is directly transported by sea from regional ports. In particular, at regional ports in Southeast Asia, container cargo handling volume by The Alliance is 13,000 TEU at Da Nang Port in Vietnam in the recreation of the present status, but 89,000 TEU with the proposed method. At Kuantan Port in Malaysia, it is 0 TEU in the recreation of present status, while 113,000 TEU with the proposed method. In addition, at Port of Semarang in Indonesia, it is 60 TEU in the recreation of present status, while 90,000 TEU with the proposed method.

    Figure 13 shows the tallying of container transport volume per region for each conference in the recreation of present status and with the proposed method. From Figure 13, it may be read that The Alliance’ share of container transport volume between East Asia and Southeast Asia, and within the Southeast Asian region, is expanding with the proposed method. On the other hand, it may also be read that there is no change in container transport volume within the East Asia region. The reason is that there was no increase in the number of ports of call within the East Asia region with the proposed method. The present Ocean Alliance status provides a point-to-point route centered on Shanghai and Hong Kong, and round-route service with few days required for one cycle, with high transport fre-quency. Hence, even if The Alliance assume a strategy to increase transport frequency and the number of ports of call among regional ports in China and Taiwan, it is still highly likely they would lose in competition to Ocean Alliance. This situation means that Ocean Alliance will build regional dominance in China and Taiwan, and The Alliance will provide new service in China and Taiwan.

    Next, we discuss how much the two evaluation indices of container transport volume and profit were improved for The Alliance. The Alliance’ share of container transport volume is 61.0% in the recreation of present status, while 63.3% with the proposed method. In addition, the profit of The Alliance is 32.4 billion yen in the recreation of present status, while 33.4 billion yen with the proposed method. From this result, a vessel assignment plan based upon a regional dominance strategy designed according to the proposed method may be said to be superior to the present vessel assignment plan. On the other hand, it is difficult to say if the results of the proposed method represent a marked improvement from present status. The reason is felt to be that Ocean Alliance is building regional dominance in China and Taiwan. However, when we consider the growth of the future global economy, the as yet immature Southeast Asian countries are expected to grow economically even more than China, which has sluggish growth. With such considerations, a regional dominance strategy based on the vessel assignment plan formulation method proposed in this study is thought to be superior in the long term as a strategy for The Alliance in Southeast Asia.

    Finally, it is considered in terms of time value. From the viewpoint of shipper time value, a regional dominant strategy is considered to be rational. This is because in recent years, manufacturers, who are the main shippers of shipping companies, have shifted from small variety mass production to high-mix low-volume production to meet a diversifying demand. Compared to small variety mass production, high-mix low-volume production makes it easier to miss demand forecasts owing to product demand pattern diversification. The more difficult it is to forecast demand, the higher the risk of demand fluctuations and shipper time value. Additionally, shippers with higher time value prefer transportation with shorter lead times. On the other hand, if the shipping company adopts the regional dominant strategy, the transport frequency will increase and shipper lead time will decrease. Consequently, more shippers are expected to select shipping companies that adopt the regional dominant strategy in the future, thus increasing their transport volume share and sales.

    9. CONCLUSION

    In this study, the relationship between lead time and shipper profit was investigated using mathematically modeled shipper inventory management methods, and the time value accounting for shipper demand fluctuation risk was estimated. In addition, we proposed a vessel assignment plan formulation method that considers inter-company competition. Then, the container transport simulation, which includes a shipper service selection model, was used to obtain the container transport volume and profit of each shipping company; the vessel assignment plans were also evaluated along with results from case studies. The strategy of each alliance was simulated, and a strategy that a Japanese shipping company might apply to survive was determined. It is expected that this study will contribute to future formulations of vessel assignment plans for shipping companies.

    Figure

    IEMS-19-2-426_F1.gif

    Main route freight index (Japan Maritime Center, 2018).

    IEMS-19-2-426_F2.gif
    IEMS-19-2-426_F3.gif

    Conceptual diagram of relationship between lead time and profit per TEU.

    IEMS-19-2-426_F4.gif

    Overall process flow for container transport simulation.

    IEMS-19-2-426_F5.gif

    Four transport routes.

    IEMS-19-2-426_F6.gif

    Printer demand data.

    IEMS-19-2-426_F7.gif

    Relationship of lead time and expected profit per TEU.

    IEMS-19-2-426_F8.gif

    Container cargo handling volume per representative port or regional port (proposed method case).

    IEMS-19-2-426_F9.gif

    Container transport volume per region (proposed method case).

    IEMS-19-2-426_F10.gif

    The Alliance profit structure (proposed method case).

    IEMS-19-2-426_F11.gif

    Ocean Alliance profit structure (proposed method case).

    IEMS-19-2-426_F12.gif

    Comparison of container cargo handling volume per representative ports or regional ports.

    IEMS-19-2-426_F13.gif

    Comparison of container transport volume per region.

    Table

    Summary of operating data

    Printer quantity and costs

    Container cargo handling volume and shares per each port (recreation of present status)

    Comparison of present vessel assignment plan and proposed vessel assignment plan

    Transport frequency and container ship combinations (Japan-Thailand)

    Weekly transport volume per vessel assignment plan (Japan-Thailand)

    Changes in shipping company profit per vessel assignment plan (Japan-Thailand)

    Container cargo handling volume and shares per each port (proposed method case)

    REFERENCES

    1. Agarwal, R. and Ergun, Ö., (2008), Ship scheduling and network design for cargo routing in liner shipping, Transportation Science, 42(2), 175-196.
    2. Akyüz, M. H. and Lee, C. Y. (2016), Service type assignment and container routing with transit time constraints and empty container repositioning for liner shipping service networks, Transportation Research Part B: Methodological, 88, 46-71.
    3. Alix, Y. , Slack, B. , and Comtois, C. (1999), Alliance or acquisition? Strategies for growth in the container shipping industry, the case of CP ships, Journal of Transport Geography, 7(3), 203-208.
    4. Brouer, B. D. , Pisinger, D. , and Spoorendonk, S. (2011), Liner shipping cargo allocation with repositioning of empty containers, Information Systems and Operational Research, 49(2), 109-124.
    5. Chen, C. and Zeng, Q. (2010), Designing container shipping network under changing demand and freight rates, Transport, 25(1), 46-57.
    6. Ding, J. F. and Liang, G. S. (2005), Using fuzzy MCDM to select partners of strategic alliances for liner shipping, Information Sciences, 173(1-3), 197-225.
    7. Dulebenets, M. A. (2018a), Application of evolutionary computation for berth scheduling at marine container terminals: Parameter tuning versus parameter control, IEEE Transactions on Intelligent Transportation Systems, 19(1), 25-37.
    8. Dulebenets, M. A. (2018b), Green vessel scheduling in liner shipping: Modeling carbon dioxide emission costs in sea and at ports of call, International Journal of Transportation Science and Technology, 7(1), 26-44.
    9. Dulebenets, M. A. (2018c), The green vessel scheduling problem with transit time requirements in a liner shipping route with Emission Control Areas, Alexandria Engineering Journal, 57(1), 331-342.
    10. Dulebenets, M. A. (2019), Minimizing the total liner shipping route service costs via application of an efficient collaborative agreement, IEEE Transactions on Intelligent Transportation Systems, 20(1), 123-136.
    11. Feng, C. M. and Chang, C. H. (2008), Empty container reposition planning for intra-Asia liner shipping, Mari-time Policy & Management, 35(5), 469-489.
    12. Imai, A. , Nishimura, E. , Papadimitriou, S. , and Liu, M. (2006), The economic viability of container mega-ships, Transportation Research Part E: Logistics and Transportation Review, 42(1), 21-41.
    13. Iyama, S. , Watanabe, T. , and Goto, S. (2012), Construction of an analytical model of freight containers in international waters using a sacrifice model, Journal of JSCE (ocean development), 68(2), 1181-1186.
    14. Japan Long Course Ferry Service Association (2016), fuel oil price trend table (1994 to 2016).
    15. Japan Maritime Center (2016), Changes in the movement of goods at sea within the Asia region.
    16. Japan Maritime Center (2018), Trends in movement of goods on main container routes.
    17. Jeon, J. W. and Yeo, G. T. (2017), Study of the optimal timing of container ship orders considering the uncertain shipping environment, The Asian Journal of Shipping and Logistics, 33(2), 85-93.
    18. Kimura, T. (2008), Study of methods of designing Japan-China container transport networks, master’s thesis, Department of the Environment and Marine Engineering, School of Engineering, The University of Tokyo.
    19. Kimura, Y. (2016), Optimization of empty container relocation planning in container distribution networks at sea, master’s thesis, Department of Systems Innovation, Faculty of Engineering, The University of Tokyo.
    20. Kurokawa, F. (2007), Efficient supply chain for automobile industry, Informatics Research, 24, 51-71.
    21. Lam, J. S. L. and Yap, W. Y. (2011), Dynamics of liner shipping network and port connectivity in supply chain systems: Analysis on East Asia, Journal of Transport Geography, 19(6), 1272-1281.
    22. Lee, C. and Song, D. (2017), Ocean container transport in global supply chains: Overview and research opportunities, Transportation Research Part B: Methodological, 95, 442-474.
    23. Liu, Z. , Wang, S. , Du, Y. , and Wang, H. (2016), Supply chain cost minimization by collaboration between liner shipping companies and port operators, Transportation Journal, 55(3), 296-314.
    24. Meng, Q. and Wang, S. (2011), Liner shipping service network design with empty container repositioning, Transportation Research Part E: Logistics and Transportation Review, 47(5), 695-708.
    25. Nihon Keizai Shimbun (2017), Consolidation of three container shipping companies, ‘It will be a company that surpasses the parent company’ launched ONE, cited 2020 April 21, Available from: https://www.nikkei.com/article/DGXLASDZ10HO5_Q7A710C1000000/.
    26. Ocean Commerce Ltd. (2015), International Transport Handbook (2015 edition).
    27. Panayides, P. M. and Wiedmer, R. (2011), Strategic alliances in container liner shipping, Research in Transportation Economics, 32(1), 25-38.
    28. Qi, X. and Song, D. P. (2012), Minimizing fuel emissions by optimizing vessel schedules in liner shipping with uncertain port times, Transportation Research Part E: Logistics and Transportation Review, 48(4), 863-880.
    29. Seta, T. and Matsukura, H. (2014), Disaggregated time value estimation from actual results data using a sacrifice model, Journal of Marine Science and Technology (JMST), 19, 227-233.
    30. Shintani, K. , Imai, A. , Nishimura, E. , and Papadimitriou, S. (2007), The container shipping network design problem with empty container repositioning, Transportation Research Part E: Logistics and Transportation Review, 43(1), 39-59.
    31. Song, D. P. and Dong, J. X. (2012), Cargo routing and empty container repositioning in multiple shipping service routes, Transportation Research Part B: Methodological, 46(10), 1556-1575.
    32. Takizawa, H. (2009), International container transport network design and hub port selection, bachelor’s thesis, Department of Systems Innovation, Faculty of Engineering, The University of Tokyo.
    33. Tezuka, M. , Hiji, M. , Munetomo, M. , and Akama, K. (2006), Risk visualization and decision support for supply planning under uncertain demand, Journal of Information Processing Society of Japan, 47(3), 701-710.
    34. Ueno, N. , Han, H. , Uno, T. , Taguchi, Y. , and Okuhara, K. (2015), Proposed weekly production planning method combining game theory and AVaR, Journal of the Department of Management Information Systems, Prefectural University of Hiroshima, (7), 203-224.
    35. Wang, S. and Meng, Q. (2012), Robust schedule design for liner shipping services, Transportation Research Part E: Logistics and Transportation Review, 48(6), 1093-1106.
    36. Wang, T. , Meng, Q. , Wang, S. , and Tan, Z. (2013), Risk management in liner ship fleet deployment: A joint chance constrained programming model, Transportation Research Part E: Logistics and Transportation Review, 60, 1-12.
    37. Yanagisawa, H. , Tsuruta, S. , and Kurokawa, H. (2003), Comparative study of shipping conferences, Journal of Japan Logistics Society, 11, 113-119.
    38. Yang, D. , Liu, M. , and Shi, X. (2011), Verifying liner Shipping Alliance’s stability by applying core theory, Research in Transportation Economics, 32(1), 15-24.
    Do not open for a day Close