A decision support system using soft computing for modern international container transportation services
Introduction
Since the 1970s, the container transportation has been rapidly developed. Nowadays, more than 60% of maritime transportation employs containers with steady 6.4% increase each year. Moreover, it achieves 100% in some developed countries [1]. Fig. 1 shows the volume change of maritime shipping transportation from 1980 to 2004 around the world [2], in which the cylinder denotes the container volume of use, and the curve denotes the percentage of container in the maritime transport. Moreover, all the shipping companies use the big container vessels based on the large-scale economy advantage, and the container type has changed from a single size of 20-foot to that of 20-foot and 40-foot. All these factors have brought more complexity to the maritime transportation management.
There have been a number of systems available in the field of container transportation. For instance, Shen and Khoong presented a support system for the empty container distribution planning [3], Korea Logistic Net developed the port management information system [4], and LOADSTAR was created by IBM [5], etc. However, some of these systems focus on only part scenarios of the whole transportation process, or the problem models solved by the other approaches do not meet the demands of modern logistics. Besides the global economic integration, the container transportation service field faces more challenges, and the companies also need to increase profits. All of these require new software systems to support the decision-making.
Based on the modern demands proposed by COSCO (China Ocean Shipping COmpany), the research team designs and develops a software system named PROFITS (PRofit Optimization For International Transportation Service). This decision support system can be used in all aspects of the maritime transport management. It is a flexible and efficient computer system enabling officers to improve decision-making. PROFITS not only meets all the requirements of the classification society, but also offers strength and stability calculations and a wide range of transportation services and management facilities.
In this paper, we will introduce three modules: the demand forecasting, the stowage planning, and the shipping line optimization. The rest of the paper is organized as follows: in Section 2, we present the system architecture of the software that gives an overview of the proposed solution model. Section 3 discusses the demand forecasting, in which some classical algorithms are employed. In section 4, a detailed description of container stowage planning is presented, and a mathematical model is constructed and optimized by the linear programming method. Section 5 describes the shipping line optimization with genetic algorithm and sequence alignment. Finally, the conclusion and future extensions of the system are summarized in Section 6.
Section snippets
Software architecture
The PROFITS system is developed using the Eclipse1 platform and related technologies. The system architecture is shown in Fig. 2.
The container distribution and stowage planning modules belong to the container level; the demand forecasting, the stowage planning(repeat) and slot pricing and allocation modules are at the vessel and voyage level; the contribution analysis and shipping line optimization are at the shipping line level. The detailed descriptions of the
Problem introduction
Forecasting technicians might say that the art of forecasting consists of generating unbiased estimates of the future value of some variable, the company wants to use forecast to support business, for example, they can make the detailed solution obtaining the commercial trend, therefore gain the initiative in the business. In the marine, one of the major jobs is to detect the booking volume and make the corresponding plan in advance, which assists companies to upgrade the ability on demand.
Container stowage planning
In the container operation, the container unproductive re-handle is a very costly activity. Each upload or download costs about $40–60 [9] in the port. Where does container re-handle come from? In a voyage with a sequence of ports, there may be upload or download in the intermediate ports. If the containers near the destination are stowed under those to farther destination, extra re-handles are wasted. The shipping companies cost too much in this regard, and they are eager to a solution to
Problem introduction
There are many time-sensitive data during the shipping period. For example, the trade volume and profitability differ from port to port, in some transportation case, the shipping company only changes one vessel journey (for instance extend the other port) and cancels another one to achieve the transportation task. in addition, according to the recurrent historical journey changes, the company needs a decision-making supporter to make a suggestion to adjust the existing shipping line network.
Conclusion and future work
PROFITS can be used for serving modern international container transportation. It includes the demand forecasting, the stowage planning, the shipping line optimization, slot pricing and allocation, the container distribution, and the contribution analysis modules. In this paper, we present the first three modules. The system relates to the whole links of maritime transportation, and the software uses various features of decision support system by creating a framework for forecasting and
Acknowledgements
Greatly acknowledge the Business Optimization team at IBM China Research Laboratory.
This software project and research works are supported by European Commission under grant No. TH/Asia Link/010 (111084); National Natural Science Foundation of China under Grant No. 60433020, 60673099, 60773095, project 20091022 supported by graduate innovation fund of Jilin University, and 863 project 2007AA04Z114.
References (19)
- et al.
A dss for empty container distribution planning
Decision Support Systems
(1995) - et al.
Stowing a containership: the master bay plan problem
Transportation Research Part A
(2004) - et al.
Container terminal operation and operations research: a classification and literature review
OR Spectrum
(2004) - et al.
An analysis of marine container transportation in the asian region
- KLN Korea logistics network, Tech. rep., http://www.klnet.co.kr,...
- IBM Loadstar, Tech. rep., http://www-05.ibm.com/services/dk/bcs/transport/en/shipping/loadstar.html,...
- et al.
The fundamental theorem of exponential smoothing
Operations Research
(1960) Applied Linear Regression
(2005)- et al.
Computational Intelligence
(2001)
Cited by (21)
Container terminal daily gate in and gate out forecasting using machine learning methods
2023, Transport PolicyA novel secondary decomposition learning paradigm with kernel extreme learning machine for multi-step forecasting of container throughput
2019, Physica A: Statistical Mechanics and its ApplicationsCitation Excerpt :They are primarily based on a model of emulating the process of the human system of identification, which employs the related spatial and temporal characteristics from the historical data patterns [18–22]. Because these AI techniques can approximate any complex level without prior knowledge of the problem solution, they have been widely applied in port throughput forecasting [23–26]. Xie et al. implemented a comparative study based on the LSSVR model for container throughput forecasting [27].
Analytics meets port logistics: A decision support system for container stacking operations
2019, Decision Support SystemsA decision support system for integrated container handling in a transshipment hub
2018, Decision Support SystemsData characteristic analysis and model selection for container throughput forecasting within a decomposition-ensemble methodology
2017, Transportation Research Part E: Logistics and Transportation ReviewCitation Excerpt :Some econometric models have previously been used for container throughput forecasting. Those are the error correction model (ECM) (Fung, 2002; Hui et al., 2004), the multivariate regression model (Chou et al., 2008; Veenstra and Haralambides, 2001), and the seasonal autoregressive integrated moving average (SARIMA) model (Schulze and Prinz, 2009), as well as other models, for example, exponential smoothing (Liu et al., 2010) and the Vector Auto-Regression (VAR) model (Tian et al., 2013). However, econometric models are built on linear assumptions.
A HPSO for solving dynamic and discrete berth allocation problem and dynamic quay crane assignment problem simultaneously
2016, Swarm and Evolutionary Computation