The objective of this research is to develop a DSS model that can accommodate criteria and risks in making a decision and optimisa- tion on a multimodal transportation route for LSPs or SMEs. The value of this research is in a methodology for integrating quantitative and qualitative criteria and assessing in a real world situation. The quantitative criteria are budget and time. The qualitative criteria are risk of freight damaged, risk of infrastructure and equipment and risk of other factors of a multimodal transportation route. The DSS model (as illustrated in Fig. 1) is the combination of a number of models beginning with the multimodal transport cost-model to achieve cost and time of each multimodal transportation route, risk identified by factor analysis and assessed by using the Delphi method from experts, followed by AHP to weight quantitative and qualitative criteria for ZOGP in the next step to optimise route for user needs in each criterion on the model that has been examined in an in-depth collaboration with major logistics firms in Thailand and Vietnam.
The contribution of this research lies in the development of a new approach that is flexible and applicable to LSPs or SMEs, in selecting a route under user needs in quantitative and qualitative criteria for minimising time and/or cost. Normally quantitativeanalysis e.g. GP, etc. is one of the most popular models in a decision making problem. Using quantitative criteria alone, however, can mislead and be inadequate – Kengpol and O’Brien (2001), Kengpol (2004) and Kengpol (2008) – and therefore there is a need to utilise quantitative and qualitative analysis together. This research has applied AHP to overcome the burden on qualitative analysis by designing a computer programme to assist the decision maker dealing with AHP and ZOGP calcula- tion. This DSS is simple and flexible for users who have limited budget, time and information of risks for transportation. The advantage of this research is that a user can select the optimal multimodal transportation route and give the significant weight as needed.
In terms of limitations, there are difficulties for a single logistics company to utilise DSS because of lack of information provided for the database. This DSS database is in need of a great number of updated data on all routes and modes between Thai- land and Vietnam such as costs of routes, time of transportation, the change of infrastructure and equipment of new route and law, etc. This DSS should be able to link the Thailand and Vietnam logistics associations that have a number of networks for updat- ing data in the database. Both associations should maintain and update information in DSS.
For further studies, the first issue is the combination of environmental criteria for making decisions such as energy consumption, where carbon dioxide (CO2) emission should be explored. Some risks of route vary seasonally and the routes in GMS countries are very dynamic in fields such as: legal, infra- structure and material handling. The second need of the DSS is to obtain accurate and reliable forecasting risk of each route. According to Huang et al. (2006) and Sun et al. (2007) the extreme learning machine (ELM) can be applied to a faster learning algorithm to find weight for neural network forecasting. At present, GMS countries lack collection of data in some risk factors that may affect the training data for adopting this tool. Finally, the model can be enhanced to multiple objective deci- sions making for designing an optimal route based upon user needs in quantitative and qualitative criteria and other criteria e.g. scheduled transportation modes and transportation econo- mies of scale. Sometimes, the cost of transportation is incon- sistent because some routes have a higher demand than supply, then the costs depend on satisfaction of LSP: this means the cost may be unreasonably high.