Tuesday 23 March 2010

Multicriteria Decision Making For Highway Management System

Generally, infrastructure engineers and planners encounter problems when they engage in planning for operations in both long term and short term projects. In order to make a successful decision, it is the most important concern to apply efficient and effective process of how different objectives, facilities and functions are evaluated. In addition, they need to think properly how measures of value or worth are placed on the various alternatives (Li and Sinha, 2004). The main objective of a decision-making process is to maximize the benefits as well as minimize the cost. As highway management system includes various facilities, objectives and functions having different measurement scales, the benefits under different dimensions are measured into different units that make the decision-making process more complex. To diminish this complexity, all the units of benefits are needed to convert into a non-dimensional uniform unit so that decision-making action can be implemented under same platform (Li and Sinha, 2004).

At present, infrastructure maintenance and rehabilitation decision-making for both network and project level is based on current measured and future predicted facilities conditions (Madanat et al., 1997). However decision-making in highway management is a complex process due to limited budget constraint and multiple criteria of influencing factors. In Thailand, there are still some decision-making processes of highway management. It needs to evaluate the performance and effectiveness of those processes whether all objectives are fulfilled or some objectives are violated. Although, highway management system consists of several operational functions, maintenance is the most important function for highway management system where necessary infrastructures have already been constructed. In addition, highway management system includes various objectives such as service, condition, cost, safety, socioeconomic factors, energy etc. So, it is required to know the amount of their influences in decision-making. For this reason, weighting factors of all objectives should be identified so that they can be easily used in any calculation of decision-making. Moreover, objectives are influenced by some performance indicators such as ADT, lane width, shoulder width, travel speed, skid resistance for safety, IRI for condition, population for socioeconomic demand etc. Therefore, to develop the utility function, those performance indicators and their influences are required to know.

If utility function and weighting factor of each individual are identified, function would give a non-dimensional unit. After accumulating all the utility functions, it might be easy to rank them for making a decision. If a decision-making process concerns all the criteria, it would be efficient and effective for any management system. So, all countries including Thailand should have an effective multi-criteria decision-making process for highway management system.

Mr. Mohammad Mamunur Rashid, made a study which tried to develop and introduce utility functions for various objectives and their performance indicators of highway management system. In addition to this utility function, weighting factors of individual objectives in different facilities provides a non-dimensional output so that planners or decision makers can use those functions for decision-making about a suitable action in highway management system.

The main objective of his study was to develop a multi-criteria decision-making process for highway management system. The basis of decision-making is to maximize the overall benefits as well as to minimize the cost within available resources. In order to achieve the main objective in a systematic and sequential manner, secondary objectives were to : (1) overview the existing decision-making process and accumulating factors; (2) determine the facilities of highway management system and their objectives and factors which influence decision and their performance indicators as well as their weighting factors; (3) develop the utility functions of different performance indicators, objectives and facilities in highway management system; and (4) develop an overall decision-making process for the highway management system.

Conclusion

1. Overview of Existing Process
At the beginning of this study, some existing decision-making processes for the highway facilities maintenance operations were reviewed and it was found that the optimum treatment matrix of TPMS budgeting module is commonly used in Thailand for the pavement maintenance. For the other facilities such as bridge, roadside element and traffic control device, there is no specific decision-making process for the maintenance operation. According to the treatment matrix, maintenance decisions are made by using roughness index (IRI), major deterioration, minor deterioration and traffic volume (AADT). In addition to that there are several important factors such as structural capacity, noise level, cost, skid resistance etc. which are absence in this process. Therefore, it would be a better idea to have such a decision-making process that accumulates as many factors as possible.

2. Determination of Objectives, Performance Indicators and Weighting Factors
At the introductory stage of the questionnaire survey, this study deals in only the identification of objectives and their respective performance indicators that would be necessary for having a maintenance decision and planning. For all facilities, objectives are almost similar such as condition, service, safety, cost, socio-economic factors, energy etc. but those objectives have different performance indicators for different facilities.

Preliminary questionnaire survey was carried out in this study in order to achieve the goal for the identification of all the indicators. Questionnaire included some objectives and their performance indicators of four different facilities such as pavement, bridge, roadside elements and traffic control devices found from previous research and experts were allowed to select all possible necessary objectives and indicators in addition to add new factors that would be important. All the selected objectives and their indicators found from the preliminary questionnaire survey were enlisted as necessary factors and used for developing the final questionnaire.

Final questionnaire survey carried out in this study deals in two objectives such as the determination of the different weights of different objectives and performance indicators and development of utility function for every individual performance indicator. On the basis of the identified factors in the preliminary survey and having some brief interview of experts helped to develop the final questionnaire. In this study, the objective of the determination of relative weighting factors is to differentiate the influences of different objectives and their performance indicators in terms of numeric percentage value so that those can be used to make a decision.

3. Development of Utility Functions
Development of utility functions of different indicators that was the most important part of this study was finally achieved by the final questionnaire survey in two steps. Experts were asked to put relative utility values for different units of one performance indicators. Regression analysis that is one of the most reputed processes was used to have the best fitted equation accumulating the opinions of all experts for every different performance indicator.

The second step of the final questionnaire survey was carried out in the point of non-consistency. Sometimes, expert opinions were contradictory and to diminish this problem, they were asked only on that non-consistent issue. Finally, all the data were in the same trend and used for regression analysis.

4. Development of Overall Decision-making Process
Final decision for the prioritization in maintenance operation would be made accumulating all the weighting factors and utility functions of the performance indicators and objectives of respective facilities. The existing condition data and future predicted data after maintenance should be listed for every alternative. Total utility value of each alternative would be determined by the multiplication of utility values of performance indicators and their overall relative weighting factors. The summation of all multiplication provides the total utility value for each alternative. In another way, utility values found for every performance indicators of one alternative would be multiplied by the weights of the performance indicator in their respective objectives and the weights of the objectives. In this case, summation of all multiplication would provide the same value of total utility value. The main purpose of dividing the performance indicators into different objectives is to provide a breakdown structure so that the limitations can easily be identified and have some correction works effectively. Finally, accumulating all the total utility values, alternative having higher utility value should have the higher priority for the maintenance operations.

His thesis abstract is copied and posted.

ABSTRACT

An effective highway management system includes multi-dimensional features such as several numbers of facilities, operations and objectives. Therefore, the decision-making process in a highway management system is very complex interrelated with multiple dimensions. Moreover, measurement with non-commensurable units makes the process more difficult and complicated. Although having complexity in its decision-making process, it is present demand and very useful for an effective highway management system so that all objectives and dimensions can be fulfilled as closely as possible.

In general, pavements, bridges, roadside elements and traffic control devices are the facilities of highway management system that also includes various objectives such as service, condition, cost, safety, socioeconomic factors, energy etc. In addition, it includes several operations like planning, design, construction, condition evaluation, maintenance, improvement and data management. This study tries to develop a decision-making process for the maintenance operation only that would accumulate all the necessary objectives and their performance indicators in order to have an effective maintenance decision. As the measurement unit of different factors are different and it is very difficult to convert all the units into monitory terms, utility functions of various objectives and their performance indicators are developed that provide a non-dimensional output. Moreover, the influences of different factors are not equal for the maintenance operation. Therefore, weighting factors of all the objectives and their performance indicators are determined so that they can be easily used according to their influences in any decision-making. Finally, the decision-making process that accumulate all the weighting factors and utility functions would help the planners for getting an effective decision for appropriate maintenance operation of different facilities of highway management such as pavement, bridge, roadside elements and traffic control devices.

No comments: