A decision support system for supplier selection

Abstract

Background Supplier selection is the procedure of selecting the most suitable supplier to deliver the project’s goal within the defined parameters of time, cost and quality. In the construction service industry, it has been realized over time that the evaluation of suppliers is a multi-criteria based process used to determine the competence of potential suppliers to perform the work if awarded, as opposed to using a single factor like lowest offered price. In Kenya, procurement in governmental organizations is regulated by the Public Procurement and Asset Disposal act, which requires the evaluation of suppliers to be done both in technical and financial basis, taking into consideration price, quality, time and service. The “quality” and “service” factors are imprecise and subjective during an evaluation process hence fuzzy in nature. The common way of evaluating suppliers is by forming technical committees which may be characterized by biases and the evaluation process itself takes time and have associated costs. Due to the advancement of e-procurement in supply chain management there is a need to automate the supplier selection process to enhance efficiency and effectiveness. Objectives The objectives of our study were to determine the most important criteria in supplier selection and to design, develop and assess a decision support system that can be used in supplier selection using the fuzzy set theory. Method This study presented an approach to help decision-makers evaluate potential suppliers by utilizing fuzzy inference system. Initially, the main quantitative and qualitative criteria used in supplier selection process in the construction service industry were identified from experts and literatures. A structured questionnaire was designed and sent to experts in the construction industry in Nairobi who scored these criteria based on the importance of their usage and their weightage was used to rank them. The most critical criteria in the evaluation process were picked for our study and subsequently Matlab Fuzzy Inference System (FIS) was utilized to develop the selection model. The applicability of the proposed model was tested using a real tender for government works and the model evaluated by the users. Results The most important criteria used for supplier selection in the construction service industry were established and a decision support system developed successfully. The validation of the developed model by the decision makers yielded a score of 79.75 % on testing with sample tenders. Data entry to the system took less than a minute while results were provided within seconds. Limitation The proposed prototype can only be accessed via Matlab® software which is not readily available in many governmental organizations and its usage requires technical skills on the software which may limit number of users. Also defining fuzzy sets and membership functions for a given system can be a subjective task. It is normally performed by a collaborative effort of users, process owners and those who possess expert knowledge in the relevant field. There are disagreements among each party when attempting to formulate the fuzzy sets, rules and membership functions and this could compromise the degree of match between system offer and actual user's need. Conclusion It was concluded that using the Fuzzy inference system resulted in an optimum solution and thus can support in decision making when selecting suppliers. One major advantage of the proposed method is that the supplier selection process takes a considerable shorter time and the system makes the selection process more systematic and realistic as the use of fuzzy set theory allows the decision maker to express their assessment on contractors’ performance in linguistic terms rather than a crisp value.