LE/Dok! fit/Ob i h ‘VCB % 2 / \b EXPERT SYSTEM ON MARK - UP SIZE DECISION IN COMPETITIVE BIDDING BY m mMTWk $&¥%& G R S PATHIRANA THE THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENT FOR THE DEGREE OF MASTER OF SCIENCE (CONSTRUCTION PROJECT MANAGEMENT) SUPERVISED BY Dr. A A D A J PERERA &£ V *05 b°[\ 6 5 DEPARTMENT OF CIVIL ENGINEERING FACULTY OF ENGINEERING UNIVERSITY OF MORATUWA MORATUWA, SRI LANKA University of Moratuwa OCT 2005 85965 8 5 9 ti 5 %6 9 65 Declaration '1 certify that this thesis does not incorporate without acknowledgement any material previously submitted for a degree or diploma in any University to the best of my knowledge and believe it does not contain any material previously published, written or orally communicated by another person or myself except where due reference is made in the text. I also hereby give consent for my dissertation, if accepted, to be made available for photocopying and for interlibrary loans, and for the title and summary to be made available to outside organizations' | Signature of Candidate Date To the best of my knowledge, the above particulars are correct. Supervisor Dr, Asoka Perera BSc Eng, MSc, PhD, CEng, M!E(SL) To my Parents for their endeavours Acknowledgments It is a great pleasure to express my gratitude to the project supervisor Dr. AADAJ Perera in appreciation of the valuable guidance and encouragement given by him and his generosity with invaluable time spent on providing helpful suggestions to make this project a success, is also highly appropriated. I should also express my sincere gratitude to Mr. K Imrias of Construction Project Management division of University of Moratuwa and Ms. Kumuduni Ratnayake for assisting me on both technical and practical aspects for the research. I also wish to acknowledge the assistance given by my employer Sri Lanka Telecom, especially Mr. U P Hattella (DGM-Civil) for easing the burden of my regular duties to attend to this research. Further, the special gratitude is due to my wife for her patience and understanding during the period of my study programme and to my daughter Vedithya Oct 2005 G RS Pathirana Abstract In line with the advancement of society, the construction industry with no exception, has been upgrading rapidly. As the contributors to the construction industry raises the complexity of the industry too raises demanding us to explore and analyze the industry with much needed attention. The bidders who can cope up with the competitive nature of the bidding only will survive. Therefore it is very much in importance to have a bidding strategy, which leads to win- win situation to both the client and the contractor. The critical early analysis of factors affecting the mark- up size decision for any given project plays a vital role. The contractors’ behavior of bidding affects by a large number of factors which ranging from the construction company characteristics (internal to the company considered) to Macro Economic environment (society at large) including the project specific characteristics. In this context the bidding decisions are of highly complex, unstructured where clear guidelines are difficult to set up. The decisions on bidding will be usually made based on the intuition and experience of the domain experts. The aim of this exercise is to develop a Knowledge Based System (KBS) to help the contractors to streamline their attention on to most critical factors identified, which are affecting bidding decision and to suggest a reasonable range of mark up size for a given project under specific context. There were ten important factors selected through intensive literature review; Availability of Projects, Need for work, Owner Client relation, Past profits in similar projects, Rate of return in investment, Experience in similar projects, Cash flow (negative), Current work load, Competition, and the KBS was developed using the Fuzzy logic tool box on Matlab platform. This KBS enables the decision makers to evaluate the impact of said factors on a specific bid situation. Given the subjective nature of the mark up size decision the Fuzzy set theory, which is a sub branch of Artificial Intelligence (AI), enables the assessments to be arrived in qualitative and approximate terms. Seven decision rules were constructed based on the expert comments. Seven sets of data analysis were carried out in this system. The quality of information and awareness of the decision on mark up size of a particular tentative project that can be gained from this model may help the construction companies to obtain a competitive edge in bidding. Declaration Acknowledgements Abstract Table of Contents List of Figures List of Tables 1. INTRODUCTION 1 1.1 BACKGROUND MAIN OBJECTIVES METHODOLOGY MAIN FINDINGS PLAN OF THE THESIS 1 61.2 1.3 6 1.4 8 101.5 LITERATURE REVIEW2 12 FACTORS AFFECTING IN BIDDING2.1 12 Introduction2.1.1 12 Previous Researches2.1.2 13 2.1.3 Comments 38 2.2 REVISITING OF BIDDING MODELS 40 Introduction2.2.1 40 Previous Models2.2.2 42 2.2.3 Comments 53 ARTIFICIAL INTELLIGENCE2.3 59 Introduction 2.3.1.1 Components of Expert Systems 2.3.1.2 Common uses of an Expert system 2.3.1 59 64 64 Models on Expert System platform 2.3.2.1 Neural Networks 2.3.2.2 Fuzzy logic 2.3.2 67 68 69 2.3.3 Comments 70 3 FUZZY KBS FOR MARK UP SIZE DECISION 75 3.1. 75INTRODUCTION 3.2. SYSTEM DEFINITIONS AND DEVELOPMENT 75 3.2.1. 75Conceptual Design System Development Data Collection Fuzzification Normalization of Fuzzy Variables Quantifying of Fuzzy Variables Mapping of the Fuzzy Membership Functions Fuzzy Rule Base Inference by the System Defuzzification 3.2.2. 76 7932.2.1. 3.2.22. 3.22.3. 3.22.4. 32.2.5. 3.22.6. 3.22.7. 3.22.8. 80 81 82 85 86 90 92 3.3 KBS FOR BID MARK -UP SIZE ( CombiD) 95 3.3.1 Components of CombiD 3.3.1.1 FIS Editor 3.3.12 3.3.1.3 3.3.1.4 3.3.1.5 95 96 Membership Function Editor Rule Editor Rule Viewer Surface Viewer 98 101 102 104 Testing of the System Advantages of the System Limitations of the System 3.3.2 105 3.3.3 108 3.3.4 109 4 CONCLUSIONS AND RECOMMENDATIONS FOR FURTHER RESEARCH 110 CONCLUSIONS4.1 110 RECOMMENDATIONS FOR FURTHER RESEARCH4.2 114 References Appendix List of Figures Fig. PageDescription No. No. Contributors of the Construction Industry 2 Research Methodology 3 Multi-Attribute Bid Reasoning Model 4 Hierarchy of Competition 5 Hierarchy of Risk 6 Hierarchy of Need for Work Hierarchy of Company’s position in Bidding Probability of beating the Contractor Vs Mark-Up 9 Data flow of a CBR 10 Process of CBR 11 Components of Competitive Bidding Strategy Model and those Relationship 12 Task Domains of Artificial Intelligence 13 Fuzzy and Neural Function Estimators 14 Flow of Data in a KBS 15 A model of an Artificial Intelligence 16 Process of formulating Fuzzy Membership Functions 17 A Membership Function 18 GMP and GMT evaluation of Fuzzy Linguistic Descriptions 19a De fuzzification with COA 19b De fuzzification (Descriptive) 20 Fuzzy Inference System 21a FIS Editor with its Features 21b FIS Editor of CombiD 22a Membership Function Editor with its features 22b Membership Function Editor (of CombiD) 23a Rule Editor with its Features 23b Rule Editor (of CombiD) 24a Rule Viewer with its Features 24b Rule Viewer (of CombiD) 25a Surface Viewer with its Features 25b Surface Viewer (of CombiD) 26 Comparison of Results 1 3 7 16 20 21 22 7 22 8 43 46 48 52 61 72 76 77 81 85 91 94 94 95 97 97 98 99 101 102 103 104 105 106 107 List of Tables Fig- Description Page No. No. 1 Selected Factors with Sources List of Factors affecting in Bidding Factors affecting Bid/No Bid Decision Factors affecting Mark-up size Decision Factors affecting the Chance of Winning a Project Factors that make contractors feel Desperate to obtain a Job Factors affecting Bid Mark-up size Decision Importance Indices and Rank order of the Factors considered by Medium size and Large size contractors Application of Expert System Values obtained from Interviews Factors considered with Importance indices and Rank order Sample set of Crisp values Results of Inference using an Example Assigned Ranks by an Expert Results of the Testing of the System 9 2 18 3 24 4 26 5 28 6 28 7 33 8 35 9 66 10 84 11 86 12 91 13 92 14 106 15 107