Assaf, G and Assaad, R H (2024) A data-driven decision-support tool for selecting the optimal project delivery method for bundled projects: Integrating machine learning and expert domain knowledge. Journal of Construction Engineering and Management, 150(12): 04024181, ISSN 0733-9364
Abstract
Project bundling is an innovative project delivery approach that combines several projects under a single contract. While previous studies have provided important information about different project bundling-related aspects, none have developed guidelines to choosing the best project delivery method (PDM) for bundled projects/contracts. Also, despite that some of the existing research efforts have offered tools to identify the optimal PDM, such studies were conducted for single projects rather than for bundled projects which significantly differ from a normal project in terms of complexity and implementation considerations. Hence, this paper develops a data-driven decision support tool that helps agencies in identifying the optimal PDM for their bundled projects by leveraging machine learning algorithms and domain knowledge while also considering the characteristics and goals of the bundled program. This proposed tool considers and compares the following 5 PDMs: design bid build (DBB); design build (DB); construction manager/general contractor (CM/GC); indefinite delivery/indefinite quantity (IDIQ); and public private partnership (PPP). First, data from previous project bundling case studies were used to identify bundling opportunities (on the program or strategic level) as well as bundling objectives (on the project or contract level). Second, a machine learning model (i.e., multinomial naïve Bayes classifier) was developed to generate a probabilistic distribution for the relative suitability of the five PDMs on the strategic bundling program level. Third, a survey was developed and distributed to collect expert's domain knowledge on the importance of the different project bundling objectives (i.e., on the project or contract level). Lastly, an easy-to-use decision-support tool was developed to calculate individual scores for the different 5 PDMs so that the best PDM could be identified. Ultimately, this paper presents an intuitive and easy to implement tool for selecting PDMs for bundled projects based on the integration of machine learning algorithms and domain knowledge.
| Item Type: | Article |
|---|---|
| Index terms: | multinomial, case study, public private partnership, general contractor, indefinite quantity, implementation, design-bid-build, machine learning, survey, bundling, complexity, project delivery, construction manager, integration, design build, program, indefinite delivery, agency, decision support, suitability |
| Subjects: | practitioner, data collection methods, design criteria, profession, sociology, software systems, financial and cost management, organizational analysis, systems engineering, contractual arrangements, contractual condition, decision analysis, statistical analysis, artificial intelligence, project delivery, payment |
| Topics: | Procurement, Cost Management, Roles and Professions, Risk Management, Digital Applications, Design Practice, Engineering Principles, Organizational Design, Contract Administration, Research Practice |
| Descriptive scope: | 5 PCTEA |
N.B. Descriptive scope is a count of how many of the five facets of empirical research are indicated by the words used in title, abstract and keywords. It is not intended as a judgement on the research; merely a count of the kind of word we would expect to indicate Phenomenon, Concepts, Theoretical framing, Empirical techniques, Analytical techniques. If all five are present, then a code of “5 PCTEA” will indicate this. If you feel the coding for this record is questionable, we welcome discussion around the terms we matched or the way we categorized them. The facet you would expect may not be coded, or a facet may be coded inappropriately. This can also bear on a larger question, of which facets should be treated as defining in construction management research. Please get in touch, and we will look at it. More details here