Identifying the Opportunities and Challenges of Project Bundling: Modeling and Discovering Key Patterns Using Unsupervised Machine Learning

Abstract

Project bundling is a strategy that combines several infrastructure projects into a single contract to improve the overall performance of projects. While previous research efforts have been conducted on certain aspects of project bundling, no research particularly focused on studying the opportunities and challenges of project bundling and the associated patterns between them. To this end, this paper addresses this knowledge gap. Based on data from 30 case studies that implemented project bundling strategies in the US, various opportunities and challenges were extracted. In addition, spectral clustering was implemented to cluster the identified opportunities and challenges based on the strength of their interconnectivities. Also, association rules mining analysis was conducted to determine key patterns. The results identified a total of 27 opportunities and 27 challenges for project bundling. Furthermore, the most critical associations between the opportunities and challenges were determined within each of the obtained clusters. The outcomes also reflected that while many opportunities and challenges could individually affect the performance of bundled projects, other opportunities and challenges could also result due to a combination of factors that might not be perceived to be critical on the individual level but rather become critical when combined with other factors. This paper adds to the body of knowledge by helping project stakeholders in capitalizing on the opportunities of project bundling while also minimizing the associated challenges.