Benchmarking is a fundamental tool for improving performance in the delivery of transport infrastructure projects. The sector has seen a significant boost in funding in recent years, and there is an urgent need to use benchmarking to inform decision-making processes so that projects benefit from collaboration and deliver future-proof infrastructure while ensuring value for money. But benchmarking cannot take place without suitable standardised data – particularly data for assessing all aspects of project costs. Unfortunately, as the TIES Living Lab data research team (the Analytical Consortium) acknowledged from the outset, there is a lack of consistency in the way costs are reported across the construction supply chain and client organisations, making it very difficult to implement robust comparison within organisations, or more widely across the sector and internationally. This paper describes a project to demonstrate the possibilities of using artificial intelligence (AI) to extract, transform and classify project cost data in a standardised way using “data mining”. The objective of the work, carried out under the project on Artificial Intelligence for Data Mining (and feeding in to the project on Metrics, Benchmarking & Repository) was to prove the concept in a “live” situation, taking data from a variety of sources and showing how AI can classify infrastructure project cost data into a common standard.
Mahdjoubi, L. (2022). IP5a standardisation and classification of project cost data. UK: Living Lab