Join the Estimating & Cost Community of Practice as they deliver a webinar where they will explore the use of AI in estimating cost of major projects.  The webinar will be delivered by Paul Baguley & Richard Kirkham from University of Manchester and three Masters students: Minglei Chen, Maria Ndawula, Junqian Wang.

Dr Paul Baguley is a lecturer in project cost management at the University of Manchester and on the Masters courses in Management of Projects. Paul has worked for 26 years in applied research in cost engineering including over twenty-six companies and has been working with advanced cost modelling methods since 1999. Paul started his work in Artificial Intelligence (AI) as far back as 1999 with work in fuzzy logic and manufacturing process time modelling in aerospace and has since used advanced machine learning methods and now recently generative AI.

Dr Richard Kirkham is a Reader at the University of Manchester. Richard's research on cost estimating spans > 20 years and includes notable contributions to the discipline through his work on whole life-cycle costing. His interests in the applications of AI to cost estimating originate at The University of Liverpool's Construction Cost Engineering Research Unit where, during the late 1990's, he worked on the application of artificial neural networks and neuro-fuzzy methods to a range of estimating problems in construction projects and asset management.

Abstract

Artificial Intelligence (AI) includes generative AI, machine learning, and uncertainty modelling with quantitative, qualitative, image and various forms of data. UK gov set a framework in which to develop the use of AI in the management of major projects in the forthcoming years (https://www.gov.uk/government/publications/data-analytics-and-ai-in-government-project-delivery/data-analytics-and-ai-in-government-project-delivery). Part of this framework is to experiment with the use of AI.

Masters dissertation projects have been conducted at the University of Manchester in order to explore the use of AI in Major Projects. The main objectives of the three projects were to:

• Capture initial possible estimating requirements for AI from estimating for major infrastructure projects;
• Identify previous AI methods used in major projects estimating research and assess their success against AI type and availability of data;
• Develop a cost assurance checklist for use of AI methods in estimating major projects;
• Develop a machine learning model from publicly available data from major projects and assess its possible utility for future projects;

Systematic literature review methodology, use of public domain secondary cost data and a cost model development process were the main research methods used. The outputs of the research into a conceptual framework include:

• A range of early estimating stage cost drivers and also project delivery stage cost drivers;
• A range of secondary cost data sets for major projects;
• Initial set of estimating requirements for major projects;
• Review of the performance of AI methods in the research literature;
• A prototype assurance check list including explainability for cost estimating;
• A machine learning model developed with historical public domain data from a UK major project.

The presentation will explain an emerging exploratory conceptual framework and how it is currently being extended by further work in generative AI.

The link for the webinar will be provided upon registration.