I-Corps: Translation Potential of Automated Analysis of Construction Contracts for Risk Identification Using National Language Processing and Deep Learning

Purdue University, West Lafayette IN

Investigators

Abstract

The broader impact of this I-Corps project is the development of a secure and automated construction contract risk identification technology that combines and leverages the state-of-the-art natural language processing, deep learning, and rule-based techniques to precisely extract risk from construction contracts. About 70% of construction projects end up in claims and disputes due to an improper understanding of the project requirements and conditions specified in contract documents and specifications. These claims result in delays and budget overruns, with millions of dollars spent on conflict resolution. Traditional contract review and risk assessment require deep expertise, extensive time, and manual efforts. The technology used in this project aims to improve the contract review process by automating the extraction of risks from construction contracts. The technology has the potential to save 60% to 70% of the costs associated with the traditional contract review processes and potentially reduce project cost overruns by 10-15%. This innovation will help project teams and stakeholders promptly address risks before construction begins, directly contributing to project successes, boosting commercial competitiveness of the construction industry through data-driven decision-making, and seizing emerging opportunities. This I-Corps project utilizes experiential learning coupled with a first-hand investigation of the industry ecosystem to assess the translation potential of the technology. This solution is based on the development of an innovative technology that harnesses the power of artificial intelligence to improve the analysis of construction contracts. At the core of the technology is a sophisticated neural network architecture designed and trained with state-of-the-art semantic Natural Language Processing (NLP) techniques and Large Language Models to comprehend the intricate language. The approach also considers multiple project specific parameters like stakeholders, demographics, geometry, and geography of the project, as they pertain to the specific situations of the site. This additional context enables early prediction of the potential risks before the construction starts. An NLP–based Automated Building Code Compliance Checking Framework serves as the technology foundation of this project. Moreover, the integration of advanced deep neural architecture, machine learning, and rule-based techniques enables the identification of complex risk patterns and trends that may not be readily apparent through traditional methods while maintaining the trustworthiness of the results. This ability enhances foresight and enables proactive risk management. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.