Shafaqna English- Researchers at Iran’s Amirkabir University of Technology have developed an artificial intelligence-based framework that predicts the likelihood of success of contractors’ claims in construction projects, potentially helping project managers and contract specialists make more informed decisions before pursuing costly disputes.
The framework, known as the Claimability Index, converts factors affecting the outcome of construction claims into a numerical model that uses machine learning to estimate the probability of success. Its developers say the approach could reduce reliance on individual experience and manual assessment of contractual documents.
The research was conducted by Mohsen Asgharnia, a graduate of construction management at Amirkabir University, under the supervision of Eghbal Shakeri, a faculty member at the university’s Civil Engineering Department.
Construction claims can lead to higher project costs, delays and contractual disputes, sometimes resulting in lengthy arbitration or legal proceedings. According to Asgharnia, decisions on whether to pursue or abandon such claims are often based largely on professional experience and manual document reviews, while companies lack a standardized, data-driven measure for assessing their prospects.
To develop the framework, the researchers reviewed scientific literature on claims management, contractual disputes and the use of AI in the construction industry. They then identified factors that can influence the success of a claim, including the type and value of the claim, duration of delays, contract type, quality of documentation and final claim outcomes.
Machine-learning models were subsequently trained using these variables, with their results translated into a simplified index designed for practical use by construction and contract professionals.
One of the main challenges was converting a largely qualitative issue involving technical, financial, managerial and legal considerations into variables that could be processed by an AI model while remaining understandable to industry specialists.
The resulting system produces a score between zero and one representing the estimated likelihood of a claim succeeding.
The framework could be used in construction, project and contract management, claims management and dispute resolution, with potential applications for contractors, employers, consultants, project management offices, claims teams, legal advisers and arbitration bodies.
The researchers also identified potential uses across building projects, road construction, transportation, oil and gas, energy, water and wastewater projects, and urban infrastructure.
A research paper based on the project, titled “Claimability Index: An AI Framework for Predicting Successful Construction Claims,” was published in the Q1 ASCE Journal of Legal Affairs and Dispute Resolution in Engineering and Construction, according to the researcher.
The project has also received a $2,000 research full-funding award for an in-person presentation at the ASCE 2027 conference in Philadelphia.
The initial stages of patent registration are also under way through Amirkabir University, while further development is focused on turning the framework into a technology product for the construction industry.
The system could eventually be developed as standalone software, a web-based platform or a module integrated with project management and document-management systems. The researchers also envision training customized models using individual companies’ historical data.
Source: Isna