Application of Artificial Neural Networks in Delay Modeling of Sports Construction Projects

Background and Purpose
Today, sports are widely recognized as a valuable tool that can exert numerous positive effects on society. Research indicates that sports can be effectively utilized as a development tool, fostering social, economic, and health benefits. This recognition has motivated countries and societies to harness the potential of sports through strategic planning and policy-making to promote sports development and leverage its benefits at the community level. In Iran, numerous plans have been formulated to advance sports development. For instance, the comprehensive document on sports development in the country outlines two broad dimensions: soft approaches and hard approaches. Soft approaches include the development of inputs such as technology, communication, human resources, legal frameworks, financial resources, and planning management. Hard approaches focus on the development of physical infrastructure, including sports facilities, equipment, and related amenities (Hanifa, 2020; Savadi, 2017).
Among these dimensions, the development of infrastructure stands out as a critical indicator of sports progress. It is widely acknowledged that sports development in any society cannot be realized without the establishment and enhancement of fundamental sports infrastructure, including venues and facilities (Kumar et al., 2018). Construction and infrastructure projects are therefore considered one of the most vital components in the advancement of sports in Iran (Malahi Koohi et al., 2017). Given the pivotal role of construction projects and infrastructure in meeting the sports needs of the country, delays in the implementation of these projects can lead to significant challenges and hinder sports development (Abraham, 2019).
Timely execution of sports development plans represents a crucial opportunity for the growth of sports in the country. Conversely, delays transform this opportunity into a threat, impeding progress and diminishing potential benefits. At the time of this research, official statistics from relevant authorities indicated that nearly 4,000 sports-related projects in Iran were incomplete and suffering from implementation delays. This substantial number highlights a critical research gap and underscores the lack of scientific evidence regarding the factors contributing to these delays, their management, and strategies to mitigate them.
The present study aims to identify the factors influencing delays in sports construction projects and to develop a predictive model using artificial neural networks (ANN). The ultimate goal is to provide a scientific basis for timely intervention and reduction of delays in sports infrastructure projects, thereby facilitating the country’s sports development objectives.
 
Materials and Methods
The research employed a mixed-methods approach, integrating qualitative and quantitative methodologies. The qualitative phase focused on identifying delay factors in sports construction projects through expert interviews. Thirteen experts were purposively selected using snowball sampling. These experts included four university professors, four managers and engineers from the Development and Equipment Company responsible for sports venues, and five managers and technical supervisors from the Ministry of Sports and Youth’s sports plans and venues department. Semi-structured interviews were conducted, and data were analyzed using a phenomenological approach with open, axial, and selective coding techniques. This process yielded 51 open codes, which were further distilled into ten selective codes representing the main categories of delay factors: financial, engineering, estimation, infrastructural, human, legal, support, supervisory, natural, and managerial.
In the quantitative phase, the focus shifted to modeling delay prediction using artificial neural networks. A sample of 750 project managers, including contractors, executive engineers, executive directors, and civil assistants from the general administration of provinces, was selected. These individuals were associated with 30 sports construction projects, with approximately 25 respondents per project. Based on the 51 open codes identified qualitatively, a comprehensive questionnaire was developed and validated by sports management professors and experts. Reliability testing using Cronbach’s alpha confirmed the instrument’s internal consistency.
The collected quantitative data served as input variables for the ANN model, while the duration of project delay (measured in months) was the criterion variable. The model was developed and validated using the Lowe method, with MATLAB 2018 software employed for coding and analysis.
 
Findings
The study identified ten key factors affecting delays in sports construction projects: financial constraints, engineering challenges, inaccurate estimations, infrastructural deficiencies, human resource issues, legal obstacles, lack of support, supervisory weaknesses, natural factors, and managerial shortcomings. These factors collectively contribute to the complexity and prevalence of delays in sports infrastructure development.
The ANN model demonstrated robust predictive capability, with an average prediction accuracy of 88.34 percent based on the input indicators. This high level of accuracy indicates that the model effectively captures the multifaceted influences on project delays and can serve as a reliable tool for forecasting delay durations in sports construction projects.
The findings suggest that by employing this predictive model, stakeholders can anticipate potential delays early in the project lifecycle and implement corrective measures proactively. This capability is particularly valuable for managing resources, optimizing scheduling, and enhancing the overall efficiency of sports infrastructure development.
 
Conclusion
Given the critical role of sports construction projects in promoting sports development and delivering community benefits, it is imperative to address and mitigate the factors causing delays. The present study’s findings highlight the importance of recognizing and managing financial, technical, legal, and managerial challenges that impede timely project completion.
The proven efficacy of the ANN model in predicting delays based on identified factors offers a practical framework for early detection and intervention. By integrating this model into project management practices, decision-makers can better allocate resources, streamline processes, and reduce the incidence and impact of delays.
Furthermore, the study underscores the necessity of fostering a culture of accountability, transparency, and continuous improvement within sports construction management. This includes enhancing expertise among project managers, improving legal and supervisory frameworks, and ensuring adequate support and infrastructure to facilitate smooth project execution.
Ultimately, the application of advanced predictive analytics, combined with strategic planning and stakeholder collaboration, can significantly contribute to accelerating sports infrastructure development in Iran. This progress will enable the country to harness the full potential of sports as a driver of social, economic, and health benefits at the community level.