Yazar "Tokdemir, Onur Behzat" seçeneğine göre listele

Listeleniyor 1 - 4 / 4
  • Küçük Resim
    Öğe
    Predicting Accident Outcomes in Cross-Border Pipeline Construction Projects Using Machine Learning Algorithms
    (Springer Heidelberg, 2023) Mammadov, Ahmad; Kazar, Gökhan; Koç, Kerim; Tokdemir, Onur Behzat
    Pipeline construction projects are necessary to provide gas and liquid energy transportation. Although various studies have investigated the contributing factors to accidents occurring in pipeline construction projects, there is a need for a predictive model for such incidents. The main purpose of this study is to provide an artificial intelligence-based model to predict the outcomes of occupational accidents. In this context, 1184 incident cases, including injury, near-miss, and asset-product damage, taken from a pipeline construction project constituted the primary dataset. Twelve prediction models are formed by changing the input domain according to the type of incident, time, and cause type attributes, leading to 12 distinct sub-datasets. Then, each dataset is tested with 11 different machine learning (ML) algorithms to derive an effective prescriptive model. The descriptive results show that low awareness of job hazards and improper vehicle operations were the most critical immediate causes, while failure in risk recognition and site supervision were the major root causes of pipeline construction accidents. Among the ML methods, the deep learning algorithm performed better than its counterparts in eight sub-datasets. Finally, a prescriptive model incorporating the ML application procedure is recommended for construction companies to reduce occupational accidents. Overall, the proposed model and findings are expected to contribute to preventing and reducing construction accidents in pipeline projects by adopting relevant strategies.
  • Küçük Resim
    Öğe
    Predicting the Cost Outcome of Construction Quality Problems Using Case-Based Reasoning (CBR)
    (MDPI, ST ALBAN-ANLAGE 66, CH-4052 BASEL, SWITZERLAND, 2022) Doğan, Neşet Berkay; Ayhan, Bilal Umut; Kazar, Gökhan; Saygılı, Murathan; Ayözen, Yunus Emre; Tokdemir, Onur Behzat
    Quality problems are crucial in construction projects since poor quality might lead to delays, low productivity, and cost overruns. In case preventive actions are absent, a lack of quality results in a chain of problems. As a solution, this study deals with non-conformities proactively by adopting an AI-based predictive model approach. The main objective of this study is to provide an automated solution structured on the data recording system for the adverse impacts of construction quality failures. For this purpose, we collected 2527 non-conformance reports from 59 diverse construction projects to develop a predictive model regarding the cost impact of the quality problems. The first of three stages forming the backbone of the study determines crucial attributes linked to quality problems through a literature survey and the Delphi method. Secondly, the Analytical Hierarchy Process (AHP) and a Genetic Algorithm (GA) were used to determine the attribute weights. In the final stage, we developed models to predict the cost impacts of non-conformities, using Casebased Reasoning (CBR). We made a comparison between the developed models to select the most precise one. The results show that the performance of CBR-GA using an automated weighting model is slightly better than CBR-AHP based on a subjective weighting system, whereas the case is the opposite in standard deviation in forecasting the cost outcome of the quality failures. Using both automated and expert systems, the study forecasts the cost impact of failures and reveals the factors linked to poor record-keeping. Ultimately, we concluded that the outcome of non-conformities can be predicted and prevented using past events via the developed AI-based predictive model.
  • Küçük Resim
    Öğe
    Prioritization of risk mitigation strategies for contact with sharp object accidents using hybrid bow-tie approach
    (Elsevier, 2023) Kuzucuoğlu, Dilsen; Koç, Kerim; Kazar, Gökhan; Tokdemir, Onur Behzat
    Occupational accidents are still a priority for the construction industry and safety professionals. Appropriate and proactive risk management strategies for complex construction safety management issues are essential to transform the if-it-breaks-fix-it approach into fix-it-so-as-not-to-break. This study introduces a novel hybrid bow-tie (H-BT) model to identify the riskiest causal paths and the most effective protective and preventive strategies. A dataset comprising 829 contact-with-sharp-object (CwSO) injury cases collected from 63 projects carried out by three companies is used to define the pairwise relationships among risk factors and define the risk paths using a mutual information matrix. A Delphi analysis is conducted to achieve the final strategies and calculate the risk reduction effect (& UDelta;RR) and cost of each strategy. Finally, protective and preventive strategies are prioritized according to the & UDelta;RR/cost ratio. The findings show that to reduce the number and severity of CwSO accidents, stop working authority and staffing for safety are the most effective preventative and protective strategies, respectively. Besides, inadequate supervision/management leadership is found to be the originating risk factor in many of the risk paths. This study incorporates mutual information between risk factors, BT, and preventive and protective barriers in the same, easy-to-understand visual model to examine a particular injury type, which is the primary contribution to theory. Overall, the risk assessment model developed is ex-pected to improve the effectiveness of safety management applications in construction sites and eventually contribute to the minimization of incidents in construction projects.
  • [ N/A ]
    Öğe
    Quality Failures-Based Critical Cost Impact Factors: Logistic Regression Analysis
    (Asce-Amer Soc Civil Engineers, 2022) Kazar, Gökhan; Doğan, Neset Berkay; Ayhan, Bilal Umut; Tokdemir, Onur Behzat
    Due to the various dynamic conditions of construction sites, quality failures have become part and parcel of the industry. Many studies have identified the causal factors of construction site quality failures and their cost impacts. However, limited studies have been made evaluating the domino effects of these on one another and the correlation between the cost impact and frequency of each attribute. In this study, made in the context of ongoing research into related artificial intelligence (AI)-based predictive models, a total of 2,527 nonconformance reports (NCRs) collected from 59 construction projects within the scope of a previous study were analyzed using the Delphi method and logistic regression analysis. According to the Delphi results, 25 critical cost impact factors were refined and categorized into five main groups: Materials, Design, Installation, Operation, and Process. Then, five main hypotheses were developed to test each attribute's cost impact and interaction by logistic regression. The results showed that although some attributes (from the Materials and Operation groups) have a significant impact on the cost of quality if observed in a failure report individually, others may become a critical cost-impact factor when interacting with other attributes. No significant correlation was observed between the frequency and cost impact of the attributes. Finally, a holistically based quality control system that considers the domino effects of causal factors from planning to operation was proposed for construction practitioners to reduce quality failures causing cost and time overruns.