Level:
TCM Section(s):
Venue: 2023 AACE International Conference & Expo
Abstract: Loss of productivity, continues to be the most difficult matter to resolve in construction disputes. Practitioners and experts’ analysts in the construction industry alike acknowledge that the current methodologies used to estimate productivity loss rely on many assumptions (subjective) , which has led to ongoing controversies that are difficult to resolve. To enhance the current improved measured mile technique [3], a new model is proposed to illustrate and contrast the lost working hours due to disruption. Three methods will be used, including elbow method to determine the number of optimized clusters, the Gaussian mixture model (GMM) to softly assign each productivity data point to every cluster with its membership probability value, and probability theory to calculate the expected value of productivity (EVP) that must be linked to the best possible root causes. Furthermore, Python language classes were used to develop the model. This paper outlines an integrated approach for evaluating the accurate impact of construction productivity changes. To show the benefits and drawbacks of this improved procedure, we undertake two case studies, including a real-world construction claim.
TCM Section(s):
Venue: 2023 AACE International Conference & Expo
Abstract: Loss of productivity, continues to be the most difficult matter to resolve in construction disputes. Practitioners and experts’ analysts in the construction industry alike acknowledge that the current methodologies used to estimate productivity loss rely on many assumptions (subjective) , which has led to ongoing controversies that are difficult to resolve. To enhance the current improved measured mile technique [3], a new model is proposed to illustrate and contrast the lost working hours due to disruption. Three methods will be used, including elbow method to determine the number of optimized clusters, the Gaussian mixture model (GMM) to softly assign each productivity data point to every cluster with its membership probability value, and probability theory to calculate the expected value of productivity (EVP) that must be linked to the best possible root causes. Furthermore, Python language classes were used to develop the model. This paper outlines an integrated approach for evaluating the accurate impact of construction productivity changes. To show the benefits and drawbacks of this improved procedure, we undertake two case studies, including a real-world construction claim.