![]() | Only 14 pages are availabe for public view |
Abstract Productivity of construction engineers’ staff represents a significant concern to all construction organizations. The optimization of the production of the total organization can be achieved through the coordinated improvement of the productivity of the engineers’ staff. As the most valuable asset of any organization is the productive workforce. If the engineers’ staff is not satisfied with their jobs and not motivated to fulfill their tasks and achieve their goals, the organization cannot attain success. Productivity is directly linked to motivation. Nevertheless, just a few construction organizations consider the human capital as being their main asset, capable of leading them to success. So, it is necessary to exert efforts toward improving productivity by heads of organizations in recent times. As Egypt is seeking to promote a leading role in the international community and to overcome most of the problems affecting its development, construction productivity is of high concern. Construction engineers’ staff productivity is affected by many factors. Motivation is one of these factors. Therefore, the first objective of this study focuses on identifying the most significant factors affecting engineers’ staff productivity in Egypt in order to find the root solution to motivate and improve construction engineers’ staff productivity. The second objective is the development of an Artificial Neural Network model that assists any construction organization’s decision maker to achieve a reliable assessment of the expected construction engineers’ staff productivity improvement. This thesis is mainly belonging to the engineers in construction industry in the Greater Cairo region in Egypt. A review for the previous research results in line with this thesis was performed. Several factors which affect III productivity are collected from previous studies and construction experts. A questionnaire survey among construction experts in the Egyptian market was carried out. Twelve out of 59 factors were identified as the most significant factors that affect construction engineers’ staff productivity. from construction projects in Egypt 123 records for the identified factors were collected. Projects and HR managers at construction field were provided feedback for the records. An Artificial Neural Network model was prepared to predict the percentage of construction engineers’ staff productivity improvement in Egypt. Educational version for MATLAB Neural Network (Matlab 2014b) software using feed forward network, and the back propagation algorithm was used to develop the model. A total of 123 records were collected and had been divided into 3 parts which are training, validating, and testing set as follow; 80 (65%), 22 (18%), and 21 (17%) for training, validating, and testing of the network respectively. Several trials had been carried out for determining the best model. Therefore, 60 trials with different network structures were experimented. The selected model for this problem is that with the least error for training, validating, and test set and with the large number of neurons in the hidden layer. This model consists of an input layer with seventeen (17) input nodes, and one hidden layer with twenty four (24) neurons, and one output layer with three (3) output nodes. Sigmoid activation function was used, and learning rate is (0.1), and no. of iteration is (500). The results of the testing process gave 14% test error. This demonstrates the viability of the model as a powerful tool. Model can help decision-maker in any construction organization to arrive at a reliable assessment of the expected construction engineers’ staff productivity improvement in any construction project in Egypt. |