PhD Oral Exam - Elie Otayek, Building Engineering

When studying for a doctoral degree (PhD), candidates submit a thesis that provides a critical review of the current state of knowledge of the thesis subject as well as the student’s own contributions to the subject. The distinguishing criterion of doctoral graduate research is a significant and original contribution to knowledge.

Once accepted, the candidate presents the thesis orally. This oral exam is open to the public.

Abstract

Most of the Researchers’ interest is focusing on Bridge Management System (BMS), by its Maintenance, Repair & Rehabilitation (MR&R) segments, while a good practice of bridge management should cover the conceptual design phase of a new bridge. Although conceptual design is one stage of a bridge life whereas there is still limited knowledge on how it should be conducted. Most often engineers’ decisions are based on their past experience and standard solutions, which are affected by the subjectivity. This approach, most probably, does not reflect the effective way to address different problems of consistency in the decision. To design a new bridge, many factors, like the cost and aesthetic behavior, have to be considered due to their ability to affect the final decision. Generally, decision-makers will make their final design and decisions based on those factors in addition to human subjectivity.

The objective of this research is to propose a methodology to develop systematic procedures that can help decision-makers select the most appropriate bridge type with its diverse components and to forecast its Life-Cycle Cost (LCC) and other characteristics such as the public satisfaction and environment sustainability of the selected bridge type. The proposed methodology integrates a decision support system with a relevant database within an artificial intelligence environment (AI) and bridge information management tools in order to reduce the impact of human subjectivity on the decision taken during the conceptual design stage of bridges life.

For this purpose, Machine Learning technique, which is a part of the Artificial Intelligence (AI) environment, will be the core of the proposed methodology. The Artificial Neural Network (ANN) with its back-propagation algorithm is adopted in order to identify the appropriate solution by setting up its engine guidelines. A polynomial regression method will first be integrated in the DSS: 1) to formulate the database in order to establish the needed functions; 2) to define the input values for the ANN; and 3) to verify the intermediate ANN weightiness of the proposed methodology engine based on historic data retrieved from the database. An investigation will be conducted in order to define bridge characteristics and parameters, to be used in the ANN, and that affect the final decision.

Elements of the ANN layers (Engine model) which include: Input, Hidden & output layers, have to be described based on a systematic and standardized process. The proposed methodology has the potential to be used in lower levels to determine other bridge components such as vertical structures, foundations and connection types. The objectives of the proposed methodology can be summarized as follow:

1. Avoiding, as much as possible, the influence of human subjectivity on the decision-making process.
2. Listing and Ranking all the other possible alternatives.
3. Ensuring equivalent and fair consideration for all factors affecting the decision, and especially reducing the possibility to miss and overlook the impact of some factors that could be ignored while proceeding with making the right decision by using conventional decision-making approaches.
4. Developing a model that can be considered as a guideline for further use that can be applied within any decision-making environment and based on a relevant historic database and experts’ input.