This thesis proposes a novel working paradigm for transportation infrastructure asset management by viewing the transportation networks as key components (or nodes) of a broader network of resources, which includes infrastructure linked with society's ecological, social, and economic systems. An extensive review of network science literature suggested that to understand the behavior of a complex network is imperative to characterize its topology. Consequently, this thesis focused on developing a framework to characterize the topology of the transportation infrastructure systems, and understanding how the unveiling topology can be used for supporting transportation asset management decisions. The proposed methodology determines whether the transportation infrastructure networks can be modeled as scale-free or exponential networks, using a framework for characterizing the agents of the network, their direct and indirect interactions among each other, and their importance as elements of a complex network, and utilizes these data to support transportation asset management. The methodology consist of seven steps: (1) define the networks of interest; (2) identify their intrinsic components; (3) visualize the identified networks using GIS maps; (4) identify direct and indirect interactions through superposition of the networks; (5) represent the relationship between the nodes and their linkages by frequency diagrams in order to determine the intrinsic topology of the network; (6) illustrate (graphically) the overall transportation infrastructure with the help of GIS; and (7) analyze the TINs from the decision-maker point of view, identifying the elements that are more relevant or need more attention on the network. The procedure is then implemented in a small network in a localized area (Town of Blacksburg, Virginia) to show its practicality, and recommendations for further development and mathematical modeling in order to allow its implementation in larger networks are provided. Based on frequency analysis of the nodes and their connectivity, it was concluded that the transportation infrastructure networks in the case study behave as exponential networks. The study showed that the links determine how the infrastructure network grows and that problems like congestion can be addressed by analyzing other factors related with topology, such as speed, unit size, and lane width. The proposed methodology was found to be useful as an asset management tool. Finally, a list of findings and recommendations for further research are presented as opportunities to enhance the management of transportation infrastructure networks. / Master of Science

Thesis (M.A.)--University of Hong Kong, 2001. / Includes bibliographical references (leaves 112-113).

Thesis (M.A.)--University of Hong Kong, 2001. / Includes bibliographical references (leaves 112-113). Also available in print.

Thesis (M. Soc. Sc.)--University of Hong Kong, 1984.

Thesis (Ph. D.)--Georgia Institute of Technology, 2007. / Includes bibliographical references (leaves 129-137).

Thesis (M.Soc.Sc.)--University of Hong Kong, 1984. / Also available in print.

Thesis (Ph. D.)--University of Texas at Austin, 2009. / Title from PDF title page (University of Texas Digital Repository, viewed on Sept. 17, 2009). Vita. Includes bibliographical references.

No / Transportation infrastructure projects are prone to cost and schedule overruns. At the time of contract award, a construction contingency budget is often used to accommodate for unplanned events such as scope changes. Recent empirical research has shown that rework during construction as a result of design changes, errors and omission are the major contributors of overruns in projects. The statistical characteristics of rework, and cost and schedule overruns that are experienced from a project's contract award for 58 Australian transportation infrastructure projects are analysed. Theoretical probability distributions are fitted to the rework, cost and schedule overrun data. Goodness of fit tests are used in conjunction with probability-probability (P-P) plots to compare the sample distribution from the known theoretical distribution. A Generalised Logistic probability density function is found to describe the behaviour of cost-overruns and provides the best overall distribution fit. The best fitting distribution for schedule overruns and rework data were the Four Parameter Burr and a Johnson SB distribution, respectively. The distributions are used to calculate the probability of rework, cost and schedule overruns being experienced. A case illustration is presented and discussed to demonstrate how the derived probabilities may be utilised in practice.

This study attempts to develop realistic and relevant investment planning models for inland container transportation systems. An inland container transport system model has been constructed consisting of three sub-models: forecasting future total export container demand, the inland container traffic allocation model and the optimum port capacity model. The models may be utilised to identify the most effective investment plan for inland transportation infrastructure development and to evaluate the inland container transportation system. The procedure enables determination of the optimal locations, sizes and time of container port developments as well as the optimal container cargo flows through transportation networks. A Heuristic algorithm was developed for the purpose of evaluating alternative investment plans. Dynamic and Linear programming methods are applied to each of the two planning problems: the former for the optimum container port capacity development problem and the latter for the optimal allocation of inland container traffic movements. Finally, the model has been applied to concrete inland container transportation system problems in Korea. The results are reported and analysed. It is hoped that they may provide a guideline for actual development.

This thesis provides an assessment of the effectiveness of micro unmanned aerial vehicles (MUAVs) as a tool for collecting condition data for transportation infrastructure based on multiple field experiments. The primary experiment entails performing a level of service (LOS) condition assessment on multiple roadside sample units at various locations across the state of Texas. A secondary field experiment entails performing a pavement condition index (PCI) survey on airfield pavements. The condition of these sample units were assessed twice: onsite (i.e., ground truth) and by observing digital images (still and video) collected via a MUAV. The results of these surveys are then analyzed to determine if there are statistically significant differences in the standard deviation and mean values of the condition ratings. This study shows that in favorable site conditions, the MUAV demonstrates promise for improving current roadway inspection methods. However, limitations of the MUAVs field performance show that there is need for improvement in this technology before it can be implemented.