Abstracts Computer Science

An analysis of the city of Austin pipe networks using network and information theory metrics

by Tess Marian Haegele

Austin’s rapid population growth over the past few decades has given rise to the need for additional water infrastructure and supply. There are limited funds for investment in water infrastructure so it should be spent with the goal of optimizing system robustness. A robust system comes from a balance of efficiency and redundancy. There are two methods used in this analysis to establish baseline metrics. Information Theory and Network Theory are based on the connectivity of the system looking at efficiency and redundancy. These theories are used by first converting the water pipe networks into a graph of nodes and links, extracting a connectivity matrix, and converting the data to “igraph” format in the statistical computing software R for analysis. The Network Theory calculations are built in to the “igraph” package in R and the Information Theory calculations are based on the equations developed by Robert Ulanowicz. The starting point metrics of this study can be replicated for the main and wastewater systems and built upon considering operational and hydraulic characteristics unique to the system in future work, and eventually inform utility decisions. Advisors/Committee Members: King, Carey Wayne, 1974 (advisor), Huber, Karen (advisor), Faust, Kasey (advisor).