Date of Award
Doctor of Philosophy
Civil, Construction, and Environmental Engineering
Civil, Constructionand Environmental Engineering
As the per capita cost of congestion rises every year, transportation agencies across the world are facing numerous challenges to handle such situations. With two-thirds of the distance being traveled at or near signalized intersections, maintaining them can suffice a great benefit to the system. This thesis presents a series of works aimed to identify locations which are controlled by intersections in the traffic network where there is some kind of issues. The first study presents anomaly detection to find out atypical behavior in traffic flow. After that we propose five performance measures which can be incorporated to identify problematic locations of an urban arterial corridor or a number of corridors. The demonstrated methodology can be applied to identify problematic segments in the future. The tool can assist in identifying locations where delay is high, day-to-day traffic patterns are dynamic, or the minute-to-minute demands at signalized intersections are highly variable. Using this methodology, agencies can come up with their own thresholds or they can directly use these thresholds to determine the performance of the different segments. The next study uses the serverless, cloud computing systems of Google Cloud Services to enhance and reuse some ATSPMs to evaluate the impact of Covid-19 pandemic lockdown on traffic signals operation across the entire state of Utah. In this study we determine the various changes taken by Utah Department of Transportation, its impact on the traffic signals' operations, and the factors that can further improve the operations of the signalized intersections. Using these measures, 51.6\% of agencies that could (or did) not implement the changes would get an idea about how to implement changes in traffic signals timing successfully. The final study is the evaluation of Adaptive Control System which seems to be a trending topic in transportation over the last few decades. Here the performance is evaluated using numerous data sources on two arterial corridors of Omaha. Study is conducted at numerous stages of lockdown and the before-after condition of ASCT. Final conclusions are drawn based upon delay per vehicle performance and travel time. With the availability of huge volume of data and robust computational systems both offline as well as on the web, agencies can choose their Measures of Effectiveness (MOEs) based upon the availability of funds and resources. Once these are determined, the agencies can apply some of the MOEs discussed here and identify the problematic locations within their system.
Poddar, Subhadipto, "Operational performance evaluation of traffic signals using big data analytics" (2020). Graduate Theses and Dissertations. 18378.
Available for download on Saturday, January 07, 2023