A Web-Based Personal Driving Assistant Using Real-Time Data and a Dynamic Programming Model
Alamdar Yazdi, Mohammad Ali
Type of DegreePhD Dissertation
DepartmentIndustrial and Systems Engineering
Restriction TypeAuburn University Users
MetadataShow full item record
Eco-driving and eco-routing have recently attracted paramount attention from both academia and transportation industry. Multiple studies have approached this problem from a variety of angles. The goal of the first two parts of this dissertation is to leverage the abundance of freely available real-time data and the advanced computer technology to solve this problem. To this aim, several types of real-time information are required. Unfortunately, there is no comprehensive data resource that contains in one place all these different and independent information. To overcome this hurdle, three data sources of GoogleMaps, OpenStreetMap, and OpenWeatherMap are deployed in this dissertation, from which multiple types of information such as road, weather, and elevations information are data-mined and extracted. The extracted information is then passed to a dynamic programming model which is implemented in Python. The dynamic programming model is constructed to optimize the speed, gear, and route based on forthcoming route information data-mined from different web APIs and online databases. The result of this part of this dissertation is an eco-driving and eco-routing real-time web application. Nowadays, due to the advances in the technology of GPS systems, wearable and portable devices such as smartwatches and smartphones, storing the locations data of a moving entity has become easy and common. GPS locations databases have a large size due to the high sampling frequency, contain a lot of conditionally redundant data and therefore are not well suited for further optimization and statistical analysis. The third goal of this dissertation is to develop an R-based web application that takes in a GPS locations raw database along with two databases of Incidents and Accidents, if they are available, and automates the following: construction of a new database that resolves the aforementioned issues of the original database; generation of interactive visualizations that illustrate different statistics of the trip-based database; enriching the constructed trip-based database with the weather conditions (data-mined from National Oceanic and Atmospheric Administration) that a specific trip encountered; and enriching the newly created trip-based database with the information of Accidents and Incidents that occurred during a specific trip.