This summer, co-PI Zuefle spent six weeks of his unpaid summer to full-time supervise excellent high-school students in a research summer internship at GMU.
Three of the five student were working on project related research topics. Their research posters can be found below:
Erik Lin: Poster
Abstract: In this work, we propose a novel approach to forecast the metro network flow of passengers, which is exceptionally useful for city planning. We propose a micro-prediction approach to predict the number of passengers exiting from a station using farecard data. Our model’s network flow predictions can provide extremely valuable insight in deciding where and when to add new trains and stations.
Katie Zhang: Poster
Abstract: The Washington DC Metro frequently experiences delays and inefficiencies due to congested areas, especially during the morning and evening rush hours. In this work, we develop a data-driven system to automatically identify and monitor congestions by interpolating passenger trajectories of the Washington DC metro network. Our system supports urban planning by indicating times and locations where more resources need to be allocated.
Jack Snowdon: Poster
Abstract: Prediction of road traffic volume is paramount in many applications such as route planning, traffic prediction, and smart urban planning. Existing work focuses on estimating traffic speeds using GPS data, but is not able to estimat the actual traffic volume. Using GPS data provided by TomTom, and linking it to publicly available traffic loop data taken from Northern Virginia, we build a model that allows to estimate the coverage of GPS data at any time and location, thus providing a robust way to estimate traffic volume in locations where no traffic loops are installed.