September 9 – 11, 2018
The 2018 Annual Conference of ITS PA and MASITE was held in Pittsburgh, PA from September 9 – 11. Over 230 ITS professionals attended the 3 day event. The conference agenda included twelve sessions on various intelligent transportation topics.
Several UTC researchers, staff and students were featured during the conference:
Dr. Steve Smith spoke on smart infrastructure for future urban mobility and innovative real-time traffic controls. He discussed how real-time traffic signal control presents a difficult challenge in urban road networks where (unlike simpler arterial settings) there are competing dominant traffic flows that shift through the day. And that further complicating matters, urban environments require attention to multi-modal traffic flows (vehicles, pedestrians, bicyclists, buses) that move at different speeds and may be given different priorities.
He shared recent research that has produced an innovative new technology that uniquely addresses these challenges, by taking a fresh perspective and combining principles from artificial intelligence, multi-agent systems, and traffic theory. The initial deployment of this technology, called Surtrac, in the Pittsburgh East End produced significant performance improvements and the technology is now being installed in other US cities.
Dr. Smith explained that starting from this technology backbone, a broader future vision of smart transportation infrastructure where, as vehicles become more connected and more autonomous, the intersection increasingly becomes the gateway to real-time traffic information and navigation intelligence.
He explained that the current technology development efforts centering on additional use of direct vehicle- (and pedestrian-) to-infrastructure communication to further enhance mobility, on online analysis of traffic flow information for real-time incident detection, and on integrated optimization of signal control and route choice decisions are summarized.
Dr. John Paul Shen and Abhinav Jauhri spoke on synthetic data generation for modeling human mobility in urban areas. In their work, they demonstrate a novel technique to generate synthetic data sets of human mobility patterns for a number of large cities in the world. The explained that the availability of large datasets has been a long-standing challenge for the research community. Human mobility patterns are highly dynamic, for meaningful and reproducible research results the availability of representative datasets is essential.
Their work highlights methods to extract spatial and temporal properties which help characterize each city’s human mobility patterns. Such properties can be used to validate similarity between synthetic and real human mobility datasets. To capture human mobility, they specifically target at generating datasets similar to those accessible to ride-sharing services which have become ubiquitous in major cities in the world.
They explained that there are multiple motivations for generating such synthetic datasets. First, it demonstrates a non-trivial application of recent advances in Al techniques. Second, to generate synthetic dataset can be used in various what-if scenarios by civic authorities and researchers to provide insights on future patterns of urban mobility and traffic congestion to help advance planning by civic authorities. Researchers that do not have access to large real datasets can now use the synthetic datasets as benchmarks for research on traffic congestion, pre-placement of vehicles, passenger pooling and many others.
Lisa Kay Schweyer and Joohyun (Sarah) Cho provided the Carnegie Mellon University’s Traffic21 Institute rapid-fire research overview. Lisa Kay provided an overview of Traffic21 Institute, explaining it is a multi-disciplinary research institute of Carnegie Mellon University and its goal is to design, test, deploy and evaluate information and communications technology based solutions to address the real-world problems facing the transportation system of the Pittsburgh region and the nation and that the Pittsburgh region serves as a “learning lab,” deploying solutions that can be applied around the nation and the globe.
She further explained that Traffic21 accomplishes this through a comprehensive program of interdisciplinary research; education and workforce development with a focus on diversity; collaboration with university, deployment, and government partners; and technology transfer and leadership efforts.
Sarah then provided a rapid-fire overview of some of the Traffic21 research currently being conducted and the new technologies being developed.
In addition, the University Transportation Center sponsored 7 students so they could attend the conference and learn from the conference presentations.