Computer Science Team's Summer Research Helps Improve Passenger Experience and Safety at Major Transportation Hubs
Before they arrived at their summer research facility at Rutgers University, Vishnu Nair, Greg Olmschenk, and Zhigang Zhu, Ph.D., were well aware of the frenetic scrambling of rush hour commuters at subway and train platforms across the country. Their research project sought to reduce not only the added stress and challenges faced by tourists or individuals with disabilities but also the safety risks inherent in these noisy, densely populated stations.
“Our research was focused on integrating three things for travel assistance in a transportation facility: crowd analysis using surveillance cameras and deep learning-based machine learning techniques; user localization using emerging Bluetooth beacon technology; and 3-D modeling using advanced laser range scanners,” said Zhu, professor of computer science at The City College of New York (CCNY) and director of the CCNY Visual Computing Lab. “Everyone who encounters challenges in a complicated transportation center can benefit.”
Nair, Olmschenk and Zhu conducted their research at the Command, Control and Interoperability Center for Advanced Data Analysis at Rutgers through the U.S. Department of Homeland Security (DHS) Summer Research Team (SRT) Program. The program aims to increase and enhance the scientific leadership at Minority-Serving Institutions (MSIs) in research areas that support the mission and goals of DHS.
This program provides faculty and student research teams the opportunity to conduct real-world, meaningful research at the university-based DHS Centers of Excellence (DHS Centers). The SRT Program and DHS Centers are sponsored by the DHS Science and Technology Directorate Office of University Programs.
Ultimately, the team envisioned transforming transportation facilities into “smart transportation hubs” linked by a smartphone app. The app could be used like an indoor GPS to help prevent the pervasive issue of first-time travelers getting “lost” or “turned around” at large, complex facilities in cites like New York City or D.C. This would ensure a smoother flow of traffic, which inherently improves the safety and security for everyone. The technology could also serve those with visual impairments through voice command or large, easy-to-decipher text and graphics on-screen.
Nair worked on this part of the project, which used small Bluetooth “beacons” rather than GPS satellites to provide location-based services.
“Before this project, I had no idea there were so many researchers and startups looking to create their own versions of ‘indoor GPS.’ It was exciting to come up with a preliminary, yet working, version of my own,” said Nair, an honors student in computer engineering at CCNY.
Olmschenk, a doctoral candidate in computer science at The City University of New York (CUNY) Graduate Center and a member of the CCNY Visual Computing Lab, instead explored crowd surveillance and analysis. He created a computer program to automatically count the number of people in a surveillance camera image through facial recognition, which required ‘training’ the program to recognize what humans looked like, even in partial view.
“By the end of the summer, we successfully produced charts of pedestrian traffic for a given location over the course of the day,” he said, adding the next layer of research involves training the program to analyze the crowd’s behaviors to identify potential security concerns. “In general, the ability to count how many people pass through different areas of the building allows the facility to improve security and efficiency.”
Although each member of the team was responsible for distinct parts of the project, cross collaboration was critical to shared success.
“One of the most important experiences in the program is learning to working with a larger group,” said Olmschenk. “Everyone has great ideas, but the project is usually too large for one person to do alone. The challenge is in getting each person’s effort to connect and work together toward a greater system.”
Zhu, the team lead, selected both Olmschenk and Nair because of their high levels of motivation and creativity. Both students were quick to pick up on new technologies, Zhu added.
“My favorite part of the program is the collaborative environment. I actually have learned a lot about emerging technologies in deep machine learning and mobile computing from my team members,” said Zhu, explaining that the program provides a platform to develop very close collaborations with other researchers across disciplines. “Yes, a professor can learn a lot from students!”
The team, which was mentored by Jie Gong, Ph.D., assistant professor in the Department of Civil & Environmental Engineering, submitted their research to the Transportation Research Board (TRB) 96th Annual Meeting as well as to the Institute of Electrical and Electronics Engineers Winter Conference on Applications of Computer Vision in 2017. The team was also invited to present their research in the TRB annual meeting, a flagship conference in the field of transportation engineering.
Overall, Zhu believes the DHS SRT MSI Program provides an ideal platform to infuse research into the larger context of government, industry, academia and end users.
“I would consider leading another summer research team and would also recommend the program to others,” said Zhu. “This is an excellent research experience.”
The DHS SRT MSI Program is funded by DHS and administered through the U.S. Department of Energy’s (DOE) Oak Ridge Institute for Science and Education (ORISE) through an interagency agreement between DOE and DHS. ORISE is managed by ORAU for DOE.