Consider your most recent experience at an airport. Did you notice the particulars of how your bags were screened, or were you more focused on taking off your shoes or collecting your laptop at the end of the conveyor belt? Most people would say the latter.
When Binghamton University faculty members Sarah Lam and Mohammad Khasawneh fly, it’s a different story. She likes to check out the X-ray images of her bag. He observes the baggage screener and where his or her eyes move.
Their interest stems from a research project that’s still in its infancy. It focuses on ways to improve the efficiency of baggage screening at airports, in part by making better use of computers to train X-ray operators. It has the potential to make baggage screening faster and more thorough, which would be good news for the Transportation Security Administration (TSA) as well as travelers.
“Baggage screening remains the last line of defense for air travelers against terrorists,” Lam said. “Therefore, a systems approach that considers human factors and intelligent training will be used to model the selection, training, certifying and monitoring of baggage screeners at airports.”
The project would integrate the components of airport security systems, including humans, equipment, machines and information systems, along with support from the physical and organizational environment.
Lam, who’s from Hong Kong, holds a doctorate from the University of Pittsburgh and came to Binghamton in 1999. She’s an associate professor of systems science and industrial engineering and specializes in system modeling, simulation optimization and neural networks.
Khasawneh, who’s originally from Jordan, came to Binghamton in 2003 after earning his doctorate from Clemson University. He’s an assistant professor of systems science and industrial engineering and he focuses on human factors in engineering and design.
They’re both in their 30s and see this project as one that could provide substantial direction for their careers.
“Since this research will lead to a greater understanding of aviation safety issues, we see this area as both challenging and rewarding,” Khasawneh said.
In the spring of 2004, Khasawneh and Lam began talking about a simulation that would encompass the flow of baggage and people through an airport. The idea was to bring a systems approach to an entire airport security system.
They quickly found that the scope of that project was too large for two people to tackle and decided to zero in on the baggage-screening portion of the process.
“Even with advances in screening technology, the performance of the human operator that uses these sophisticated machines remains a significant component in the entire security system,” Khasawneh said. “That’s why we decided to focus on the operators’ ability to detect threat objects such as knives or guns with an emphasis on intelligent training.” Lam and Khasawneh now hope to evaluate how much information human operators can absorb at once and determine the optimal way for the computer to present information.
Both expect to address issues related to the project in the classroom.
“This will provide a vehicle where the students can use what they learn in classes to solve real-world problems,” Lam said. “They will also get involved in research activities that will ultimately benefit not only themselves but also the general public.”
The baggage screening research is a natural, if unusual, blend of their different areas of expertise.
In graduate school, Khasawneh worked on projects that involved simulating aircraft inspections. The work, which was funded by NASA, resulted in more controlled, less expensive training sessions for Delta Airlines technicians. His dissertation focused on a “hybrid inspection simulator” – or a computer and human working together to inspect, in this case, printed circuit boards. For that work, he had to investigate the ways the computer could help the human with decision-making and how much and why the human operator would trust the computer.
“The baggage-screening environment represents a similar, but more complicated, hybrid system in which the operator and the computer work together to detect threat items,” Khasawneh said. “Even those advanced baggage screening machines are only as reliable as the human who operates them.”
Lam’s dissertation also provided a good foundation for this type of research. She looked at neural networks – artificial intelligence that was inspired by the way human brains work – and how they perform when asked to assist with decision-making and help solve complex problems.
“Incorporating artificial intelligence into the baggage-screening process will lead to better ‘hints’ from the computer to help human screeners make decisions about which items are unsafe,” Lam said.
Lam and Khasawneh say the next step is to line up at least $400,000 or $500,000 to get the project off the ground. Because the Government Accountability Office cited weaknesses in the TSA’s training for checked baggage screening in a recent report, they believe their work may attract federal funding from the Department of Homeland Security, the Federal Aviation Administration or the National Science Foundation.
They could also partner with a local company on research that would lead to a product that could be commercialized. “This research,” Lam said, “will lead to more secure airports for the flying public.”