Study: Random Boarding Likely Better for Reducing Virus Exposure

 - June 23, 2020, 4:46 PM
Several universities teamed on research that questions whether boarding from back-to-front is the most effective approach. (Photo: Embry-Riddle Aeronautical University)

This story is part of AIN's continuing coverage of the impact of the coronavirus on aviation.


While many airlines have adopted practices such as keeping middle seats open and boarding from back to front in response to the Covid-19 pandemic, preliminary academic research suggests that random boarding may be a better approach to limiting exposure rates.

Still in peer review, the research suggests that random boarding could reduce exposure rates by 50 percent, reported Embry-Riddle Aeronautical University (ERAU), which participated in the study alongside researchers from Florida State University, University of West Florida, and Arizona State University.

 “[The findings] suggest that airlines should either revert to their earlier boarding process or adopt the better random process,” said Ashok Srinivasan of University of West Florida, a co-author of the research paper.

An important aspect of this research is understanding how molecules travel, added Sirish Namilae, associate professor of aerospace engineering at ERAU and co-author to the study. Namilae previously researched how viruses spread in airport and airplane environments during the Ebola outbreak in 2017.

“While back-to-front boarding has been instituted by some airlines to try and reduce contact between people, our simulations show that high-density clusters can form as people stow their luggage while other passengers are still pushing toward the back of the aircraft,” Namilae said. “We hope that our research will prove useful to the airline industry in navigating the current health crisis.”

Using about 16,000 simulations, the study looked at “person-minute” measures of contact, or the number of minutes a person stands within six feet of someone during boarding. The simulations varied parameters such as walking speed, proximity, and the number of people, and tested several boarding patterns, such as front-to-back, back-to-front, six boarding zones, and random boarding through one zone.

“It turns out, the one-zone, random boarding model eventually results in a lower number of contacts,” Namilae said. “Other patterns tend to increase the time a passenger waits in close proximity to fellow travelers.”

Looking at an Airbus 320 at full capacity, the simulations recorded 43,000 person-minutes of contacts. This compares with 60,000 person-minutes for six-zone boarding, and 90,000 person-minutes for back-to-front boarding, ERAU said.

“Our analysis indicates that airlines that changed to a back-to-front boarding policy erred, exposing passengers to substantially higher infection risk than their original procedures,” said Srinivasan. “This result shows that good intention is not a substitute for good science when it comes to determining policies.”

Economics were considered throughout the simulations. For instance, other, more costly scenarios were evaluated and shown to be effective in significantly reducing risk. This included keeping middle seats unoccupied, which showed to be effective in significantly reducing infection risk. “The problem is, how long can airlines afford to fly with empty seats?” Namilae asked. Also effective is the use of more, but smaller aircraft to transport the same number of passengers as one larger aircraft.

But Namilae and his team recognized a middle ground must be found and believe the conversation should begin with the boarding process.

“There’s evidence of a lot of diseases, such as tuberculosis and SARS, being transmitted on airplanes and in air travel,” Namilae said. “Boarding is one of the critical aspects of air travel that contributes to the spread of diseases. Any steps we can take to reduce the spread would help.”

Tasvirul Islam of the University of West Florida served as lead author of the study. Other participants included Sadeghi Lahijani of Florida State University, and A. Mubayi and M. Scotch of Arizona State University. A National Science Foundation grant supported the research, which was conducted in part through the assistance of a supercomputer at Argonne Leadership Computing Facility, a U.S. Department of Energy Office of Science User Facility.