From the time he was a 10-year-old who wanted to know anything and everything about dinosaurs, Peter Madsen has loved learning. Now, as the recipient of a $2 million grant to research train traffic controllers’ use of algorithms, he shows no sign of slowing down. Even in his role as a professor at the BYU Marriott School of Business, he shares his love of learning with each student who steps foot in his class.
Madsen received a grant from The National Science Foundation to research how to best divide and balance transportation-related tasks between artificial intelligence (AI) and humans. Specifically, he’s looking at rail traffic controllers’ use of algorithms.
Madsen and the other researchers are now one year into their four-year and $2 million project. They are currently looking at which factors predict when a rail traffic controller will turn control over to an algorithm. These factors will help Madsen understand potential biases a controller might hold that would keep them from using the algorithm.
“These biases are suboptimal because in situations where the algorithm does a really good job, it can actually outperform humans because it never gets distracted,” explains Madsen.
The implications of this research project boil down to safety issues. He says, “Universally, people display a bias against trust in algorithms. The distrust humans show for AI causes potentially dangerous situations in transportation, because drivers tend to rely on humans in critical safety situations.”
Madsen has statistics to back his view that having an algorithm take over can save lives. “An algorithm called positive train control (PTC) has recently been implemented in the United States. The algorithm can take over for the train driver to slow the train down if the PTC system detects a likely collision or derailment,” explains Madsen. “We’re finding the use of the PTC system reduces collisions and derailments of passenger trains by about 90 percent.” Interestingly enough, the FrontRunner, a popular passenger train among students and professionals in Utah, was a part of this study.
While this is a complex project, Madsen doesn’t shy away from bringing his findings into the classroom. As a professor of organizational behavior and human resources at BYU Marriott, Madsen feels his research directly applies to the subject he teaches. His teaching focuses on how businesses and organizations implement change and how to help people accept organizational changes.
“I see a lot of similarities between my project and my class,” Madsen explains. “Change is happening in a lot of companies where AI is being adopted. Employees oftentimes have a hard time with that change because of a natural distrust of algorithms.” In class he frequently discusses his updated research because he knows firsthand it applies to what students might face in future employment.
One of Madsen’s main goals in this project is to “educate others about how to use these new technologies that are here and will continue to come.” For Madsen and his colleagues, the big picture is to find the balance so AI and humans work can together effectively, especially in areas where a mistake can lead to dangerous accidents.
Looking to the future, Madsen knows that people everywhere will face difficulties when it comes to accepting AI. He says, “I hope people will be excited about the potential of this technology, rather than fearful.”
Written by Maggie Olsen