AI assistance in pretrial scheduling could provide justice for all

The American court system, an essential pillar of American democracy, is burdened by a persistent problem in its pretrial scheduling process. A staggering one in five defendants fails to appear in court, leading to immense costs for the judiciary and detrimental consequences for defendants, particularly those facing economic insecurity, caregiving responsibilities or transportation limitations.

William Yeoh, associate professor of computer science & engineering in the McKelvey School of Engineering at Washington University in St. Louis, and collaborators at Syracuse University in New York received a $600,000 grant from the National Science Foundation (NSF) to reform pretrial scheduling to be fairer and more equitable with the help of artificial intelligence. Through integrating machine learning and optimization techniques, the team’s approach promises to enhance fairness and provide meaningful explanations for scheduling decisions, thereby mitigating the negative impacts associated with the current system.

“From a computational perspective, scheduling is a general problem that you often encounter, such as the scheduling of classes, flights, or anything else with a complex set of constraints and preferences,” Yeoh said. “My collaborator Ferdinando Fioretto and I are particularly interested in the topic of fairness and how AI can be used to promote social good and equity. After learning about the issues involved in pretrial scheduling, the inequities that come with it and the inefficiencies of the current scheduling system, we knew this is an area where new AI techniques could potentially have big positive impacts.”

Yeoh and his co-investigators, Fioretto, an assistant professor of electrical engineering and computer science, and Lauryn Gouldin, the Crandall Melvin Professor of Law and director of the Syracuse Civics Initiative, both at Syracuse University, will combine their expertise in AI, optimization, fairness and criminal justice reform with insights from domain experts – including people who actually create pretrial schedules – to build a fairer, more equitable and more efficient scheduling system.

“We’re starting in the New York court system because our team has contacts there, and we can begin by deploying our scheduling technique in a single district,” Yeoh said. “Improving court scheduling, even at a small scale, I already see that as a success, especially during the three-year period of this project. Looking ahead, longer-term goals include generalizing our technique to other court systems, finding more ways to improve efficiency in terms of the attendance rates of pretrial defendants, and expanding to include other forms of fairness that people in the scheduling area might not be aware of that they should take into account.”

Click on the topics below for more stories in those areas

• Research
• Computer Science & Engineering

Back to News