Lehigh engineering professors and students worked with the Pennsylvania Department of Corrections to create an algorithm that saves Pennsylvania taxpayers $3 million per year.
The project started five years ago when Dan Li, ’11G, a second-year doctorate student, interned with the Pennsylvania Department of Corrections. Li started the project after her internship to create a more efficient way of assigning inmates to correctional facilities.
The systematic approach takes into account various factors of optimization to create an algorithm, which with a few keystrokes and a couple of seconds does the work that used to take seven men one week.
Since then, the algorithm has been improved by three professors and four graduate and doctoral students. The algorithm will be completed in the spring of 2018, though it has been in operation since December 2016 in Pennsylvania’s Camp Hill correctional institution.
The algorithm will be recognized by the Pennsylvania State Senate on Oct. 17.
Tamas Terlaky is the former department chair of the Industrial and Systems Engineering department. Terlaky, along with other Lehigh professors and students, partnered with the Pennsylvania Department of Corrections to develop an algorithm that helps with assigning inmates to correctional facilities. (Courtesy of Tamas Terlaky)
“We didn’t know from the beginning if we could solve it, but we did know that optimization had never been used in a correctional facility before,” said Tamas Terlaky, the former department chair of Lehigh’s industrial and systems engineering department. Terlaky took over from professors Lou Plebani and George Wilson, who were both present during the algorithm’s creation.
After Li obtained her doctorate degree in 2013, doctoral candidate Mohammad Shahabsafa followed in Li’s footsteps.
Shahabsafa currently owns the project and has spent the last three and a half years completing the algorithm. In the process, Shahabsafa and his colleagues were contenders for the Wagner Prize, which is awarded annually by the Institute for Operations for Research and the Management Sciences.
The Wagner Prize contenders are nominated through a three-part application. It is one of the most lauded international prizes in optimization engineering, and winners have access to the most sought-after engineering positions.
“It would be an honor to get the prize…(but) the experience of being a part of the creation of the algorithm is more than enough,” Shahabsafa said, “as I will use the skills in the future that I have used in the process of working on the algorithm.”
Graduate students Anshul Sharma and Chaitanya Gudapati created the algorithm’s graphic interface that accesses all the inmate information in the Department of Corrections’ system. It then allows the user to assess and recommend inmate assignments based on gender, crime, duration of stay and past behavior, among others variables.
Shahabsafa said the algorithm had a three-step creation process. The first step was studying the inmate assignment process. The second step was creating a mathematical model that actively fixed the problem. The third step was the longest step, and it tested the mathematical model and refined the algorithm several times until all the bugs were worked out.
Bill Nicklow, the population management and sentence computation director for the Department of Corrections, said he had nothing but praise for Shahabsafa and the rest of the team that completed the algorithm.
“Mohammad’s response to problems and the overall turnaround time on (the algorithm) was tremendous,” Nicklow said. “It was a good cohesive partnership, and I can’t say enough good things about Lehigh’s faculty and staff.”
The algorithm can only be used in Pennsylvania because the variables and information that it processes are specific to the state and facility that the algorithm operates. Future algorithms can be modified to fit the locale of specific centers.
Shahabsafa said the algorithm isn’t patented yet but will be in the future.
Comment policy
Comments posted to The Brown and White website are reviewed by a moderator before being approved. Incendiary speech or harassing language, including comments targeted at individuals, may be deemed unacceptable and not published. Spam and other soliciting will also be declined.
The Brown and White also reserves the right to not publish entirely anonymous comments.