When Michael Scott Brown teaches the capstone software engineering course at University of Maryland University College (UMUC), he starts a special discussion thread asking students to recommend organizations that they want to work with on projects.
During the course, student teams spend 12 weeks building software applications for an outside institution, and Brown wants to give them a chance to be masters of their own project destinies.
Normally, he gets no recommendations for projects from his students, and when he gets something, it’s typically “too big in scope,” said Brown, who directs UMUC’s software engineering program and also serves as a professor in the Graduate School.
Last summer, though, things played out differently when one of his students, bioinformatics software engineer Tyler Marrs, nominated his employer—Children’s Mercy Hospital in Kansas City, Mo.
Brown, who does bio-research himself, was initially skeptical about whether others in the class would be interested in such a narrow topic. But, quickly, a “critical mass” of students with scientific backgrounds, including a biologist, wanted to be involved in the project. So Brown contacted the hospital and formalized the details.
Marrs led the UMUC team that included classmates Shaun Einolf, Daniel Epstein, Robert Holloway, and Katie Heasley. The group worked with a contact at the hospital to build an application that researchers at the hospital, and beyond, can use to share information about something called DNA motifs, or patterns.
In more technical terms, the team’s application stores variants as haplotypes―that is, it stores sets comprising similar genetic-mutations sequences, Marrs explained.
“The main goal was to store these haplotype sets, so that researchers can use this information for diagnostic purposes,” he said. At Children’s Mercy, some of the children have issues metabolizing certain drugs, which can harm them, so the application aims to reduce errors in dosage.
Brown noted that an application that allows researchers to share information of this sort can “help tremendously” in the fields of medical research and, in particular, cancer treatment.
“If you’re a researcher, you want to know what the research community [as a whole] knows about this motif,” Brown said. Previously, to find the answer researchers had to search and hunt in a variety of places. The centralized database “is useful for all kinds of things.”
“It was exciting to create something that is being used to reduce drug dosing errors. There are many deaths daily due to dosing errors,” Brown said. “It is very rewarding to know that the application is being used, and the cause it is being used for.”
Marrs, who earned his Master of Science in Information Technology Software Engineering from UMUC in 2015, said he found the project rewarding for the same reason. Leading the project, and specifically addressing the challenges of distilling complex information to those who are non-experts, was particularly meaningful he added.
That, said Brown, is a fundamental point of the project, which is designed to give students, many of whom are hoping to change careers, first-hand experience at solving real-world problems.
For instance, Brown said he heard two years ago from a student who had taught biology for a decade and wanted a change. When he applied to SRA International, now CSRA International, the student found that his SWEN 670 project gave him a competitive advantage.
“He told me that they counted what he did at this capstone as work experience,” Brown said, noting that other students who went on to Fortune 500 companies had related similar experiences.
It’s not surprising that companies weigh SWEN 670 projects so heavily when making hiring decisions, because they are essentially soup-to-nuts training.
“Students go and do full life-cycle development from gathering requirements, to designing, to building and testing and, in many cases, documenting the software,” Brown said. “And then they turn it over to the sponsor to use in the organization or company.”
Sponsors receive free software, licensing, documentation and source code. “It’s kind of a win-win from the administration side, the student side and the corporate sponsor side,” he added.
There’s a big added plus, according to Brown, who said the industry standard in IT tends to be roughly 50 percent of projects completed on deadline, meaning 95 percent of requirements are met. The typical rate for SWEN 670 is well above that at 70 percent. “That is pretty good in this industry,” he said.