|Ken Yarnall Leads Student Team in Mobile App Development
There is a mobile app in development at Lebanon Valley College that is part tour guide and part day planner. It can become a sort of social network that will tell you what’s happening on campus and show you whether your friends are on their way to events or not.
The development project is led by Dr. Ken Yarnall, chair and associate professor of mathematics and coordinator of the computer science program at LVC. Now in its second year, the project is funded by the Arnold Grant program. Student developers work with Yarnall in all phases of planning and execution.
“It was an area of development that I didn’t have any experience in, so we were really coming at it from scratch,” Yarnall said. “As the first summer went on, and we learned more about what we were doing, and adopted the Android platform. We began to see how developing apps for this platform worked.
“A month into that first summer we settled on a loosely defined idea to ‘build something that could be potentially useful on campus.’ None of us really knew what that meant at the time, but the students and I settled on trying to develop an app that would let students and visitors know what was going on at the College.”
Last summer, Yarnall hired John Makatche ’12 and Rob Hosler ’13 after asking students in his department what types of projects they would be interested in working on with Arnold Grant support. Makatche and Hosler both insisted on mobile development—something that Yarnall, himself, wanted to explore.
This summer, Yarnall was able to retain the services of Hosler and add Caitlyn Light ’13, a mathematics major. The idea behind rotating students throughout the life of the project is to have current students advance the app and set up ideas for future students to work on as technology progresses.
Currently, Yarnall and his staff work on algorithms that will accurately triangulate a user’s position on campus using Wi-Fi access points around campus and floor plans of campus buildings. If the algorithms can read signal strength gained from each access point, they can begin to pinpoint where the user is to a degree higher than standard GPS signals.
“One of the problems I wanted to try to understand was how to do that without investing in a very expensive infrastructure,” Yarnall said. “How can the phone figure out where it is, and once it does, what can we do with that information?”
As solutions become more feasible, ideas continue to form regarding what to do with the information. Theoretically, the app will be able to tell you which building or landmark you are standing at, give you information about the landmark with a multimedia content presentation, and list upcoming events in an area of campus. It would also allow users to opt into letting friends see where they are on campus, increasing the social ability of the app tenfold.
“One of the nice things about computer science compared to physics or chemistry or biology is that nothing has to die or blow up when you do something wrong,” Yarnall said. “Your experiments can go bad and it doesn’t hurt too much.”
Experimenting has gotten the team this far. Though the programming language of mobile app development was familiar, the constraints of the devices were not. For example, a cell phone doesn’t have the seemingly endless amounts of memory and CPU speed that a desktop or laptop has.
Another area of concern is in the app’s aesthetics. Yarnall says that so much time is spent on technical development and coding that it becomes difficult to focus on making the app “pretty” at this point. As functionality and success increases, the interface will become more of a primary focus.
It’s a balancing act. In coding, Yarnall and his students are mindful that one day, the entire interface may need to be ripped apart and redone. “If you’re lazy about it and just tinker in the code early on without thinking about the future, you end up with things that are too tightly bound together to do what you want to do without rejiggering everything.”
Stripped down, the application could be released today. It could, for example, provide an events calendar. The idea is to hold off until it is a more well-rounded and useful app that can quickly and effectively pull information straight from the LVC server and automatically display it in a usable format.
The program will require continued funding—whether that comes from continued support through Arnold Grants or through contributions from various departments on campus.