2015-03-21 - By Robert Elder
I was recently invited to be a referee for the 2015 SE Design Symposium at UW, which was quite a wake up call for the fact that it has been exactly one year since I was back in school doing the same thing. During the symposium, students in engineering programs present their fourth year design projects. These projects represent a body of work that teams of students have been working on for about two years at the point of their presentation. The idea is to prepare engineering students to be able to manage all aspects of a project, including all of the non-technical aspects like presentation and communicating with customers.
The first group I refereed was 'Sonicle'. Sonicle is a mobile app designed with the idea of using data about your physical location and movements to assist in recommending music more effectively. Other music services only use information like your friends' listening preferences to recommend you music. An example of how this extra information is purportedly useful would be to have a higher chance of recommending energetic music while at a gym.
The second group I refereed was 'Map2Go'. Map2Go was pitched as a route planning tool that allows for efficient route planning after users enter various kinds of information like constraints, a home point, and frequent destinations.
Of the two projects I would say that 'Sonicle' was the most interesting one. The idea of using sensory feedback to reinforce music recommendations definitely seems like it would have merit, although I'm not convinced of the marketability of this idea in isolation. If one were to pursue an app like this commercially, I would expect most of your technical challenges to come from things other than the recommendation engine.
Unfortunately, for the Map2Go team I was unable to catch up with them before their presentation, despite hanging around their unattended booth for about 15 minutes until I had to leave for the first presentation. Despite this no-show, I decided to leave any judgments until after I had seen their presentation. After the presentation I still wasn't able to see how the app provides substantial practical benefit over vanilla Google Maps. Despite this, I gave both groups a generous grade since I can sympathize with their pain having been in their shoes only a year ago.
In terms of other software engineering projects, one that was memorable was homebrü. Their homepage doesn't say much at the moment, but the idea is to allow identity based access to your wifi network. I didn't review the details of how it is done, but the high-level promise was to allow people to sign into wifi networks with Facebook or other identify based services.
An interesting non-software project was spillOpill. The promise of this project is a substance that is capable of economically absorbing sub-surface oil in ocean water and returning it to the surface. If I recall the process correctly, a special oil absorbing material is covered in slowly disolving salt. The substance sinks to a certain depth, absorbs oil, and then returns to the surface where it can be collected.