In April, we started a series of posts which showcased some of the top semester projects for an electronics course from a class of Electrical Engineering students. To test the students and get some creative results, the constraint was that each project had to use a Bluetooth module that communicated with a smart phone application.
So far, we have had a look at a device that can make any speaker wireless, a wearable communication device and a car automation system. In this final entry of the series, we will look at another impressive and interesting project, named “Mution”. It lets the user control a music player on his/her phone with hand gestures.
Project name: “Mution”
Students: Ali Athar, Armaghan Ahmed Khan and Azlan Shaukat.
(All students of BEE-4 at SEECS, NUST)
The inspiration for the group working on this project came from the fact that users have to remove gloves during winters in order to do simple things like changing songs or volume of their media player, which might be uncomfortable. If they could shift the operation of a mobile phone from touch to gesture, it would be easier to use in situations where they are wearing gloves or driving. So they started with a music application that could be controlled with hand gestures.
Design and Operation:
Like most other projects, this one also used the RFduino module which has a Bluetooth 4 chip on it. To recognize different gestures that signified different actions, they used an IMU (Inertial Measurement Unit, a combination of an accelerometer and a gyrometer). Both were put on a glove. The RFduino was programmed to decipher the gesture made from the IMU’s data. It would then communicate this to the music player on the smart phone, which would carry out the appropriate operation.
The music player application on the smart phone was purpose built for the project. The reason for this was that accessing the already installed, default music player was not possible and a music player made specifically for this purpose would ensure smoother and more efficient operation.
The group demonstrated control of song selection and volume of the music player application.
Like everyone else, this group was also using the RFduino for the first time and had their fair share of troubles. It was also their first time using an IMU.
One of the problems they mentioned encountering was that they initially used an algorithm for gesture recognition that later turned out to be too crude for their use. Although they had to work out a finer algorithm half way through the project, it eventually had better results.
The newer algorithm utilized running average on the IMU’s data for more reliable data processing, whereas the initial one worked on raw values from the accelerometer and gyroscope.
Although the demonstrated version of the project did everything the group had set out to accomplish, the group thought there was room to improve and extend. For starters, the group wanted to replace the make shift glove with a proper data glove. That would not only be more aesthetically pleasing, but also easier to work with.
Separate from the project, the group has also worked with a smart phone camera and demonstrated its control via gestures.