I used several prototyping exercises to explore my ideas, establish design criteria, and iterate my design solution.

I explored the use audio and voice cues while running with a series of improvised role-play scenarios. In these scenarios we prototyped using voice control as a way to respond to a phone call that is received during a run. I learned three key things about control: that giving control to runners over which calls they will accept is very important; that passive control, such as a one-word command, feels unobtrusive and manageable; and that pre-set filters work well as they limit interruptions and give the runner the flexibility to customize the list of which calls get through.

After the role-play scenarios, I developed a second round of prototyping exercises to gather criteria for for the device's physical design. I used simple foam core pieces to test weight and positioning of the device, and how individual components might be distributed on the body. Each component was a different size and weight, and users were asked to attach them to themselves using Velcro, and an armband or a belt. The users then went through a series of pre-defined tasks, and to verbally describe their thoughts.

To validate my design concept I turned once again to role-play. I wrote a narrative about the events in the day of a runner (based on user research and my own experience), and then documented the scenario in video to illustrate the concept and test its validity. In working with a runner to visualize the narrative, I was able to pinpoint the issues that needed further thought. For example, my initial concept for switching among the features of the device was to press down on a small button at the top of the remote, but tapping the remote repeatedly to tab through the features (in this scenario, worn on the arm) quickly proved to be impractical. I found the video prototype a valuable tool for evaluating my concept and learning where to focus further iterations of my design.

At this point I also implemented my first technical prototype, a script designed to provide a preliminary idea of how audio feedback to interaction with the device would work best. I tested two simple interactions: Tabbing through the device features, and receiving a phone call. Although tabbing through the device features proved to be laborious, I found the audio cue was an effective way to orient users as to what feature they were accessing.

To solve the task of tabbing through features using two buttons, I created a series of flash prototypes which simulated interface variations for accessing the main features of the device. The features I tested included: the phone, MP3 player, chronometer, speedometer, and heart rate monitor. I also tested starting and stopping a function and checking its status. I tested various interaction mappings and navigation flows through the functionality of the device. By simulating physical input and audio output, I was able to quickly pinpoint the issues with each interface layout. This exercise confirmed that independent access to the main functions of the device, and therefore multiple buttons, were necessary. I then tested several versions of a multiple-button layout, and refined each one until the combination of physical input and audio output was satisfactory.

Once I narrowed the possibilities to a few, I tested the placement of each layout on the arms runners to see which orientation worked best. To do this, I created a simple prototype using foam pieces and a felt armband. The foam pieces represented a range of buttons positioned vertically, horizontally, and in a circle. I found that finger memory and accuracy while in motion was most successful with a vertical layout, because the button placement mirrored the natural position of the fingers.

My next step was to create a physical prototype of the device interface and define the form of the device. I used forest foam to cast the imprint of my fingers across my arm and my wrist. This prototype helped me understand the curve of the fingers and define the ideal placement of the buttons within a three-dimensional form. Next I created a series of sketches and physical prototypes of the device to determine its ideal shape and size. I tested a number of ideas and evolved the shape of the device from 3D sketch into a defined interface layout.

I also created paper prototypes of the screen interface runners would use to set up the actively mobile device, and designed experience prototypes to explore and my ideas for service concepts. Through this process of continual refinement and testing I was able to grow my idea and ground it in user needs, to create a realistic and usable mobile solution for running.