In addition to our video prototype, we had a ton of fun prototyping some of our ideas over the last several weeks. Although we lacked the technology and resources to prototype some of our more interesting ideas, like AR Knitting, we did the best we could with what we had, and learned a lot in the process.

Below are some first iteration examples of our quick-and-dirty prototypes.

01. Skillshare Prototype

We prototyped our skillshare idea, a synchronous and remote experience, first by using two phones, two computers, and a four-way call on Google Hangouts, but quickly switching a phone out for a webcam to eliminate some technical difficulties (screen freezing, lag, blurriness, etc.)

Our lessons began face-to-face for introductions, utilizing the computer cameras and allowing learner and expert to spend a few minutes meeting each other, increasing comfort for both participants.

Although participants were given the option to alternate between views, participants never chose to return to the face view once they had switched to the first person perspective.

We received very positive feedback about our prototype. Participants noted the friendly tone of the experience, encouraging, positive feedback, and immediate corrections as some of the more positive aspects of the experience. Alternatively, they were at times frustrated by the technical difficulties experienced due to the limitations of using mobile phones.

It was also brought to our attention that our idea was, while feasible, perhaps a bit too safe and shallow, and that we should push ourselves a little bit harder to come up with a more novel and compelling idea.

02. Prototyping The Sphere

What makes The Sphere so compelling, is the 3-dimensional view it offers the learner. When trying to figure out ways to prototype a 3D experience, however, we struggled to think of a feasible idea due to unavoidable issues, such as available technology. Ultimately, we realized the only way to really and truly build a 3D prototype within our constraints, was to use an actual human.

Or, more specifically, for Lauren to act as a human-robotic sphere.

Unlike a sphere, a human would of course understand when and where a learner is struggling, and adjust the tutorial to meet the needs of the user. In order to eliminate as much bias as possible, Lauren played loud music through her headphones, and sat facing away from the user.

To provide the user with control over the tutorial, we placed paper controls for play, pause, and back in front of our robot, and Aaron pointed to the user-issued command.

Our first participant, Tayseer, learned faster than any other participant using any other prototype observed so far. He changed position multiple times throughout the tutorial, and peered around for different views, noting the ability to access multiple views and angles due to the 3-dimensional experience as incredibly helpful.

Although a very quick and dirty prototype, we had the best results of any asynchronous prototype tested thus far, and so we began moving forward with the concept.

The Sphere:
Human-Computer Interaction

After organizing our information architecture, we got to work on creating a paper prototype as a low fidelity, first iteration of a testable interface.

During our first internal evaluation of our paper prototype, we uncovered an incredibly valuable insight that shaped the direction of our concept.

Because she had taught several lessons prior, during evaluation of the skillshare prototype, Lauren had become accustomed to slowing down instruction and including extra explanation for struggling students, while ramping up the pace and streamlining explanation for participants who caught on quickly.

While functioning as a human Sphere, however, Lauren lacked the visual and auditory cues required to customize the tutorial for the user, one of the more valuable aspects of our synchronous skillshare concept.

What our prototype did include, however, was control over the speed of the lesson. While intended to function as traditional video speed controls, Lauren naturally adjusted the pace of the tutorial when instructed to speed up or slow down, recognizing the likelihood that it was an indication of struggling or success by the student; she organically adjusted the pace of the tutorial to fit the needs of the user, instead of just the tutorial speed.

We quickly realized that this revelation provided an excellent opportunity for our system to understand user context, and incorporated a pacing feature that adjusts to user needs based on a combination of a user-tracking algorithm, and periodically checking-in with the user.

Our system incorporates the following three paths for a tailored learning experience:

1. Easy — Slow Learner
The easy path includes additional instruction and more detailed explanations, as well as a slowed down video, to assist slower learners through the experience, and help prevent frustration.

2. Default — Average Learner
The default path reflects the needs of the average learner. The video is played at a normal speed, and a moderate amount of instruction is included.

3. Accelerated — Fast Learner
The accelerated path eliminates some of the repetition, and streamlines instruction to include only what is necessary.