Exploring Human-AI Interface for Future AR 

Topic: User Study for the Future AR Platforms

Facebook Reality Labs (FRL)

Summer, 2019, 4 months

My role: Research intern

Methods: Ideation workshop, affinity clustering, semi-structured interviews, card-sorting, body-storming. 

Overview of the Topic, Study Design, and Execution

At the beginning of my internship, the team proposed three potential project topics, and I was given the flexibility to choose the one that interested me the most. The topic I chose made me super excited. It had broad research scope and very ambitious goals. Briefly, we were interested in how to deliver personalized Human-AI experience based on users' context, intentions, and behaviors through AR. To make sure it was feasible for a four-month internship, we narrowed down the project goal to the following: consolidate AR experience concepts using specific scenarios. My starting point was the pre-generated AR experience concepts.

Methods

To achieve the project goal, I employed several methods, including ideation brainstorming workshops, affinity clustering, and semi-structured interviews. 

• Body-storming

  • I took the consideration of multiple inputs in the AR experience. For each interview, I designed a body-storming session and created a set of toolkits for participants to act out their thoughts. Questions were also asked for participants to talk through the reasons.

• Card-sorting

  • For each interview, there was a card-sorting session to better understand the user-centered values and design strategies of multi-inputs.

Deliverables

In summary, the final results inform the team's future research direction. By the end of the internship, I did an internal presentation to demonstrate the study design process and findings. The main deliverables and contributions are summarized here:

●      User-centered value - Explored the user-centered value of each predefined AR experience concept.

●      Refinement - Discussed how the team might leverage the user-centered values to inspire future refinements of the experience.

●      Use case generation - Proposed methods to generate more use cases based on the new insights; suggested additional use cases for each experience concept.