eight actions
We started from what the TV needed to understand: unlock, select, back, pass ownership to another viewer, volume, home, and navigate left and right.
- unlock
- select
- back
- pass ownership
- volume
- home
- navigate left
- navigate right
design · 2012
A gesture library for a smart TV interface. Instead of inventing gestures, we looked for the ones people already use without thinking, then narrowed them down with the engineers who would build the recognition.
We started from what the TV needed to understand: unlock, select, back, pass ownership to another viewer, volume, home, and navigate left and right.
Each action had its own study format: behavioural observation, interviews, card placement, task analysis and questionnaires. From video, photos and recordings of people interacting, we extracted the gestures they made naturally.

The insights fed a brainstorm that produced a shortlist. We went through each gesture with the technical team against what the sensor could reliably detect, then sketched scenarios to explain the use cases to the developers.

Internal tests checked each implemented gesture for affordance, function and false triggers. The results led to new threshold values, so the TV ignored everyday movement and responded to intended gestures.
