At this stage we know a lot about the brand, the product itself and the user. I know what I'm trying to achieve. It's just time to start creating the product and figuring out how it works.
I wanted to start broad. Before thinking about interactions or how the thing moves, I needed to understand what Pane OS looks like when it is doing nothing. What does it look like when nobody needs it?
My first sketches focused on standby mode. The idea that most of the time Pane OS is not being actively used and in those moments it should sit quietly and do something simple. My instinct was a clock. Maybe a date, maybe weather, all contained in one small clean card on the wall. Unobtrusive. Almost furniture.
That immediately raised a question I had not considered. How big is it? In standby mode I imagined something roughly 15cm tall and 30cm wide. Noticeable but not dominant. From there I had to ask what happens when it scales up. Doubling felt too mechanical. Tripling felt better, more intentional.
I also sketched a couple of alternative static modes, music playback being the obvious one. These sketches are not particularly exciting and I knew that when I was drawing them. But they were worth doing to understand the range of quiet states the product moves between without being asked.
The most useful thing to come out of these sketches was actually a question about physicality. How thick is this object? I noted that I liked the idea of a very slight protrusion, somewhere between 1 and 3mm. Enough to feel like a real object rather than a sticker. Enough to cast a faint shadow and feel present without demanding attention. That physical quality felt important to establish early.
Movement is such a central part of this concept that I wanted to visualise it properly rather than assume it.
These sketches looked at what Pane OS looks like in motion. The Flubber reference from my mood board became directly relevant here. The idea that in motion the panel should not remain a perfect rectangle but should elongate, stretching in the direction of travel as though being pulled. That sense of drag makes it feel alive rather than mechanical. It gives the product character before anyone has even interacted with it.
From there I had to answer what happens when it stops. I felt that because so much character had already been established in the movement, the stop needed to match it. A spring stop felt right. Pane OS overshoots its destination slightly and then pulls back into position. That small moment of personality feels very on brand and reinforces the idea that this is something alive rather than something robotic.
I then looked at how it navigates the home. Rather than tracks or rails I landed on the idea of gravitational surfaces. Pane OS is subject to gravity the way we are, moving across walls, floors and ceilings rather than through open air. The surfaces are its ground. That constraint makes it feel more human and less mechanical and it also answered the corner question naturally. If Pane OS is free forming and rubbery in its physical quality it simply wraps around edges rather than stopping at them. That assumption resolves a lot of navigation complexity in one go.
Finally the mitosis moment. One user is watching a film. Another calls Pane OS from a different room. Rather than interrupting the film the panel divides, the requested portion travelling to the new user in that same elongated motion form while the original screen continues uninterrupted. I felt both voice and gesture command should trigger this. Voice is fast but disruptive in a quiet room. Gesture is silent but requires proximity. Together they cover every situation a family might find themselves in.
