Hiding a screenless AI module in a plush: placement is the design

Screenless companion plush put every pixel of their personality into voice and motion, which makes module placement the central design act. The same module in two positions yields two different products.

The microphone wants the face

Sound entry points belong near where a speaker stands in front of the plush, not where manufacturing finds room. Port placement behind thick pile or a sewn label muffles everything downstream; we keep the mic openings on the face plane, with a short, protected air path.

The speaker wants a chamber

Small speakers sound like small speakers unless the module design gives them an enclosed rear volume. Inside a plush, stuffing is not a chamber — it is an acoustic sponge. A module with a real rear enclosure, even a modest one, doubles the perceived warmth of the voice. This is the least visible BOM line and the one buyers can hear.

Weight is a posture decision

The module and battery together are the plush’s centre of mass. Sit-posing plush want weight low and forward; hang-tags plush want it centred. Getting this wrong yields a toy that slumps, tips or sits in a way the character “wouldn’t”.

The antenna is part of the silhouette

Bluetooth antennas hate proximity to stuffing density and love the open. We plan antenna position with the pattern maker, not after — the fabric cut-out that frees the antenna becomes part of the silhouette instead of a patch hidden inside.