Every XSMITHS system — from a single talking installation to a whole responsive building — is engineered as four layers. The visitor only ever meets the last one.
Presence, position, gesture, touch, and identity — captured by sensors engineered into the architecture, not bolted onto it. Cameras, depth sensors, NFC touchpoints, pressure surfaces, and microphones, selected for the venue and its privacy constraints.
Software decides what happens next: AI models for conversation and recognition, show control logic for sequencing, and rules that keep behavior useful rather than novel. This is where an experience gets its judgment.
Media surfaces, projection, lighting, sound, scent, kinetic elements, and robotics move with intent. The response layer is tuned so the visitor experiences a room that cared enough to answer — not a stack of output devices.
Commissioning, monitoring, remote support, and maintenance keep the experience running unattended, in public, for as long as the space needs it. Operation is designed in from the first day — not added after handover.
The four layers are engineered together by one team, through the studio's method — strategy and design through X-Consult and X-Lab, engineering and fabrication through X-Build, operation through X-Ops. That is what prevents an interactive project from becoming a collection of vendors' parts.