Animal-Machine
Exhibition
Personal work
Animal-Machine is an interactive installation exploring artificial agency through a physical robotic creature and its virtual counterpart. The project questions how simple programmed behaviours can be perceived as intention, understanding and personality.
Overview
The project distributes a single behavioural identity across two simultaneous forms: a 3D-printed creature mounted on wheels and a Unity-based virtual agent displayed through an embedded screen. The physical body moves autonomously through space, while the virtual counterpart responds to specific voice phrases. Rather than creating a truly intelligent system, the project deliberately relies on simple rules and scripted responses. The contrast between physical movement and simulated conversation creates the impression of a single entity that can act, react and communicate.
Interactive UI for BPM Scanning & Aura Generation
technical process
The physical creature was 3D-printed and controlled with an Arduino microcontroller. An obstacle detection sensor continuously influences its movement, producing non-deterministic trajectories rather than following predefined paths. In parallel, a Unity environment handles the virtual character and voice interaction. Speech input is processed through phrase detection, triggering scripted dialogue responses. The Arduino and Unity systems operate independently, with the embedded screen acting as the perceptual bridge between the physical and virtual versions.
Speech-recognition demo in Unity
Closing note
Animal-Machine investigates the gap between behaviour and understanding. The machine does not actually understand language or make decisions, but its consistent responses and autonomous movements can make it appear intentional. The project asks whether perceived agency comes from cognition itself, or from the way observers interpret behaviour.
3D Printing
Speech Recognition
Unity
Arduino
Obstacle Sensor
Embedded Display
Interactive / R&D
