Airy: Reading Robot Intent through Height and Sky
Baoyang Chen, Xian Xu, Huamin Qu
TL;DR
Airy investigates how to render multi-agent reinforcement learning behavior legible in public contexts by mapping internal control signals to perceptible cues. It proposes a triadic design: a competition metric (height) paired with a familiar gesture (bedsheet snapping) and sensor-to-sense mappings (forest-altimeter and atmospheric visualizer) to reveal coordination and conflict in real time. The system uses two RL-controlled robot arms trained in cloth simulation with domain randomization, with a blue-hush safety mechanism and a live forest sky narrative that evolves with performance. Field deployments show audiences consistently interpret strategies and emotions from the visuals, suggesting a path to turning 'black-box' AI into a publicly navigable interface, while acknowledging the need for broader evaluations and accessibility improvements.
Abstract
As industrial robots move into shared human spaces, their opaque decision making threatens safety, trust, and public oversight. This artwork, Airy, asks whether complex multi agent AI can become intuitively understandable by staging a competition between two reinforcement trained robot arms that snap a bedsheet skyward. Building on three design principles, competition as a clear metric (who lifts higher), embodied familiarity (audiences recognize fabric snapping), and sensor to sense mapping (robot cooperation or rivalry shown through forest and weather projections), the installation gives viewers a visceral way to read machine intent. Observations from five international exhibitions indicate that audiences consistently read the robots' strategies, conflict, and cooperation in real time, with emotional reactions that mirror the system's internal state. The project shows how sensory metaphors can turn a black box into a public interface.