Apple has introduced a new artificial intelligence (AI) framework designed to help non-humanoid robots communicate intent through motion. Named ELEGNT, the framework allows robots to express intention, attention, and even emotions using movement, posture, and gestures. Unlike humanoid robots, which naturally engage users due to their familiar design, non-humanoid robots often struggle to convey meaning. ELEGNT addresses this gap, making human-robot interaction more intuitive and engaging. Apple researchers tested the framework with human participants to assess its effectiveness in real-world scenarios.
The Cupertino-based company detailed the framework in a recent post, emphasizing its focus on non-anthropomorphic robots—machines that lack human-like features such as limbs or facial expressions. While these robots excel at task execution, their rigid and mechanical nature can make interactions feel impersonal. By introducing expressive movements that do not interfere with task completion, Apple aims to create a more natural and immersive collaboration between humans and robots. The framework includes both hardware design considerations and software-based training techniques to enable fluid, expressive behavior.
To bring ELEGNT to life, Apple researchers developed a set of interaction scenario storyboards. These scenarios map out how a robot’s movement can enhance engagement and improve user perception. The findings suggest that adding expressive movements significantly boosts a robot’s perceived intelligence and sociability. A research paper detailing this approach has been published on the pre-print server arXiv, highlighting the potential for more emotionally resonant robotic systems in various industries, from home assistants to industrial automation.
A demonstration showcased the expressive capabilities of ELEGNT through a lamp-like robot prototype. In a video released by Apple, the robot, resembling Pixar’s Luxo Jr., responded to human gestures by directing light toward indicated spots. Its movements mimicked behaviors like understanding commands, nodding in agreement, and executing tasks efficiently. This experiment illustrates how even simple, non-humanoid robots can engage users on a deeper level through intentional motion, redefining how machines and humans collaborate.