Plant-Inspired Robot Design Metaphors for Ambient HRI

Technical Summary: Plant-Inspired Robot Design Metaphors for Ambient HRI

Problem Statement
Current Human-Robot Interaction (HRI) design is predominantly grounded in anthropomorphic (human-like) and zoomorphic (animal-like) paradigms. These models often produce overt, high-demand forms of engagement that require focused attention and explicit interaction. In contrast, plants offer a paradoxical model for interaction: they are ambient, low-demand presences that nonetheless shape atmosphere, routines, and relationships through temporal rhythms and subtle expressions. The paper identifies a gap in HRI research regarding how to leverage plant metaphors to reframe robotic presence as ambient, gradual, and interpretive rather than dominant or transactional.

Methodology: Research through Design (RtD)
The authors employ a Research through Design (RtD) methodology to explore plant metaphors as a resource for reorienting robot design. The process involved iterative cycles of ideation, prototyping, and reflection, structured around three research questions:

1. What design primitives emerge from plant metaphors and morphologies?
2. How can these primitives be combined into expressive robotic forms?
3. How do people perceive and imagine plant-inspired robots?

Design Process and Prototyping

• Design Primitives: The team identified four key plant elements—leaves, stems, flowers, and pots—as design primitives.
  • Leaves: Developed using a lightweight, string-driven mechanism with TPU (Thermoplastic Polyurethane) to exploit flexibility. A central spine with triangular wedge segments allows the leaf to transition between a naturally curled state and a flattened state via tension, enabling gestures like wilting, perking, or unfurling.
  • Stems: Utilized a concentric bellow mechanism (inner and outer bellows) combining TPU and PET-G. This allows for smooth, reversible bending through lateral displacement, supporting gestures like leaning, bowing, and swaying.
  • Actuation: The team iterated from noisy servo motors to near-silent stepper motors with TMC drivers and microstepping to preserve ambient presence. Material choices (PLA, TPU, PET-G) were selected to balance rigidity, elasticity, and translucency.
• Speculative Prototypes: Four open-source prototypes were constructed to probe the design space:
  1. Snake Plant: Combines three actuated leaves with a rotatable pot to simulate leaning and swaying, intended for posture correction and daily rhythm marking.
  2. Flower Plant: Features an actuated stem with a flower and passive leaves, designed to mediate social interactions (e.g., turning toward speakers or inviting inclusion).
  3. Keeper Plant: Inspired by the Venus flytrap, uses eight leaves (four actuated, four passive) to function as a pill organizer that grips and releases based on time.
  4. Dancing Pot: A mobile robot with a fern-inspired leaf arrangement that performs subtle dances near water dispensers.

Workshops and Elicitation
To investigate user perceptions (RQ3), the authors conducted three 90-minute design workshops with 11 participants of diverse ages and backgrounds. The workshops followed a three-phase structure:

1. Storyboard Annotation & Critical Reflection: Participants annotated video storyboards of the prototypes' "imaginaries" (e.g., the Snake Plant correcting posture).
2. Micro-Interaction Design: Participants modified storyboards or sketched new interaction sequences, using "bodystorming" to test timing and gesture.
3. New Imaginaries: Participants sketched speculative scenarios for plant-inspired robots in new contexts.

Data was analyzed via thematic analysis of transcripts, annotated storyboards, and sketches.

Key Results and Findings
The study yielded three primary contributions:

1. A Suite of Plant-Inspired Robotic Artifacts: The authors released open-source prototypes and primitives that demonstrate how plant-like expressivity (wilting, swaying, pulsing) can be operationalized in robots without literal mimicry.
2. Designerly Insights on Perception:
   • Multiplicity of Role: Participants questioned the necessity of adding robots to cluttered spaces. Plants are justified by their decorative, ritualistic, and ecological roles; robots must similarly embody multiple roles to be accepted.
   • Calibration of Meaning: Plant gestures are open to interpretation (e.g., a drooping leaf could mean fatigue or growth). While this invites poetic engagement, it risks confusion. Designers must balance legibility with interpretability.
   • Ambient vs. Apparent: Expressions must be subtle enough to blend in but noticeable enough to be perceived, especially for users with limited peripheral vision.
   • Non-Reactivity: Unlike facial robots, plant-inspired robots can offer "intentional negative space" by not reacting, which users found less judgmental for sensitive topics like mental health.
3. Design Considerations for Ambient HRI: The authors consolidated findings into four axes for future design:
   • Presence: Robots should support "cultivation" (user customization) and function as distributed ecologies rather than isolated agents.
   • Palette: Expressive gestures should be calibrated to avoid ambiguity while leveraging the "liveliness" of natural variability (e.g., using Perlin noise) to prevent repetition.
   • Rhythms: Interaction should utilize multiple timescales (rapid nastic movements vs. slow seasonal arcs) to mediate intrusiveness. Timing is critical; gradual cues accumulate meaning over time.
   • Shape: Morphology can range from literal to abstract. Materiality (softness, translucency) and scale are critical levers for shaping presence and acceptance.

Significance and Claims
The paper claims to contribute a new framework for HRI that moves beyond the dominance of anthropomorphism. By treating plants as metaphoric resources rather than blueprints, the authors demonstrate that robots can be designed to be "ambient, gradual, and interpretive." The work challenges the assumption that robots must be overtly interactive to be useful, proposing instead that they can shape environments and relationships through subtle, temporal rhythms.

The authors modestly frame their work as speculative, noting that the prototypes are not fully functional systems but "epistemic tools" to probe "what is the right thing to design." They acknowledge engineering constraints (noise, heat, torque) that currently limit the grace of these artifacts. Furthermore, they raise critical questions about sustainability, noting that robotic artifacts carry a larger ecological footprint than living plants and could potentially displace real botanical life, a tension they identify as a direction for future inquiry.