Self-Resetting Soft Ring Robot

• The paper demonstrates the design of a self-resetting soft ring robot that transitions between distinct gaits using controlled pneumatic actuation.
• The robot leverages a central inflatable bladder and peripheral actuators optimized through simulation to enhance locomotion across varied terrains.
• Material choices like Dragon Skin silicone and variable-friction feet ensure reversible deformations and robust performance in diverse environments.

A self-resetting soft ring robot is a class of adaptive, deformable, toroidal robotic system designed to autonomously change its resting geometry—typically between multiple distinct shapes—with the purpose of transitioning between incompatible locomotion modes and adapting to diverse environments. In such systems, "self-resetting" refers to the robot’s ability to actively reconfigure its body without human intervention, typically through actuation of internal structures, resulting in reversible and repeatable transformations. The design is inspired by shape-morphing strategies seen in biological organisms, such as spiders and caterpillars, which reconfigure their bodies to optimize movement across variable terrain (Shah et al., 2020).

1. Structural Mechanisms for Adaptive Shape Change

A prototypical self-resetting soft ring robot utilizes a hierarchical actuation and structural scheme:

• Central Inflatable Bladder: The primary morphing element is an internal core bladder, fabricated from highly stretchable silicone elastomer (Dragon Skin 10), which regulates the robot's global cross-sectional geometry. Inflation (≥12 kPa) yields a cylindrical, rolling-optimized form; deflation causes flattening, suitable for crawling or inchworm gaits.
• Peripheral Pneumatic Bladders: Eight external pneumatic actuators, constructed from stiffer silicone (Dragon Skin 30), are arrayed peripherally. Sequential actuation induces localized bending (for inchworm motion) or facilitates circumferential rolling.
• Variable-Friction Feet: At each end, two specialized feet (latex balloon and silicone composite sheathed in cotton) provide surface-dependent friction, actively modulated according to internal bladder pressure. Inflation increases friction by 25–35%, enabling controlled gripping during crawling modes.

The robot’s active self-resetting is enacted through pneumatic control of the central bladder: inflation/deflation cycles orchestrate the switch between functional shapes and gaits, with actuation timing and pressure regulated by an external controller receiving optimized policies from simulation (Shah et al., 2020).

2. Control Policy Optimization and Implementation

The functional versatility of the system derives from a model-based optimization pipeline:

• Parameterization: Control variables comprise shape (core bladder pressure $p$), orientation ($\theta$), and discrete actuator firing policy ($S$), encoded as a binary matrix representing the timing and selection of pneumatic bladders.
• Optimization Algorithm: A hill-climbing algorithm explores the control space (shape, orientation, actuator policy) within physics-based simulation (Voxelyze), seeking to maximize displacement per unit time (body-lengths per second, BL/s) for each terrain.
  • For a given environment (e.g., flat, incline; prescribed in tests), shape is discretized ($p \in [0, 12]~\mathrm{kPa}$), orientation varies over a relevant range, and actuator sequences are searched.
• Policy Transfer: The optimal (shape, orientation, control) triple for each terrain is translated to robot hardware, with adjustment for actuator variance. Terrain adaptation in experimental runs is manually triggered, with future work suggesting sensor-driven autonomous switching.

This approach allows the robot to dynamically “self-reset” to the optimal configuration for the encountered terrain, outperforming non-morphing controls by a significant margin (Shah et al., 2020).

3. Material Selection and Fabrication for Reversibility

Material systems are chosen for high stretchability, resilience, and durability during repeated large deformations:

Component	Material	Property Leveraged
Core bladder	Dragon Skin 10	High flexibility/stretch
Peripheral bladders	Dragon Skin 30	Stiffness for actuation
Structural layer	Silicone/cotton	Flexibility, structural constraint
Feet	Latex/silicone	Friction modulation

These choices ensure that shape transformations are both reversible (key for cyclic self-resetting operation) and robust under repeated morphing cycles.

4. Comparative Locomotion Performance

Shape-morphing capability yields quantifiable locomotion advantages:

Locomotion Metric	Shape-Morphing Robot	Non-Morphing Robot
Flat terrain (rolling, BL/s)	0.1888	0.1576
Inclined terrain (inchworm, BL/s)	0.0158	−0.0023
Combined (Table 1, BL/s)	0.1023	0.0777
Gait transition	Rolling ↔ Inchworm	Single gait only
Environment adaptation	Flat, incline, granular	Flat only
Self-resetting	Yes (inflate/deflate core)	No

Empirical results reveal that non-morphing robots cannot traverse sloped terrain (rolling backward), while morphing systems, through self-resetting, ascend inclines up to ~14°.

5. Simulation and Hardware Validation

Simulation in Voxelyze enables robust model-based optimization, with direct hardware transfer:

• Hardware Validation: Locomotion controllers derived from simulation reliably drive the physical robot, with only minor manual retuning for actuator variability.
• Friction Testing: Inflatable feet, when modulated, provide measurable increases in friction ($\mu$), facilitating crawling gaits.
• Shape Transition Time: Inflation to rolling shape requires ~11.5 s at 50 kPa; flattening is faster (~4.7 s at −80 kPa).

Physical realization demonstrates seamless, reliable self-resetting during environment transitions, validating the system architecture.

6. Implications for Design and Future Directions

Autonomous, reversible shape change via self-resetting supports complex adaptive robotics:

• Functional Multiplicity: A single robot accesses multiple, otherwise incompatible gaits (rolling/inchworm), controlled by morphing into distinct geometric states.
• Design Automation: Integration of simulation-based policy search with soft actuation mechanisms yields an end-to-end adaptive robot design pipeline.
• Scalability: Future enhancements include multi-core morphing, expanded morphospace exploration, advanced control algorithms, and sensor integration for robust autonomous environmental recognition and online gait switching.

A plausible implication is that self-resetting morphing mechanisms will expand operational domains for soft robots, enabling dynamic adaptation to unstructured, unpredictable, or evolving environments—critical for field robotics, search-and-rescue, and autonomous inspection.

7. Key Mathematical Formulations

The system’s shape and bending ability are characterized quantitatively:

• Curvature when flat: $\kappa = \frac{M}{E I}$, lower $I$ in flat configuration increases bending efficacy for crawling.
• Locomotion fitness: Calculated as body-lengths per second—benchmark for gait efficiency across shapes/environments.

Fluid transition between rolling and crawling gaits is thus not only physically implemented but rigorously modeled, providing design and control insight for future self-resetting ring robots.

---

In summary, the self-resetting soft ring robot achieves terrain-adaptive, reversible morphing via controllable pneumatic bladders and variable-friction feet, orchestrated through optimized actuation policies that transition between rolling and crawling gaits. Simulation and experimental validation confirm enhanced performance over non-morphing systems, with the architecture laying critical groundwork for broadly adaptive morphing robots in complex environments (Shah et al., 2020).