PaperlayerView exact revision on arXiv

Source-linked AI summary

SpiRobs: Logarithmic Spiral-shaped Robots for Versatile Grasping Across Scales

Zhanchi Wang, Nikolaos M. Freris, Xi Wei

arXiv:2303.09861v2cs.RO

TL;DR

Soft manipulators are difficult to design because their kinematics and surface geometry are hard to predict together, while a precise logarithmic-spiral design principle has been lacking. SpiRobs use that morphology with cable-controlled curling and uncurling for adaptive grasping, and the paper demonstrates variants spanning millimeter to meter scales. The authors report adaptable object handling and three applications, while noting that contact during grasping can push some objects away.

  • Problem

    Soft manipulators are often case-specific because compliant deformation makes their kinematics and surface geometry difficult to predict together, and the curling and uncurling of a logarithmic spiral had not been precisely described as a design principle.

    Source: INTRODUCTION · pp. 1–2

  • Method

    SpiRobs discretize a logarithmic spiral and use cable-controlled curling and uncurling, with contact deformation supporting adaptation to object shape and size.

    Source: INTRODUCTION · p. 2, RESULTS · p. 2

  • Results

    SpiRobs demonstrate adaptable handling of objects that vary in size and shape, with applications spanning a miniaturized robot, a one-meter manipulator, and an entangling multi-robot gripper.

    Source: INTRODUCTION · p. 2, CONCLUSION · p. 11

  • Takeaways & Limitations

    The logarithmic-spiral model provides a basis for deriving robot design and fabrication across diverse application scenarios rather than relying on trial and error or extensive simulations.

    Source: DISCUSSION · p. 10

    Uncurling along an object requires contact force that may push it away; this was not an issue when humans handed objects to the tested two-cable robot.

    Source: DISCUSSION · p. 11

Abstract

from arXiv · show

Realizing a soft manipulator with biologically comparable flexibility and versatility often requires careful selection of materials and actuation, as well as attentive design of its structure, perception, and control. Here, we report a new class of soft robots (SpiRobs) that morphologically replicates the logarithmic spiral pattern observed in natural appendages (e.g., octopus arms, elephant trunks, etc.). This allows for a common design principle across different scales and a speedy and inexpensive fabrication process. We further present a grasping strategy inspired by the octopus that can automatically adapt to a target object's size and shape. We illustrate the dexterity of SpiRobs and the ability to tightly grasp objects that vary in size by more than two orders of magnitude and up to 260 times self-weight. We demonstrate scalability via three additional variants: a miniaturized gripper (mm), a one-meter-long manipulator, and an array of SpiRobs that can tangle up various objects.

INTRODUCTION

SpiRobs use a logarithmic-spiral morphology to address the challenge of designing soft manipulators whose kinematics and surface geometry work together. The paper presents a cross-scale design and an adaptive grasping strategy inspired by animal appendages.

INTRODUCTION

  • Soft manipulators are often case-specific because their compliant deformation makes kinematics and surface geometry difficult to predict and integrate.Existing prototypes rely on long simulations and trial and error to design both.Source: INTRODUCTION · p. 1
  • Animal appendages across species and scales share a tightly packed logarithmic-spiral shape, but its curling and uncurling had not been precisely described as a robot-design principle.Earlier work used other spiral forms or observed spiral-like shapes emerging through deformation.Source: INTRODUCTION · p. 2
  • The paper aims to specify the design, fabrication, and operation principles of soft robots from a holistic logarithmic-spiral geometry.It introduces a discretized spiral design and an octopus-inspired grasping strategy, alongside miniaturized, one-meter-long, and multi-robot applications.Source: INTRODUCTION · p. 2
  • SpiRobs combine a cross-scale logarithmic-spiral design with cable-controlled curling and uncurling to wrap around objects of varied shapes and sizes.The strategy uses deformation upon contact to adapt, without requiring precise feedback or complex planning and control.Source: INTRODUCTION · p. 2

RESULTS

SpiRobs use discretized logarithmic-spiral units to couple body geometry with curling and uncurling. Results cover cable-controlled grasping and variants spanning miniature, meter-long, and array scales.

Design of SpiRobs

  • Angularly discretizing and mirroring a logarithmic spiral forms robot units with a fixed size ratio; a cable and elastic axis produce curling and uncurling.The units are extruded, cut into a conical body, and fabricated with cable holes.Source: Design of SpiRobs · pp. 2–3, RESULTS · p. 2
  • The CC-T workspace is about one-fifth of SpiRob’s, while greater taper increases load capacity for a fixed object diameter and smaller taper expands workspace.A 5% elastic-layer thickness was selected for the two-cable robots; thinner layers risk self-collision or buckling, while thicker layers require greater curling force.Source: Design of SpiRobs · pp. 4–5, Figure 2 · p. 4
  • 5.6–115 mm graspable diameters and a 10 kg load (about 260 times the 38.4 g robot weight) were demonstrated with the 15° SpiRob.The figure shows the 5.6 mm cable, 115 mm tape, and 10 kg load.Source: Figure 2 · p. 4, Design of SpiRobs · p. 5

Bioinspired grasping strategy

  • Antagonistic control of two cables moves the robot from packing through reaching and surface climbing to wrapping and grasping.Current increases detect object contact; the strategy achieved 94.4% success across six objects at six positions using approximate object location rather than high-frequency feedback or precise motion planning.Source: Bioinspired grasping strategy · pp. 5, 7, Figure 3 · p. 6
  • Grasp success depends on object size and position: smaller objects can be grasped farther toward the workspace boundary, and an object may fail when placed too far or too close.Tests used 25 mm, 50 mm, and 100 mm objects at grid points in the workspace.Source: Bioinspired grasping strategy · pp. 5–6

Scalability

  • A robot scaled down 70 times to about 1 cm handled and manipulated an ant without damage.The miniaturized robot has 29 units and is actuated through two cables connected to sliders.Source: Scalability · p. 7, Figure 4 · p. 8
  • The 100 cm, three-cable SpiRob demonstrated surface wrapping, dynamic whipping, and grasping and lifting a headset within 1 s.With a drone under remote control, it grasped a barrel across a wide range of positions without precise hovering.Source: Scalability · p. 8, Figure 5 · p. 9
  • A six-arm SpiRob array achieved more than 90% grasp success across ten objects; load capacity increased with branch complexity but fell for the 16-branch target.The array grasps by entanglement, without perception or feedback control; excessive branching left insufficient space for full entanglement.Source: Scalability · pp. 9–10, Figure 6 · p. 10

DISCUSSION

The discussion frames the logarithmic spiral as a model-first design target linking SpiRob geometry and deformation, while noting grasping limits and possible extensions to its simplified biological abstraction.

DISCUSSION

  • The logarithmic-spiral model comes before hardware design, supporting application-specific SpiRobs without trial-and-error builds or extensive simulations.The authors identify this model-first principle as the paper’s main novelty.Source: DISCUSSION · p. 10
  • The logarithmic spiral defines both robot kinematics and geometry, offering a unified design principle for soft manipulators.Soft-manipulator deformation depends on materials, structures, and actuation, which makes a unified principle difficult to establish.Source: DISCUSSION · p. 10
  • For a given taper angle and size, SpiRob kinematics and geometry can be derived from the spiral parameter b.The authors propose the logarithmic spiral as a possible design target for assessing whether both are properly designed; exact logarithmic-spiral deformability in all directions had not been reported.Source: DISCUSSION · p. 11
  • SpiRobs use 2 or 3 cables for complex operations, while uncurling the spiral into a tapered shape simplifies 3D printing and cable threading.The authors also demonstrate paper-based origami fabrication.Source: DISCUSSION · p. 11
  • Graspable space is smaller than tip reach because part of the body must wrap around an object, and contact can push an initially unfixed object away.The paper demonstrates grasping unfixed strawberries or pineapples, a ball placed on a table, and a multi-arm array that alleviates the pushing problem by enveloping objects.Source: DISCUSSION · p. 11
  • The design abstracts natural forms geometrically to reproduce rich kinematic and dynamic behavior with low complexity; adding suckers or independently controllable arms remains future work.The authors suggest these additions to more closely incorporate other grasping mechanisms and octopus-like control.Source: DISCUSSION · p. 11

CONCLUSION

SpiRobs combine bioinspired morphology and operation with a scalable design principle demonstrated from millimeters to meters. The robots handle varied objects with minimal sensing and actuation, including in three demonstrated applications.

CONCLUSION

  • SpiRobs demonstrate a scalable design principle with robots from millimeter to meter scales.The conclusion also describes complex movements and adaptability to objects that vary in size and shape, using two to three cables.Source: CONCLUSION · p. 11
  • With two to three cables and minimal sensing, SpiRobs handle objects that vary in size and shape.Source: CONCLUSION · p. 11
  • Demonstrated applications include a miniaturized robot for fragile samples, drone teleoperation, and a multi-arm gripper that functions via entanglement.Source: CONCLUSION · p. 11

EXPERIMENTAL PROCEDURES

The procedures describe the spiral-based body design, its fabrication and control, and the resources available for the study. The design relates its packed curvature and deformation to a discretized robot body.

Resource availability

  • Study materials can be produced from the procedures in Notes S3 and S5, and all study data are available in the article and supplemental information.Source: Resource availability · p. 12

Conceptual design of SpiRobs

  • The robot body is formed by discretizing and mirroring spiral-based units, while its radius of curvature changes linearly and its deformation ratio is independent of the segment considered.At b=0.22, the deformation ratio is 3.98; a multi-unit design addresses the large deformation required of a continuous homogeneous body.Source: Conceptual design of SpiRobs · pp. 12–13
  • The three-cable design curls in three-dimensional space by rotating a mirrored form and reshaping its cross-section to increase contact area.Source: Conceptual design of SpiRobs · p. 13

Fabrication of SpiRobs

  • The robots use 3D-printed TPU units, with adjacent units scaled by a fixed ratio; longer robots use dovetail-connected segments and UHMWPE actuation cables.Cable holes guide the cables, which are fastened at the tip with a fisherman knot.Source: Fabrication of SpiRobs · p. 13

Control terminals

  • Separate motor-control terminals serve two-cable robots, three-cable drone operation, and multi-SpiRob grasping.The drone terminal uses drone power to reduce system weight.Source: Control terminals · p. 13

Data analysis

  • Motor current, speed, and position are recorded, cable length is derived from motor rotor position, and high-speed video captures dynamic tasks.Source: Data analysis · p. 14

SUPPLEMENT INFORMATION

The supplemental information includes methods and tables, a set of figures, and videos covering robot designs and demonstrations.

SUPPLEMENT INFORMATION

  • Supplemental materials include Notes S1–S7, Tables S1–S2, and Figures S1–S15.Source: SUPPLEMENT INFORMATION · p. 14
  • Videos S1–S20 cover robot designs and demonstrations, including grasping, manipulation, dynamic motion, drone use, and multi-SpiRob grasping.Source: SUPPLEMENT INFORMATION · p. 14
  • The video list includes demonstrations of miniaturized SpiRobs, a three-cable robot, impact-resistant grasping, and whipping to grasp.Source: SUPPLEMENT INFORMATION · p. 14
Loading 2303.09861v2…