Research ArticleSOFT ROBOTS

A soft robot that navigates its environment through growth

See allHide authors and affiliations

Science Robotics  19 Jul 2017:
Vol. 2, Issue 8, eaan3028
DOI: 10.1126/scirobotics.aan3028
  • Fig. 1 Substantial lengthening from the tip with directional control enables a body to pass through a constrained environment and create a structure along its path of growth.

    (A) A body lengthens from its tip toward a target. Because only the tip moves, there is no relative movement of the body with respect to the environment (colored bands do not move). This results in the capability to move with no sliding friction through a constrained environment. As the tip moves, the body forms into a structure in the shape of the tip’s path. (B) Examples of biological systems that grow to navigate their environments. Neurons grow through constrained tissue to create structures that act as signal pathways. Pollen tubes lengthen through pistil tissue to build conduits to deliver sperm to the ovary. Sclerenchyma cells grow within the xylem and phloem to create supporting structures.

  • Fig. 2 Principle of pressure-driven eversion enables lengthening from the tip at rates much higher than those found in plant cell growth.

    (A) Implementation of principle in a soft robot. A pump pressurizes the body, which lengthens as the material everts at the tip. This material, which is compacted and stored on a reel in the base, passes through the core of the body to the tip; the rotation of the reel controls the length of the robot body. (B) Images of the lengthening body. The body diameter is 2.5 cm. (C) The relationship between lengthening rate (r) and internal pressure (P) shows a characteristic viscoplastic behavior: no extension below a yield pressure (Y) followed by a monotonic relationship between rate and pressure with a power term (n) close to 1. (D) Data show the relationship between rate and pressure above yield for the soft robot, worms with an everting proboscis (S. nudus), and a plant cell (Nitella mucronata). The extensibility φ (inverse viscosity) of a soft robot body is roughly seven orders of magnitude higher than that of the plant cell, resulting in a lengthening rate that is roughly five orders of magnitude higher. The extensibility of the soft robot body is slightly higher than the worm, which uses the same principle for lengthening.

  • Fig. 3 Principle of asymmetric lengthening of tip enables active steering.

    (A) Implementation in a soft robot uses small pneumatic control chambers and a camera mounted on the tip for visual feedback of the environment. The camera is held in place by a cable running through the body of the robot. (B) To queue an upward turn, the lower control chamber is inflated. (C) As the body grows in length, material on the inflated side lengthens as it everts, resulting in an upward turn (see Materials and Methods and Fig. 5 for details). (D) Once the chamber is deflated, the body again lengthens along a straight path, and the curved section remains. (E) A soft robot can navigate toward light using a tip-mounted camera. Inset: The view from the camera shows the target to the right. Electronically controlled solenoid valves inflate the control chamber on the left side of the robot body, resulting in the tip reorienting to the right and forming a right turn. (F) The target is straight ahead, and the robot steers straight. (G) The target is to the left, and the robot steers left. (H) The robot reaches the target. (I) Position of the target along the horizontal axis of the camera as the robot lengthens toward the target.

  • Fig. 5 Details of an implementation of a mechanism within the control chambers for selective lengthening of the sides of the soft robot.

    A series of latches are manufactured into the control chambers shown in Fig. 3 (A to D). Each latch crosses pinched material, such that when released, the side lengthens. There are four total states. State 1: When the control chamber is depressurized and the latch is on the side, the latch remains closed. State 2: When the control chamber is depressurized and the latch is at the tip, the latch remains closed. (When a control chamber is depressurized, the pressure from the main chamber keeps the latch closed regardless of whether the latch is on the side or at the tip.) State 3: When the control chamber is pressurized and the latch is on the side, the latch remains closed. State 4: When the control chamber is pressurized and the latch is at the tip, the latch opens. (When the control chamber is pressurized, the latch remains closed if it is along a side, due to the shape of the interlocking of the latch, but the latch opens if it is at the tip because the high curvature overcomes the interlocking.)

  • Fig. 4 Growth enables a soft robot to move its tip through constrained environments and to form 3D structures defined by the path of its tip.

    (A) A soft robot lengthens through various challenging constrained environments without active control. Instead, the robot passively deforms to navigate the obstacles. Yellow bodies, 2.5-cm diameter; clear bodies, 8-cm diameter. (B) The pressure required to lengthen through the gap remains relatively constant, despite vastly different surface properties of the material surrounding the gap and different displacements within the gap. Setup is shown in fig. S1. (C) The soft robot demonstrates the ability to lengthen into useful 3D structures.

  • Fig. 6 Overview of active steering control system.

    Hardware components and a physical depiction of the steering task are shown. Electrical signal formats are labeled in purple, and their semantic meanings are labeled in black.

Supplementary Materials

  • robotics.sciencemag.org/cgi/content/full/2/8/eaan3028/DC1

    Text

    Fig. S1. Experimental arrangement for collection of data shown in Figs. 2 and 4 and fig. S2.

    Fig. S2. Additional experimental results from tests to determine full model for soft robot lengthening.

    Fig. S3. Modeling of a helical antenna formed with a soft robot.

    Fig. S4. Extension of a soft robot body with preset pattern of branching.

    Fig. S5. Viscoplastic relationships for natural extending systems.

    Movie S1. Lengthening.

    Movie S2. Steering.

    Movie S3. Constrained environments.

    Movie S4. Forming structures from the body.

    References (40, 41)

  • Supplementary Materials

    Supplementary Material for:

    A soft robot that navigates its environment through growth

    Elliot W. Hawkes,* Laura H. Blumenschein, Joseph D. Greer, Allison M. Okamura

    *Corresponding author. Email: ewhawkes{at}engineering.ucsb.edu

    Published 19 July 2017, Sci. Robot. 2, eaan3028 (2017)
    DOI: 10.1126/scirobotics.aan3028

    This PDF file includes:

    • Text
    • Fig. S1. Experimental arrangement for collection of data shown in Figs. 2 and 4 and fig. S2.
    • Fig. S2. Additional experimental results from tests to determine full model for soft robot lengthening.
    • Fig. S3. Modeling of a helical antenna formed with a soft robot.
    • Fig. S4. Extension of a soft robot body with preset pattern of branching.
    • Fig. S5. Viscoplastic relationships for natural extending systems.
    • Legends for movies S1 to S4
    • References (40, 41)

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    • Movie S1 (.mp4 format). Lengthening.
    • Movie S2 (.mp4 format). Steering.
    • Movie S3 (.mp4 format). Constrained environments.
    • Movie S4 (.mp4 format). Forming structures from the body.

    Files in this Data Supplement: