IEEE Robotics & Automation Magazine - December 2020 - 32

only from the base material but, more importantly, also from
the featured structures at the microscopic scale in terms of
shape, orientation, and arrangement.
Recently, metamaterials have been receiving increasing
attention in soft robotics. The involvement of metamaterials
may lead to paradigm shifts in the design of soft robots by
directly encoding the desired complex motion within the
material architectures, leading to conformable monolithic systems [42]. Currently, the rational design optimization of
metamaterials remains in its early infancy, and we focus here
on how various metamaterials provide new insight into the
design of soft robots.
As mentioned, fibers have features of anisotropy and
can be exploited to program motions of soft robots. By

Extension

Expansion

assembling fibers with user-defined patterns, general
anisotropies that span more directions can be achieved.
Textile fabrics are such examples whose anisotropies are
directly encoded by the weaving or knitting paths at the
stage of fabrication [Figure  4(a)]; they have been widely
used in wearable robotic devices for hand, ankle, and foot
rehabilitation [36], [43], [44]. Inspired by layered human
muscles, such as the transverse abdominis, Zhu et al. [45]
recently presented a new family of fluidic fabric muscle
sheets based on composite fabric structures that admitted
design options at multiple scales, adaptability to curved
structures, and large work densities. They also investigated
how combinations of the fabric type and stitch design
modulate the patterns of stretchability.

Core
Module

Protective
Shield

Pneumatic
Legs

Explosive
Actuator

0.1

1,000 (MPa)
Young's Modulus

(b)

Input 2: Complex Kinematics
Twist
α6, L6
α5, L5
α4 L 4
α3, L3
α2, L2

Bending

α1, L1

Single
Pressure
Inlet

Output:
Fiber Angles and Segment Lengths

Input 1: Actuator Models

(c)

(a)

Rest Shape
Rigid Region

Target 1

Target 2

Constructed
Chambers

Constructed Frame Structure

Material Optimization

Fabricated Pneumatic Soft Object

(d)
Figure 3. The material optimization. (a) The fibers can passively constrain the deformation of pneumatic tubes along user-defined
directions to program the output motions (left); one may take a prescribed motion (top right) as the input and optimize the design
parameters for actuators to replicate the desired motion upon pressurization (bottom right) [24]. (b) A combustion-powered robot made
of nine different materials in terms of modulus attains a smooth stiffness gradient [32]. (c) Soft and hard materials in a cantilever beam are
automatically designed using evolutionary algorithms to enable the beam to deflect in a user-defined profile [33]. (d) A systematic design
and fabrication framework for soft pneumatic matters with desired shapes upon pressurization. Taking a rest shape and target shapes of a
heart model as the input (left), the method first optimizes the chamber arrangement and frame structures (middle left) and subsequently
optimizes the material distribution (middle right) of the frames to reproduce the motion behavior of a beating heart (right) [34].

32

*

IEEE ROBOTICS & AUTOMATION MAGAZINE

*

DECEMBER 2020
IEEE Robotics & Automation Magazine - December 2020

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - December 2020
