This article presents a brief overview of advent and future of the concept of artificial skin. Sensorized electronic skins are an essential element in wearable computing and co-robotic systems that engage in physical interaction with humans. Wearable computing is a rapidly growing sub-domain in the electronics industry and has the potential to transform how we work, play, get information, and interact with others. Due to its close connection to personal electronics and fashion, artificial skin technologies represent a good opportunity for science, technology, engineering, and math (STEM) outreach at the middle- and high-school levels. Wearable technology is a promising path to STEM education, since it is relevant to the daily experiences of teens and pre-teens, regardless of their social, cultural, and economic background. Middle-school students are an especially appropriate target, since they have adequate intellectual preparation for hardware prototyping but are still at the earliest stages of developing their academic interests.

Article

In order to match the versatility and robust mechanical properties of their natural counterparts, bio-inspired robots must be soft and elastically deformable [1]. This not only requires new classes of “artificial muscle” actuators but also artificial nervous tissue and skin to support stretchable electronic connectivity, sensing, and automation. In recent years, a broad range of materials, composites, and so-called “deterministic” micropatterned architectures have been introduced to support soft and stretchable electronic functionality [2]. These include conductive textiles, wavy circuits, graphene and nanotube films, filled elastomer composites (Figure 1), and liquid metal microfluidics (Figure 2). A common feature of these materials and architectures is their intrinsic mechanical compliance, stretchability, and low mass density. Such properties are especially important in wearable systems for personal computing and human motor assistance. By matching the mechanical properties of skin and nervous tissue, these materials can be mounted to the body or embedded in clothing without causing discomfort or injury. However, replacing rigid (and semirigid, i.e. flexible but inextensible) electronics with stretchable materials introduces new challenges in dynamic state estimation and feedback control. As they deform, soft circuit and sensing elements can exhibit dramatic changes in electrical conductivity, capacitance, or inductance. This electro-elasto coupling has important implications in circuit design and sensing functionality. Addressing these effects in soft bio-inspired systems represents a new and exciting opportunity to merge the mechanics of materials with sensing and controls.

graphic

graphic

Sensorized electronic skins are an essential element in wearable computing and co-robotic systems that engage in physical interaction with humans. These include strain and bend sensors for joint proprioception and gesture monitoring, pressure sensors for monitoring surface tractions, and tactile sensors for detecting light touch and data entry. A variety of sensing architectures and materials are currently being explored in the Soft Machines Lab (SML) at Carnegie Mellon University. This work incorporates conventional microelectronic and flex circuit technologies and also builds on principles and practices in the emerging fields of soft robotics and soft-matter engineering.

Artificial Skin for Soft Robots

Like their counterparts in nature, soft bio-inspired robots are primarily composed of lightweight elastic materials that can easily deform and adapt their shape to contacting surfaces [1,3]. This compliance and mechanical versatility allows soft robots to grasp delicate objects and navigate through tightly confined spaces with limited sensing and closed-loop control. Quadrupeds, undulators, and snake-like robots represent a particularly exciting class of soft bio-inspired systems that could have a revolutionary impact on unmanned, autonomous field exploration, particularly in unmapped or unstructured environments.

Examples of soft robot quadrupeds are presented in Figure 3. The quadruped limbs contain artificial muscles that are typically powered with pneumatics/inflation [4,5] or shape memory alloy. Ionic-polymer metal composites (IPMCs), dielectric elastomer actuators (DEAs), combustion, motor-driven cables, and series elastic actuators represent other common sources of actuation in bioinspired robotics [6-8]. The latter two are particularly popular for humanoid robots that engage in physical human interaction during assistive co-robotic tasks.

graphic

Just as soft robots depend on artificial muscles for actuation, they also require artificial skin and nervous tissue for sensing, wiring, and signal processing. As with technologies for wearable computing, these sensors and circuits must be soft, lightweight, and stretchable so that they don’t interfere with the mechanics of the host. Therefore it helps to turn to recent advancements in wearables in order to identify materials and sensing architectures for soft and elastic functionality.

Sensors for Next-Generation Wearables

Wearable computing is a rapidly growing sub-domain in the electronics industry and has the potential to transform how we work, play, get information, and interact with others. There are already thousands of wearable electronic devices on the market, with millions of users and projections for hundreds of millions more within the next 5-10 years [9]. Although most existing wearables are constructed from conventional solid-state microelectronics, there is increasing interest in incorporating novel materials and architectures. One popular approach is to use elastomeric composites in which rubber is embedded with percolating networks or films of micro/nano-scale conductive filler. Early efforts have focused on structured carbon black (see top insets in Figure 4 [10]) and metal powders, while more recent attention has shifted to co-polymer blends, silver flakes, carbon nanotubes, and graphene [2]. Another emerging technique is to interface solid state sensors with wavy circuits and other so-called “deterministic architectures” that achieve stretchable functionality through pre-buckling, wrinkling, helicity, or serpentine shapes (e.g. bottom insets in Figure 4 [11,12]).

graphic

In recent years, there has also been increasing interest in using soft microfluidics. This includes soft silicone elastomers that are embedded with microfluidic channels of ionic solution [13]. Stretching the elastomer or applying pressure to the surface alters the geometry of the channels and changes its electrical resistance. By monitoring this change in resistance, the elastic deformation can be inferred. An alternative is to replace the ionic fluid with liquid-phase metal alloy. As far back as the late 1940s, liquid metal (LM) electronics have been used for stretchable wiring and stretch sensing - most notably the mercury-based strain gauge developed by Reginald Whitney for biomechanical measurements [14]. In 2007, researchers at Harvard University discovered an approach to LM-based soft microfluidic electronics involving a safe alternative to mercury [15]. This is accomplished with eutectic gallium-indium (EGaIn; see inset of Figure 2). EGaIn is an alloy composed of 75% Ga and 25% In (by wt) that is liquid at room temperature, has high electrical conductivity (σ ∼ 3 x106 S/m), low toxicity, and negligible vapor pressure. As shown at the top of Figure 5, the liquid oxidizes in air and forms a nanometer-thin oxide skin that allows the droplet to hold its shape and be molded. Such oxidation and moldability allows EGaIn to be patterned with a variety of techniques based on soft lithography, additive manufacturing, and laser rapid prototyping [16]. It also enables wetting to soft silicones and urethanes so that the liquid channels can conform to stretch and other deformations (bottom Figure 5, [17]).

graphic

Lastly, there are opportunities to integrate conventional sensing technologies with soft materials and deterministic architectures. Referring to Figure 6, these hybrid technologies combine the sensing mechanisms (resistive, capacitive, inductive, optical) of solid-state devices with the soft, flexible, and stretchable functionality of elastomeric and thin-film technologies. Such integration is especially important for precision sensing and measurement of vitals for personal healthcare monitoring. For example, a pulse oximetry or glucose monitoring chip mounted to a wavy or soft microfluidic circuit can be embedded into a bandage or clothing and placed virtually anywhere on the skin. Current technical challenges include electrical interfacing between solid-state chips and soft-matter circuits and antennae as well as power for RF transmission.

graphic

Figure 6 also shows the difference in Young's modulus for different sensing technologies and how they compare with materials in nature. Young's modulus scales with the force required to elastically deform a material and is only defined for small (<1%) strains. In the case of soft and biological materials, there are other essential metrics for capturing elastic and rheological properties, including strain limit, shear modulus, and dynamic (shear) viscosity. Moreover, mechanical compliance and deformability represent just a few of the many properties that a sensor must exhibit for applications in soft robotics and wearable computing. Attention must also be given to the influence of materials selection and sensor architecture on dynamic range, sensitivity, bandwidth and allowable sampling rate, signal-to-noise ratio, and hysteresis.

Because of the large spectrum of materials, layups, geometries, and sensing mechanisms, there is an almost endless number of possible design combinations. In order to navigate this vast design space, it helps to have reliable theoretical models or computer-aided tools that accurately predict sensor performance under anticipated loading conditions. This includes multi-physics models that combine 3D elasticity, electrodynamics, and effective medium theory to predict the change in the electrical properties of a soft-matter circuit embedded in an elastically deforming medium. Figure 7 shows an example of one theoretical model for predicting the change in electrical resistance of an EGaIn channel as surface pressure is applied to the surrounding silicone. This strong agreement is achieved without data fitting and utilizes a combination of Ohm's law along with classical solutions in linear elasticity [18]. Models like this have the potential to dramatically reduce the number of design iterations required to select appropriate materials and geometries for soft microfluidic sensing.

graphic

Opportunities for Stem Outreach

Because of its close connection to personal electronics and fashion, artificial skin technologies represent a good opportunity for STEM outreach at the middle and high school levels. Wearable technology is a promising path to STEM education since it is relevant to the daily experiences of teens and pre-teens, regardless of their social, cultural, and economic background. Through hands-on projects that combine e-textiles and fashion accessories with rapid prototyping and open-source electronics hardware, youth can readily connect STEM topics with arts, crafts, and personal expression [19]. Middle school students are an especially appropriate target, since they have adequate intellectual preparation for hardware prototyping but are still at the earliest stages of developing their academic interests.

Members of SML have participated in several outreach events on the theme of materials for wearable sensing. This includes an interactive in which students can produce their own wearable touchpad and use it to move a cursor or play Tetris. Instead of custom laboratory- prepared materials, we provide the students with commercially available conductive fabrics and elastomers that they can cut out and bond to a soft insulating adhesive (e.g. 3M VHB tape). An example is shown in Figure 8, taken from a recent workshop hosted by the Gelfand Center for Service Learning and Outreach at CMU.

graphic

Acknowledgment

The author wishes to acknowledge the following members of SML for their contributions to the work presented in this article: Mike Bartlett, Eric Markvicka, Tong Lu, James Wissman, Andy Fassler, Hesham Zaini, Xiaonan Huang, Wenhuan Sun, and Ruthika. He also wishes to thank his collaborators on the iSkin project: Martin Weigel, Gilles Bailey, Antti Oulasvirta, and Juergen Steimle. Various aspects of this work has been supported by the following funding agencies: Office of Naval Research (#N00014120614; Program Manager: Dr. Tom McKenna), NASA (#NNX14AO49G; Research Collaborator: Dr. Bill Bluethmann), and DARPA (#N66001-12-1- 4255; Mentor: LTC Joseph Hitt, Ph.D.).

Reference

Reference
1.
C.
Majidi
,
Soft Robotics - A Perspective: Current Trends and Prospects for the Future
,
Soft Robotics
,
vol. 1
pp.
5
11
2013
.
2.
M. L.
Hammock
,
A.
Shortas
,
B. C.-K.
Tee
,
J. B.-H.
Tok
,
Z.
Bao
,
The Evolution of Electronic Skin (E-Skin): A Brief History, Design Considerations, and Recent Progress
,
Advanced Materials
,
vol. 25
pp.
5997
6036
2013
.
3.
D.
Rus
,
M. T.
Tolley
,
Design, fabrication, and control of soft robots
,
Nature
, vol.
521
pp.
467
475
2015
.
4.
R. F.
Shepherd
,
F.
Ilievski
,
W.
Choi
,
S. A.
Morin
,
A. A.
Stokes
,
A. D.
Mazzeo
,
X.
Chen
,
M.
Wang, and
G. M.
Whitesides.
Multigait soft robot
,
Proc. Nat. Acad. Sci. USA
, vol.
108
pp.
20400
20403
2011
.
5.
C.
Majidi
,
R. F.
Shepherd
,
R. K.
Kramer
,
G. M.
Whitesides
,
R. J.
Wood
,
Influence of Surface Traction on Soft Robot Undulation
,
IJRR
, vol.
32
pp.
1577
1584
2013
.
6.
J. D.
Madden
,
N. A.
Vandesteeg
,
P. A.
Anquetil
,
P. G. A.
Madden
,
A.
Takshi
,
R. Z.
Pytel
,
S. R.
Lafontaine
,
P. A.
Wieringa, and
I. W.
Hunter.
Artificial Muscle Technology: Physical Principles and Naval Prospects
,
IEEE J. Oceanic Eng
, vol.
29
pp.
706
728
2004
.
7.
J. M.
Hollerbach
,
I. W.
Hunter
,
J.
Ballantyne
,
A Comparative Analysis of Actuator Technologies for Robotics
,
The Robotics Review
, vol.
2
pp.
299
342
1992
.
8.
G.
Pratt
,
M. M.
Williamson
,
“Series Elastic Actuators,”
IEEE/RSJIROS
, vol.
1
pp.
399
406
1995
.
9.
“The Wearables Report,” BI Intelligence, Business Insider, July 27, 2015.
10.
M.
Weigel
,
T.
Lu
,
G.
Bailly
,
A.
Oulasvirta
,
C.
Majidi
,
J.
Steimle
,
iSkin: Flexible, Stretchable and Visually Customizable On-Body Touch Sensors for Mobile Computing’
ACM CHI
, pp.
2991
3000
2015
.
11.
J. A.
Rogers
,
Y.
Huan
,
A curvy, stretchy future for electronics
,
PNAS
, vol.
106
pp.
10875
10876
2009
.
12.
Dae-Hyeong
Kim
,
Nanshu
Lu
,
Rui
Ma
,
Yun-Soung
Kim
,
Rak-Hwan
Kim
,
Shuodao
Wang
,
Jian
Wu
,
Sang Min
Won
,
Hu
Tao
,
Ahmad
Islam
,
Ki
Jun Yu
,
Tae-il
Kim
,
Raeed
Chowdhury
,
Ming
Ying
,
Lizhi
Xu
,
Ming
Li
,
Hyun-Joong
Chung
,
Hohyun
Keum
,
Martin
McCormick
,
Ping
Liu
,
Yong-Wei
Zhang
,
Fiorenzo G.
Omenetto
,
Yonggang
Huang
,
Todd
Coleman, and
John A.
Rogers
,
Epidermal Electronics
,
Science
, vol.
333
pp.
838
843
2011
.
13.
K.
Noda
,
E.
Iwase
,
K.
Matsumoto
,
I.
Shimoyama
,
Stretchable liquid tactile sensor for robot-joints
,
IEEE ICRA
, pp.
4212
4217
2010
.
14.
R. J.
Whitney.
The measurement of changes in human limb-volume by means of a mercury-in-rubber strain gauge
,
Proc. Physiological Soc.
vol.
109
pp.
5P
6P
1949
.
15.
R. C.
Chiechi
,
E. A.
Weiss
,
M. D.
Dickey, and
G. M.
Whitesides.
Eutectic Gallium-Indium (EGaIn): A Moldable Liquid Metal for Electrical Characterization of Self- Assembled Monolayers
,
Angewandte Chemie
, vol.
120
pp.
148
150
2007
.
16.
I. D.
Joshipura
,
H. R.
Ayers
,
C.
Majidi
,
M. D.
Dickey
,
Methods to pattern liquid metals
,
J. Mater. Chem. C
, vol.
3
pp.
3834
3841
2015
.
17.
A.
Fassler
,
C.
Majidi
,
Soft-matter capacitors and inductors for hyperelastic strain sensing and stretchable electronics
,
Smart Mater. Struct.
, vol.
22
pp.
055023
2013
.
18.
C.
Majidi
,
Y.-L.
Park
(co-1st author),
R.
Kramer
,
P.
Berard
,
R. J.
Wood
,
Hyperelastic pressure sensing with a liquid-embedded elastomer
,
J. Micromech. Microeng.
, vol.
20
pp.
125029
2010
.
19.
B.
Barker
,
J.
Melander
,
N.
Grandgenett
,
G.
Nugent
,
Utilizing Wearable Technologies as a Pathway to STEM
,
Proc. Inter. Conf. Soc. Inform. Tech. Teacher Educ. (SITE)
, pp.
1770
1776
2015
.