![(a) Husky [8], (b) ANYmal [1], (c) CARMA 2 [6], (d) Corin [13], (e) Phantom Pro 4, (f) Matrice 600 Pro, (g) AVEXIS [2], (h) BlueROV.](https://www.researchgate.net/publication/336314697/figure/fig1/AS:811418114879488@1570468562407/a-Husky-8-b-ANYmal-1-c-CARMA-2-6-d-Corin-13-e-Phantom-Pro-4-f_Q320.jpg)
![Schematic representation of magnetic resonance coupling technology [55].](https://www.researchgate.net/publication/336314697/figure/fig2/AS:811418114863104@1570468562444/Schematic-representation-of-magnetic-resonance-coupling-technology-55_Q320.jpg)
Available via license: CC BY 4.0
Content may be subject to copyright.
Wireless Power Transfer
cambridge.org/wpt
Review Article
Cite this article: Cheah WC, Watson SA,
Lennox B (2019). Limitations of wireless power
transfer technologies for mobile robots.
Wireless Power Transfer 1–15. https://doi.org/
10.1017/wpt.2019.8
Received: 17 June 2019
Revised: 30 July 2019
Accepted: 21 August 2019
Keywords:
Robotic applications; Resonant magnetic;
Laser; Microwave
Author for correspondence:
Wei Chen Cheah, Department of Electrical and
Electronic Engineering, The University of
Manchester, Manchester, United Kingdom of
Great Britain and Northern Ireland.
Email: wei.cheah@manchester.ac.uk
© Cambridge University Press 2019. This is an
Open Access article, distributed under the
terms of the Creative Commons Attribution
licence (http://creativecommons.org/licenses/
by/4.0/), which permits unrestricted re-use,
distribution, and reproduction in any medium,
provided the original work is properly cited.
Limitations of wireless power transfer
technologies for mobile robots
Wei Chen Cheah , Simon Andrew Watson and Barry Lennox
Department of Electrical and Electronic Engineering, The University of Manchester, Manchester, United Kingdom
of Great Britain and Northern Ireland
Abstract
Advances in technology have seen mobile robots becoming a viable solution to many global
challenges. A key limitation for tetherless operation, however, is the energy density of batter-
ies. Whilst significant research is being undertaken into new battery technologies, wireless
power transfer may be an alternative solution. The majority of the available technologies
are not targeted toward the medium power requirements of mobile robots; they are either
for low powers (a few Watts) or very large powers (kW). This paper reviews existing wireless
power transfer technologies and their applications on mobile robots. The challenges of using
these technologies on mobile robots include delivering the power required, system efficiency,
human safety, transmission medium, and distance, all of which are analyzed for robots oper-
ating in a hazardous environment. The limitations of current wireless power technologies to
meet the challenges for mobile robots are discussed and scenarios which current wireless
power technologies can be used on mobile robots are presented.
1. Introduction
The demand for tasks beyond the confinements of a work cell requires the use of mobile
robots. This has seen the development of various kinds of mobile robots at an unprecedented
pace for operations in air, land, and water. Applications for such robots include oil and gas
refinery inspection [1], underwater mapping [2,3], radiation mapping [4,5], and nuclear
decommissioning [6,7].
The primary limitation for the deployment of mobile robots for an extended period of time
is its limited onboard battery capacity. While technological advances have increased the energy
capacity and power output of batteries, this capacity is still lacking for robots. Limited by this,
the options for mobile robots are to either return to a base station for charging, have its battery
replaced manually, or operate with a tether. The first two options result in downtime for the
robot while the third option presents additional challenges such as tether crossovers, limited
bending around obstacles, reduce in the robot payload from the tether system, all of which
limit the mobility and performance of the robot [8].
A possible solution to the limited capacity of batteries is the use of wireless power transfer
(WPT) technologies to provide power alongside, or replace, the robot’s onboard battery. This
solution has the potential to address the energy requirement of the robot without comprom-
ising on both the robot’s operational time and mobility associated with the options present
above. Other potential advantages include higher power transmission efficiency (PTE), elim-
ination of spark from contacts [9], and redundancy for tether.
The WPT literature is primarily focused on either high-power (kW) applications such as
electric vehicles, which typically operate with transmission distance of <0.3 m [10]orlow-
power (less than a few Watts) applications such as consumer devices at centimeter distances,
wireless sensor network (WSN) at kilometer distances [11] and medical implants at centimeter
distances [12]. However, the use of WPT for the mobile robots in this study requires mid
power (100 W) and transmission distances of 1–20 m.
The primary challenge of WPT here is to deliver the power required to a non-stationary
mobile robot without compromising the robot’s mobility and its mission. This is compounded
by additional challenges such as system efficiency, the device form factor (size, mass, ancillary
components), transmission distance, and medium. An additional consideration is human
safety in scenarios where human and robots work in the same environment.
The contribution of this study is to collate the state-of-the-art in WPT and analyze the
applicability of such technologies for mobile robots. In particular, the challenges of a WPT
system detailed above will be considered in detail for mobile robots. The mobile robots selected
for the case study in this research are a representation of the robots available in the literature
targeted for deployment in a hazardous environment. However, the outcome of this research is
applicable to mobile robots in general.
The rest of this paper is structured as follows: Section II pre-
sents the problem statement for this work, detailing the scenarios
and case study robots. Section III presents an overview of WPT
technology and recent work on WPT for mobile robots. Section
IV details the analysis of WPT technologies for the scenarios con-
sidered in this study. Section V discusses the results of the analysis
and its applicability for the robots considered in this study.
Finally, Section VI summarizes the work and sets the outlook
to the future.
2. Problem Statement
The problem statement on the use of WPT for mobile robots may
be formulated by first identifying the environments where mobile
robots are used, followed by the type of mobile robot used, and
finally the power transfer scenarios of concern in this study.
2.1. Deployment scenarios
Advances in technology have led to mobile robots being consid-
ered for deployment in environments which are hazardous to
humans to perform remote inspections and interventions. Such
environments include, but are not limited to, nuclear plants, off-
shore wind turbines, oil and gas platforms, offshore substations,
underground mining tunnels, and underwater pipes. These envir-
onments dictate the type of medium that the robot has to operate
in, typically either air or water.
Due to the hazardous nature of these environments, transmitting
power directly to these robots may not be possible and needs to pass
through different mediums. For example, a robot deployed in a
sealed room, common in nuclear facilities, will require power trans-
mission through reinforced concrete or lead glass. Robots deployed
for in-pipe inspection will require power transmission through
metal or plastic, while underwater robots may require power trans-
mission through the containing vessel which could be made of con-
crete or metal. The type of medium for power transmission is thus
broadly categorized as air, optically translucent (water and glass)
and optically opaque (concrete, reinforced concrete, and metal).
2.2. Mobile platforms
This section presents the state-of-the-art mobile robot which
represents the generic robot types for ground, aerial, and under-
water targeted for deployment in hazardous environments.
Figure 1 shows the robots considered in this study, while the para-
meters on the robot’s power and battery requirements are sum-
marized in Table 1.
Both wheel and legged robots of medium (ANYmal and
Husky) and small (Jackal and Corin) scale have been included
for ground-based robots as these platforms have gained signifi-
cant interest in the industry and research community due to
their availability or capability that previous platforms have not
been capable of achieving [8,14,15]. The AVEXIS and
BlueROV represent the mini- and small autonomous underwater
vehicle (AUV) scale robot which is more suitable for the hazard-
ous environment [16].
The two unmanned aerial vehicles (UAVs) selected here
represent the medium and large class for UAVs available in the
market. Medium size UAVs, such as the Phantom 4
1
, are typically
used with their onboard camera only with no reported payload
capability. The camera itself is sufficient to be used for remote
inspection and characterization [5,17]. The Matrice 600 Pro
2
represents the large class of UAV with payload capability. This
allows different sensors or manipulators to be mounted to it,
allowing for a larger array of tasks to be undertaken [4,18,19].
2.3. Power transfer scenarios
There are three different WPT scenarios for the robot:
1) Net increase: The power consumption of the robot, P
c
, is less
than the power received, P
R
. In this case, prolonged power
transfer will lead to a fully charged battery. The robot may
be in either operational or standby mode, the latter in which
the robot is stationary with negligible P
c
.
Fig. 1. (a) Husky [8], (b) ANYmal [1], (c) CARMA 2 [6], (d) Corin [13], (e) Phantom Pro 4, (f ) Matrice 600 Pro, (g) AVEXIS [2], (h) BlueROV.
1
https://www.dji.com/uk/phantom-4-pro/info#downloads.
2
https://www.dji.com/uk/matrice600-pro/info#specs.
2 W. Cheah et al.
2) Zero net: The difference between the net power consumption
and power received is zero (P
c
=P
R
), hence the energy of the
battery remains constant.
3) Net decrease: The power consumption is higher than the
power received, P
c
<P
R
. In this scenario, the battery energy
will eventually reach zero.
The first and second scenarios are preferred as these ensure
that the onboard battery is not depleted. A smaller battery pack
may be used on the robot in both these scenarios, allowing the
robot’s payload to be increased. This study will consider the
second scenario in increasing the robot’s operational time.
Therefore the required output power of the WPT corresponds
to the Power consumption column in Table 1.
An important factor in WPT is the transmission distance, d,
achievable for a desired power output. The following definitions
for transmission distance are used in this study as they are
more relevant to the applications:
1) Very short-range: d< 0.1 m
2) Short-range: 0.1 m ≤d≤1m
3) Mid-range: 1 m ≤d≤5m
4) Far-range: 5 m<d≤20 m
5) Very far-range: >20 m
The problem statement can thus be stated as the feasibility of a
WPT technology for delivering the power required through differ-
ent transmission mediums for a non-stationary mobile robot
operating within mid- to far-range distances. The form factor
(size and mass) of the WPT system should also be within payload
capability of the mobile robot. This research focuses on
end-to-end WPT system only, i.e. the power is transmitted dir-
ectly to the receiver on the robot without requiring relay systems,
whether stationary set up or mobile robots.
3. Related Work
This section reviews the available technologies for WPT and its
application in mobile robots. For each technology, a brief descrip-
tion of the underlying principles is first presented followed by its
application. For a more comprehensive review on the respective
technologies, the interested readers are referred to [20–23].
WPT can be broadly categorized as radiative and non-radiative
as shown in Fig. 2. The two techniques in the literature using
radiative power transfer are microwave power transfer (MWPT)
and laser power transfer (LPT). Non-radiative power transfer
technologies are inductive power transfer (IPT), magnetic reson-
ance power transfer (MRPT), capacitive power transfer (CPT),
and acoustic power transfer (APT). Each technology can be fur-
ther categorized into direct or background energy harvesting.
The difference between these two is that the former receives
energy from a source that has been set up with the purpose of
transmitting power to the receiver. On the other hand, the latter
harvest background energy that is a by-product of another process
and not targeted for powering devices, such as heat from a coolant
system and radio waves from wireless communication [24,25].
Table 2 summarizes the main characteristics of the different
types of technologies. Of interest are the system efficiency, trans-
mission distance, the ratio between transmission distance and
receiver diameter, R
tr
, maximum power transferred, and the
hazardous potential of these technologies.
3.1. APT
This technology utilizes sound for power transfer, typically in the
megahertz range (0.5–2.25 MHz) [23]. A transducer converts
electrical power to mechanical power by vibrating the active
area. The medium which the transducer is connected to, such
as air or wall, then resonates and propagates the vibration through
the entire medium. The propagated vibration that reaches the
Table 1. Robot parameters considered in this study.
System
Power consumption,
W
Operational time,
h
Battery energy,
Wh
Max. distance
travelled, m
Wired charging
time, h
Payload,
kg
Husky 160 3 480 10 000* 4 75
ANYmal 300 3 900 3600* 4* 10
CARMA 2 90 3 270 14 000* 4 20
Corin 36 1 36 10 1 0.5
Phantom 4
Pro
178 0.5 89 7000 0.56* 0.5
Matrice 600
Pro
198 0.5 99 5000 0.39* 6
AVEXIS 30 3* 90* 300 1–2* 1.5
BlueROV 133 2 266 300 1–25*
*Values inferred from datasheet or communication with authors.
Fig. 2. Wireless power transfer technology classification.
Wireless Power Transfer 3
receiver is then converted back from mechanical to electrical
power. This methodology is illustrated in Fig. 3.
Application for this technology is primarily in low-power bio-
medical implant devices (mW) [40]. Medium- and high-power
transfer of 62 W and 1 kW has been achieved although at non-air,
small transmission distance of 70 and 5 mm, respectively [41,42].
These are targeted toward sealed environments such as containers
with hazardous objects [42]. The maximum efficiency achieved
for transmission in the air was 4% at 0.045 m transmission dis-
tance [26]. Although the efficiency was lower compared to IPT
at 0.045 m, the efficiency trails off slower compared to IPT (2%
for APT compared to 0% for IPT at 0.12 m). To the author’s
best knowledge, this technology has not been used for powering
mobile robots.
3.2. CPT
The idea behind this technology is to separate the pair of capacitor
plates so that one acts as a transmitter, while the other acts as the
receiver. CPT is different in that it uses electric instead of mag-
netic field for transferring power. The coupling capacitance is
defined by
C=k1oA
d,(1)
where ε
o
= 8.85 × 10
−12
F/m is the permittivity of space, kis the
relative permittivity of dielectric between material, Ais the surface
overlap between the plates, and dis the separation distance
between the plates. Evidently, the coupling capacitance is small
due to ε
o
, thus requiring large metal plates and high voltage res-
onance operation for power transfer at larger distances.
This technology has been used on mobile robots as well
besides electric car charging [27,43]. In [44], mobile soccer robots
have been equipped with parallel capacitor plates for docking-
station type wireless charging. The proposed system has a form
factor of 0.03 × 0.09 m
2
and achieved 44.3% efficiency. In [45],
a stationary UAV was charged using CPT where the entire landing
pad acts as the transmitter. Their approach minimizes the need
for accurate alignment and achieved 50% efficiency for 12 W
power output. A similar approach was used in [46] with improve-
ments on the ancillary system, achieving a higher efficiency of
77% although at only 8 W of power output. Common to these
mobile robots applications for CPT are the small transmission
distance, typically in millimeter distances.
3.3. IPT
This technology relies on inductive coupling, which is achieved by
the electromagnetic induction between a primary and secondary
coil. An AC current passing through the primary coil generates
a magnetic field which then induces voltage across a secondary
coil. An example of such a system is the power transformer
where there are no direct connections between the primary and
secondary coil. The magnetic field, B, at any point in space, d,cre-
ated by the primary coil is defined by the Biot–Savart’sLawas
B(d)=
m
o
4
p
C
IdI×d
|d3|,(2)
where μ
o
is the permittivity of free-space and Iis the current
through the conductor at section d
I
. As seen from this relation-
ship, the magnetic field is inversely proportional to distance,
B∝1/d
3
, causing the power transfer to drop significantly over lar-
ger distances.
The use of this technology is considered matured and has been
used in short-range consumer products such as wireless tooth-
brush, medical implants, and cellphones [12]. Application of
this technique to mobile robots is limited to situations where
the robot is within close proximity to the transmitter. The use
of a power floor mat allows for a large charging area for single-
[9] and multi-robot scenarios [47,48]. In [9], the entire mat con-
sists of only a single-layer transmitter coil and multiple pickups
were used on the receiver to achieve more regular power output
across the mat.
In [47], the coils on the transmitter mat have been designed for
higher power distribution at a fixed diameter so that similar
power can be transferred to multiple robots. In [48], multiple
transmitter coils were arranged in a grid to achieve large coverage
and dynamic charging in an arena. For underwater robots, guide,
and locking mechanisms have been used to align the coils,
arranged in an outer (transmitter) and inner (receiver) loop,
and to hold the robot in place [49–51].
3.4. MRPT
This technology is considered a special case of IPT whereby strong
electromagnetic coupling is achieved by operating at the reson-
ance frequency of the coils. This principle of operation can be
Table 2. Summary of the different WPT in air.
APT CPT IPT LPT MRPT MWPT
Max. efficiency 4% [26] 90% [27] 98% [28] 14% [29] 85% [30] 62% [31]
Max. distance, m 0.05 [26] 0.3 [27]7[32] 1000 [33]3[34] 1600 [35]
R
tr
4.6 [26] 0.5 [27] 3.5 [32] 916 [33]6[36] 222 [35]
Power transferred mW [26]kW[27]kW[37]kW[33]kW[38]kW[35]
Potentially hazardous Yes Yes Yes Yes Yes Yes
Fig. 3. Basic methodology of AET technology [39].
4 W. Cheah et al.
achieved using two (similar to inductive coupling) or more coils
(see Fig. 4)[52,53], where having more intermediary coils
increases the transmission distance. There are two principles of
operation, namely power delivered to load (PDL) and PTE [54],
where a trade-off on either the power delivered or the system effi-
ciency is observed.
The transmission efficiency for MRPT follows a similar trend
to IPT where it trails off with 1/d
3
. High system efficiency is
achieved by selecting high-performance subsystems such as the
power supply [56], coil configuration [57], and coil material
[58]. Due to the coupling effect, the power transfer does not fol-
low the transmission efficiency curve. The coils may be wound
around a ferrite core to increase the mutual coupling [59] or air
core for compact and lightweight design although at the expense
of lower coupling [53,60].
While a large body of research in this technology is focused
toward electric vehicles [61], this technology has also been used
on mobile robots operating in air, ground, and underwater. In
[62], a four-coil MRPT system was used to power a small electric
helicopter with automated impedance matching for improved
transmission efficiency. In [63], the receiver coil was wound
around the UAV’s landing leg to keep the central area clear for
application-specific purposes. In [64], a 3D receiver coil was pro-
posed to address the alignment requirement. This set up allows a
single transmitter coil to be used while still achieving high effi-
ciency compared to multiple transmitters. Besides being used
for powering the robot, the feasibility of charging sensor nodes
from a UAV was shown in [65].
On ground, a docking station setup is used for charging Team
Air-K’s robot in the ARGOS challenge and the Docent robot
through a wall [66]. In [67], underground guide rails along the
navigation path allow robots traveling along these paths to be
charged dynamically [67]. Underwater power transfer using
MRPT has been proposed for an AUV [68], showing the effects
of underwater transmission compared to air for circular and spiral
coils. However, the proposed system has not been evaluated on an
actual platform.
3.5. MWPT
Microwave is used as the medium for transmitting energy in this
radiative technique. The underlying technology for this technique
is similar to wireless communication using radio frequency (RF).
Electrical energy converted into microwaves at the transmitter end
is transmitted either in an isotropic or beam-forming manner.
The former is commonly used in communications and in WSN
which allows a larger area coverage for simultaneous wireless
information power transfer applications while the latter is typic-
ally used for transmitting to a single point, such as solar power
satellites. The receiver then converts the microwave received
back into electrical energy.
The potential damaging effect from exposure to high-power
microwave radiation limits the application of this technology,
where living tissues exist, to low-power applications with the
safety regulation set at 1 mW/cm
2
[69]. Hence sensors in WSN
are typically low powered (<1 W) while high-power base station
(>1 GW) is out of bounds for human entry. The high-power
density transmission for both short and far-range distance has
been demonstrated before the turn of the century [70].
However, the transmission area is a human-free zone and the
antennas are typically large and immobile.
Applications of MWPT for mobile robots include micro UAV
with mW power requirements [71], proof-of-concept mini airship
at 0.83 W [72], a charging mat area for ground robots [73], a low-
power 4 W rover [74], and for pipe inspection robot [75]. To
adhere with safety regulation, the power transmitted was either
kept low [71], thus limiting the power output, or required the
use of waveguide to contain the radiation [73,75].
3.6. LPT
This technique uses radiative electromagnetic field, similar to
MWPT, except that the wavelengths used are near the visible or
infrared spectrum. Figure 5 shows the schematic representation
of an LPT system. The transducer produces a highly collimated
beam (beam director) which the receiver, termed as laser power
converters (LPC), then converts back to electrical power. The use
of solar panels for the conversion was used in early works but
this was inefficient as solar panels were designed to operate over
a large spectrum while laser beams are monochromatic. Higher
efficiency has been achieved by designing monochromatic-photo-
voltaic cells for a specific wavelength [76]. The short wavelength
used for LPT allows for longer transmission distance but factors
such as moisture and dust particles in the transmission medium
will attenuate the transmission efficiency [77].
Applications for LPT have been demonstrated for very far-range
transmission on a smartphone [79] and UAVs with varying sizes
[80,81]. In [80], commercially available components were used
for the LPT system to achieve around 0.3 W power output at 1 m.
In [81], the power delivered was sufficient for a 1 kg UAV to retain
flight for 12 h with rough estimates of power delivery of 100 W.
4. Analysis
This section analyzes the feasibility of the different WPT tech-
nologies in terms of power output, system efficiency, device
form factor, transmission distance, and medium for the mobile
robots selected in subsection B. In this analysis, only the use of
direct power transfer is considered. This is due to the low energy
available from background energy harvesting, typically in mW or
μW[22], which is well below the power transfer requirement for
the robots considered in this study. Meaningful energy harvesting
from nuclear radiation is also not considered as the radiation at
this level will damage the electronic components [82].
Given that there is a range of robots selected in this case study,
only the analysis for the Husky is shown here for brevity pur-
poses. However, the same analysis is used for the rest of the robots
and the results of these are discussed in Section V. From the prob-
lem statement earlier in Section II, the WPT system requirements
for the Husky robot are:
Fig. 4. Schematic representation of magnetic resonance coupling technology [55].
Wireless Power Transfer 5
1) Power output of 160 W.
2) Power transmission through different mediums.
3) System efficiency of at least 10%.
4) Coverage to cope with the robot’s motion.
5) Minimum receiver weight (<75 kg).
6) Maximum receiver dimension of 0.3 × 0.3 m.
The power output and receiver weight are based on the speci-
fications of the robot detailed in Table 1. The maximum receiver
dimension has been selected such that it can be mounted on the
rear side of the robot. The system efficiency of LPT is used as a
benchmark here since this technology has shown one of the high-
est transmission efficiency at very far-range transmission distance.
4.1. Transmission range
The feasibility of the six available WPT is considered with respect to
the transmission distance, power efficiency, and form factor. Table 3
shows the range of limitations for the different technologies.
4.1.1. Very short-range
APT is the only technology which is limited to very short-range
operation. It also has the lowest efficiency and is therefore not
considered a viable solution for mobile robot WPT.
4.1.2. Short-range
CPT is restricted to lower-end of short-range operations only
(a maximum transmission distance of 0.3 m). This limitation is
due to the small capacitance arising from the permittivity of
space, thus resulting in small R
tr
. The capacitance is increased
by increasing the capacitive plate dimensions and reducing the
transmission distance, both of which are challenging as the
receiver dimension is limited on a mobile robot.
4.1.3. Mid-range
MRPT is in the mid-range operations but has a higher R
tr
com-
pared to IPT, as shown in Table 2. The diameter of the coil in
MRPT is one of the limiting factors in transmission distance,
although improvements may be achieved by increasing the num-
ber of turns to increase the coupling factor between the coils. The
coils may also be designed to have square spirals on printed cir-
cuit boards for maximizing the coil area.
4.1.4. Far- and very far-range
The recent progress in IPT allows far-range transmission distance
to be achieved at the lower limit of this range. This is achieved
using a long ferrite core for the magnetic flux lines between the
transmitter and receiver to be linked [83]. However, the transmit-
ter used in [83] is large and bulky, resulting in a large R
tr
.
Radiative power transfer technologies, namely MWPT and
LPT, are preferred for far and very far-range distance as the trans-
mission efficiency is generally much higher than non-radiative
techniques and does not require coupling, which limits the
transmission distance. However, the primary limiting factor for
achieving high efficiency is the conversion from electrical to elec-
tromagnetic energy and vice versa, which is relatively low com-
pared to non-radiative techniques. Other drawbacks for
radiative technologies include line-of-sight (LOS) requirement
which requires accurate tracking system and the potentially
hazardous effect on humans within its radiative field.
4.1.5. Analysis
Although IPT has been shown to achieve a higher transmission
range compared to MRPT, the R
tr
for MRPT is higher compared
to IPT, indicating that higher transmission range can in fact be
achieved using MRPT. Hence, MRPT, MWPT, and LPT are inves-
tigated further in the next section for mid- and far-range trans-
mission distance.
Fig. 5. Schematic diagram of LPT technology [78].
Table 3. WPT range limitations.
APT CPT IPT MRPT LPT MWPT
Very short-range X X X X X X
Short-range X X X X X
Mid-range X X X X
Far-range X X X
Very far-range XX
6 W. Cheah et al.
4.2. Power transfer
This section analyzes the power transfer efficiency and power out-
put of MRPT, MWPT, and LPT technologies through the air,
optical opaque, and optical translucent mediums.
4.2.1. MRPT
The three-coil structure will be used in this analysis since this
structure has been shown to provide a higher efficiency and trans-
mission distance compared to the two- and four-coil arrange-
ments [52,57]. The power output and transmission efficiency,
η
trans
, for the three coil structure is described by
Pout =V2
in
R2
T2
QLM
[(1 +LM)(1 +SM)+T2
Q]2,(3)
h
trans =T2
QLM
[(1 +LM)(1 +SM)+T2
Q](1 +LM)
RL
RL+R4
,(4)
where TQ=
v
M23/
R2R3
√,LM=
v
M34/
R3RL
√,S
M
=R
s
/R
2
,M
23
and M
34
are the mutual inductance between the sending–receiv-
ing coil and the receiving–load coil, respectively.
The system efficiency is then obtained by
h
sys =
h
dc−tx
h
trans
h
rx−dc,(5)
where η
dc−tx
and η
rx−dc
are non-unitary due to losses such as
switching and rectification.
Impedance matching is used to increase the power output,
termed as PDL, or transmission efficiency, termed as PTE. The
optimal impedance for both these cases are
LM,PDL =T2
Q+SM+1
1+SM
,(6)
LM,PTE =
T2
Q+SM+1
1+SM
.(7)
The parameters of the three-coil structure are detailed in
Table 4. The values for L
2
and L
3
were calculated for a planar
spiral coil with an outer diameter of 0.3 m, an inner diameter
of 0.2 m, six turns, a wire diameter of 3 mm, and a pitch of
7mm [84]. The capacitance was selected to be similar to [60],
resulting in a resonant frequency of 8.63 MHz. The efficiency of
the driver and rectifier is assumed to be n
rx−dc
= 0.81 [85] while
the η
dc−tx
is assumed to be accounted for by R
s
.
The system efficiency and output power of using fixed load
and impedance matching for both PTE and PDL are shown in
Fig. 6. Both the impedance matching approach converges at
approximately 0.6 m. The P
out
and n
sys
are increased by up to
3.5 and 4.7 times, respectively, at distance compared to a fixed
load.
4.2.2. MWPT
The transmission field for MWPT may be broadly categorized as
reactive near-field, radiative near-field, and radiative far-field. The
first results in coupling, similar to MRPT [86], while the third is
non-viable due to the low transmission efficiency at this distance
[86]. Thus the permissible transmission distance for MWPT is
0.62
D3/
l
≤d≤2D2
l
,(8)
where dand λcorrespond to the aperture and wavelength,
respectively.
The transmission efficiency within the permissible distance
without losses is described by [86]
t
=
l
2GTGR
4
p
d2,(9)
h
trans =1−e−
t
2,(10)
where G
T
and G
R
are the transmitter and receiver gain, respect-
ively. Finally, the system efficiency is calculated from (5).
The attenuation of electromagnetic radiation is described by
Bouguer’sLaw
I=I0e−
m
ad,(11)
Table 4. Parameters of three-coil magnetic resonance setup.
Transmitter Receiver
Component Value Component Value Component Value
R
s
1ΩR
2
1ΩR
L
50 Ω
R
1
1ΩL
2
27 μHR
3
0.25 Ω
L
1
27 μHC
2
12.6 pF L
3
1μH
C
1
12.6 pF k
23
0.1 C
3
340 pF
Fig. 6. Magnetic resonance using three coils with a fixed load, optimal PTE, and opti-
mal PDL.
Wireless Power Transfer 7
where I
0
is the initial intensity, μ
a
is the linear attenuation coeffi-
cient, and dis the transmission distance. The μ
a
is dependent on
the electromagnetic wavelength, temperature, and object proper-
ties [87]. The system efficiency including the attenuation loss is
to multiply (5) with I/I
0
from (11).
Using (11), an estimation of the propagation of RF through the
air, fresh water, glass, and concrete with attenuation coefficients of 0,
2.1584 [88], 4.6 [89], and 62 [89]/m, respectively, is shown in Fig. 7.
The n
sys
of MRPT through air has been added for comparison.
4.2.3. LPT
The laser setup is based on an end-to-end LPT system [90]. The
power at each stage of the system for a defined P
out
and η
sys
is cal-
culated as
Pin =Pout/
h
sys (12)
Ptx =Pin
h
dc−tx (13)
Prx =Ptx
h
trans (14)
Pheat =Prx −Pout.(15)
The thermal impedance requirement of the heat sink is
Zth =Tmax −Tamb
Pheat
,(16)
where T
max
and T
amb
are the maximum permissible and ambient
temperatures.
Table 5 summarizes the power at each stage for P
out
= 160 W.
The area of the LPC required is 2.7 × 10
−3
m
2
for a laser intensity
of 60 kW/m
2
. Assuming that the beam transmitted is circular, the
corresponding diameter of the LPC is 60 mm.
Ignoring scattering effects and water moisture in air and glass,
the attenuation coefficients for air, pure water, and glass are
assumed to be 0, 1.678 [91], and 4.575
3
. The results are shown in
Fig. 8.Then
sys
of MRPT through the air has been added as com-
parison for transmission through water and glass since MRPT is
unaffected by these two materials, discussed further in subsection A.
4.3. Comparison of technologies in air
Table 6 shows the typical η
dc−tx
and η
rx−dc
conversion efficiency of the
three technologies selected for comparison. A constant
h
sys =11.76%and P
out
= 160 W are assumed here for LPT based
on the setup of [90] since the transmission loss, observed to be
inversely proportional to the laser power transmitted, is negligible at
the power considered in this study. The input power of the MWPT
system is assumed to be 1.36 kW, the same as LPT for comparison.
Figures 9 and 10 show the system efficiency against transmis-
sion distance and receiver diameter, respectively, of the three
technologies selected from subsection A.
5. Discussion
This section discusses the results of the analysis and draws on
recent advances in WPT technologies on possible improvements.
5.1. Propagation through mediums
5.1.1. Air
From Fig. 9, the transmission distance for MRPT and MWPT is
limited to 1.04 and 1.3 m, respectively, for a power output of
160 W. The corresponding η
sys
at this distance is lower for both
technologies compared to LPT. While the η
sys
for LPT is low at
short- and mid-range, the efficiency is approximately the same
even for long-range distances of up to 1 km [33].
Increasing the resonant frequency of the MRPT does increase
the transmission distance, although at only <0.1 m at 50 MHz.
Operating at higher frequency also incurs losses in the power elec-
tronics circuit. Hence, MRPT tends to operate around the
100 kHz to 10 MHz range [60,94].
5.1.2. Optically translucent
MRPT has been shown to be insensitive to non-metallic obstacles
[53]. The transmission efficiency underwater for saline water was
shown to decrease by 5% compared to in air while no decrease
was recorded for pure and tap water [95]. The loss is attributed
to the dissolved salt increasing the dielectric losses. In [96], insu-
lating the coils from direct contact with the seawater shows
Fig. 7. Microwave transmission efficiency through the air, fresh water, glass, and
concrete.
Table 5. Power at each stage of LPT and thermal impedance requirement for
the Husky.
P
in
P
tx
P
rx
P
heat
P
out
Z
th
1360.54
W
408.16
W
400
W
240
W
160
W
0.05°C/
W
Fig. 8. Laser transmission efficiency through the air, pure water, and glass.
3
Calculated for optical silica glass with transmittance of 90%, zero reflection and
sample thickness of 10 mm.
8 W. Cheah et al.
similar efficiency in air and underwater at atmospheric pressure.
They proceeded to show that the increase in pressure decreases
the transmission efficiency underwater at fixed transmission dis-
tance [96].
The transmission efficiency of MWPT and LPT through
mediums decreases according to the absorption coefficient, μ
a
,
in (11). A model of μ
a
is defined in [97]as
m
a=
v
me
√0.5
1+
s
ve
2
−1
1/2
.(17)
It can be seen that (17) is dependent on frequency (ω), surface
conductivity (σ), dielectric permittivity (ϵ), and permeability of
material (μ). The ϵis frequency dependent and may be described
by the Debeye or Jonschur models [98]. With each model consist-
ing of three or more parameters, there are altogether at least six
parameters that need to be identified for calculating μ
a
. The
dependency of these parameters on environment factors means
that the model is highly specific for a particular scenario.
In this study, the values for μ
a
are extracted from experimental
data in the available literature where possible, or approximated.
The μ
a
for pure water has been identified in [91] for the wave-
lengths considered in this study. The μ
a
for glass was extracted
from the 20 mm glass thickness in [89] and is assumed to be
linear.
Both LPT and MWPT attenuate quickly in water and glass as
seen in Figs 8 and 7. The maximum transmission distance to
achieve 10% efficiency is reduced to 0.1 and 0.8 m, respectively,
for water, while transmission through glass is not feasible for
both radiative technology as the efficiency is <10%.
While the η
sys
through pure water for LPT is slightly higher
than MRPT between 0.7 and 1.1 m, it should be noted that the
μ
a
used here corresponds to pure water. It is expected that
the μ
a
will be higher in practical deployment scenarios, where
the conductivity of the water has a proportional relationship to
μ
a
[99]. Furthermore, losses arising from reflection, diffraction,
and dispersion will result in even lower η
sys
. These observations
are also applicable to MWPT.
5.1.3. Optically opaque
The opaque mediums considered are concrete, reinforced con-
crete, and metal. For MRPT, possible losses in the presence of
magnetic components arise from the dielectric, hysteresis, and
eddy currents [100]. The studies on MRPT through concrete
have shown mixed results, where the η
trans
either decreases
[101], remains the same [94], or increases [102]. This variation
is attributed to the concrete mix which may be paramagnetic,
thus affecting the coupling factor and the resonance frequency.
Hence, the magnetic permeability of concrete cannot be assumed
to be equivalent to free-space and needs to be evaluated
experimentally.
The power loss for transmission through reinforced concrete
tends to arise from hysteresis and eddy currents on the steel
bar. Aligning the coils such that it sits between the space of the
steel bars reduces the power loss by 10% when air core is used
on the coils [101]. This approach will require knowledge of the
steel arrangement within the reinforced concrete so that the coil
diameter may be designed to fit within the space. In practice,
the coil alignment will require trial-and-error to identify the opti-
mal position for maximizing the power transfer. However, using
coils wrapped around a steel core results in a negligible power
loss of 1% [94], thus eliminating the need for finding the optimal
position between the steel bar although at the expense of
increased mass arising from the steel core.
The frequency of the system also dictates the losses, where high
frequency tends to have lower penetrative capability [103] and
incurs higher eddy currents thus dominating the losses [100].
This type of loss was addressed in [94,103] by using low fre-
quency, 50–60 Hz for transmission through 5–10 mm-thick
metals and 100 mm reinforced concrete, respectively. In both
these cases then, the losses are dominated by the hysteresis.
Using low frequency results in a transmission efficiency drop of
approximately 10% for a single layer of steel bar in reinforced con-
crete. For the setup in this study, the maximum transmission dis-
tance is limited to around 0.7 m for P
out
= 160 W.
Similar to the transmission through optically translucent
mediums, the μ
a
for concrete and reinforced concrete have been
Table 6. Parameters of DC-TX and RX-DC used in the analysis.
Technology η
dc−tx
η
rx−dc
3-coil magnetic resonance 0.90 [85] 0.90 [85]
Microwave 0.80 [92] 0.75 [93]
Laser 0.30 [90] 0.40 [90]
Fig. 9. System efficiency and power transfer as a function of transmission distance for
magnetic resonance, laser, and microwave technology up to 20 m.
Fig. 10. System efficiency as a function of receiver diameter.
Wireless Power Transfer 9
estimated from [89], where the average values between the differ-
ent concrete variations were used. Since the attenuation coeffi-
cient for both concrete and reinforced concrete is approximately
the same, only concrete is shown here. MWPT through concrete
shows significant attenuation where even at 0.4 m (start of radia-
tive near-field), the n
sys
is almost zero. LPT was not considered for
transmission through opaque materials as the wavelength is
blocked by these materials and will result in power being dissi-
pated as heat on the surface.
5.1.4. Comparison
Table 7 summarizes the performance of MRPT, MWPT, and LPT
for the different mediums considered in this study.
The maximum transmission distance in the air is achieved
using LPT followed by MWPT and MRPT. For transmission
through non-air mediums, MRPT provides the maximum trans-
mission distance at approximately 1 m compared to the other
technologies. From Table 7, only LPT is capable of achieving
the far-range transmission required in this study. However, this
technology is only viable for transmission in air, limiting its appli-
cation to ground robots and UAVs. Although underwater robots
can rise to the surface to be charged using LPT, doing so results in
the robot deviating from its mission, which is no different from
returning to a docking station hence is not viable.
For underwater robots, MRPT is more feasible since the
attenuation underwater is negligible if the coils are isolated
from the surrounding water as shown in [96]. The attenuation
from the surrounding pressure on the transmission efficiency
does not need to be accounted for the two robots considered
here since the maximum depth of both these robots is 100 m,
well below the depth at which attenuation becomes non-negligible
[96]. It should be noted that the transmission range will be <1 m
since the receiver area for both the AVEXIS and BlueROV is smal-
ler than the Husky (the effect of the receiver area on the transmis-
sion range will be discussed shortly). Land and air robots that do
not have LOS, such as operating behind a glass or concrete wall,
will have to rely on MRPT. Similar constraints on the range
and mobility of underwater robots and land/air robots are
observed.
It is understood that the approximations of μ
a
are a limitation
to this study as these values may vary widely. While the results in
[89] show a linear relationship between the concrete thickness
and receiver gain, a non-linear relationship is observed between
the glass thickness and receiver gain. Hence, in-situ measure-
ments will be required to determine μ
a
for the wavelength used
and subsequently the transmission efficiency through these
mediums. Despite the approximation used, the fundamental limi-
tation of propagation through a non-air medium has been high-
lighted here for the use of radiative power transfer technologies.
5.2. Receiver requirements
A typical receiver setup consists of a power converter, ancillary
electronics, and subsystems such as battery management system
and control system (allows communication between the transmit-
ter and receiver for tasks such as, but not limited to, power trans-
fer initiation and termination, power output required, impedance
matching). The focus in this section is on the required power con-
verter and its ancillary components for MRPT, MWPT, and LPT,
used on the robots highlighted in subsection B.
5.2.1. MRPT and MWPT
The dependency between the η
sys
and the receiver diameter, shown
in Fig. 9, means that the η
sys
for MRPT and MWPT will decrease
with the diameter of the receiver area. This decrease means that
robots with smaller receiver area availability will have lower system
efficiency for the same transmission distance. Power transfer of
MRPT is further limited by the heat generation on the coils arising
from conduction, hysteresis, and eddy current losses [104]. The
need for a cooling system, typically in the form of passive heat
sink, will increase the mass and volume of the system.
5.2.2. LPT
From Fig. 9, the η
sys
for LPT is independent of the receiver size
(assuming that temperature of the LPC is maintained). Hence,
the area of the receiver may be minimized to the same size
as the collimated beam while still achieving the same efficiency.
The LPC size is scaled according to the power requirement as
each cell on the LPC has a limited power output. The LPC
required for the Husky is only 2.7 × 10
−3
m
2
, much smaller com-
pared to the other two technologies which occupy the maximum
permissible receiver size (0.3 × 0.3 m
2
).
An important consideration in the use of LPT is the LOS
requirement between the receiver and transmitter. The high inten-
sity of laser used is hazardous and hence should only be pointed to
the LPC which is able to withstand the laser intensity. This require-
ment will need a tracking mechanism, such as visual servoing [81]
or laser guard beams [79] as part of the control system for initiating
the terminating power transfer according to alignment.
The LPC requires a cooling system which serves to dissipate
the heat generated due to the low efficiency of LPC, and optimizes
the efficiency of the LPC by maintaining the temperature of the
LPC at an optimal temperature. Available heat sinks indicate
that the setup for Z
th
= 0.05°C/W requires a volume of 0.3 ×
0.1 × 0.07 m
3
with two fans (2.4 W each) attached. This setup is
approximately half the maximum permissible receiver size and
is expected to weigh <2 kg. Since the Husky has a payload capabil-
ity of 75 kg, the cooling setup is realizable for this robot. A similar
or scaled-down setup can be used on the Jackal which would meet
the power requirement without compromising the payload avail-
ability of the robot.
However, the LPC for the ANYmal, Corin, Phantom, and
Matrice has to be operated at non-optimal temperatures or under-
powered. This is due to the thermal impedance and low payload
capability of these robots, which prevent the use of an active cool-
ing system. The same cooling setup on the Husky can be used for
the Matrice, although this corresponds to taking up 33% of the
payload. The limited payload capability of both the Corin and
Phatom robot severely limits the use of a heat sink for the
required thermal impedance. A possible solution to this is to
use a lower incident power or intermittent charging to prevent
damage to the LPC from overheating.
Table 7. Comparison of WPT technologies transmission distance through
various mediums.
MRPT (m) MWPT (m) LPT (m)
Air <1 <1.2 up to 1000
Water <1 <0.8 <0.1
Glass <1 <0.4 <0.1
Concrete <1 ≈0–
Reinforced concrete <0.9 ≈0–
10 W. Cheah et al.
5.3. Human safety considerations
An important factor in the use of WPT is the health and safety
issues of using these technologies in the presence of humans.
The power requirement of the robots considered, coupled with
the low system efficiency of the three WPT technologies consid-
ered, results in high-power emission of either magnetic field or
electromagnetic radiation. Exposure of humans to these emissions
needs to adhere to standards set by government and non-
government organizations.
5.3.1. Magnetic field
The emission for electromagnetic WPT technologies needs to
adhere to the ICNIRP Guidelines for Limiting Exposure to
Time-varying Electric, Magnetic, and EMF (up to 300 GHz)
[105] and the IEEE Standard for Safety Levels with Respect to
Human Exposure to Radio Frequency EMF 3 kHz to 300 GHz.
Exposure limit for humans need to be identified using EM simu-
lation software such as COSMOL in [106] or SEM-CAD X in
[107] for MRPT. Rough estimates indicate a radius of 1 m from
the transmitter location is within the safety limit.
5.3.2. Electromagnetic radiation
For MWPT, the power density limit for the ICNIRP limit at 2.4
and 5.8 GHz are 50 W/m
2
. The power density is calculated as
S=PtxGt
4
p
d2,(18)
where P
tx
is the power transmitted, G
T
is the antenna gain, and dis
the transmission distance.
Assuming P
in
= 1.6 kW fornsys =10%at d= 3.12 m and G
t
=
25 dB, the power density is 808 W/m
2
, significantly above the per-
missible limit. This renders the transmission path inaccessible to
humans. A safety boundary also needs to be included due to beam
divergence. A possible solution is to use power waveforming,
where the RF power is transmitted to the receiver through
multipath, thus distributing the power density [108]. However,
the authors have their reservation on the feasibility of
dispersingP
tx
= 1.28 kW in a multipath fashion such that the
power density safety limit is adhered to.
The high power of LPT required in this study is categorized as
a Class 4 laser. This means that necessary safety measures must be
put in place to avoid human exposure, such as dedicated rooms,
remote viewing, or the use of safety glasses by anyone within
viewing the proximity of the laser [109]. An indicator of the
accessible range along the axis of the transmitted beam is the
nominal ocular hazard distance (NOHD), described by
NOHD =
u
−1
4
f
p
M−d2
out
,(19)
where θis the beam divergence, ϕis the laser radiant power, Mis
the maximum permissible exposure, and d
out
is the output beam
diameter.
Assuming θ= 0.001 rad, ϕ= 408 W, M=25W/m
2
, and d
out
=
0.060 m, the NOHD = 4.6 km. This means that the entire length
along the transmission path is considered hazardous. However,
the small beam divergence and output beam diameter mean
that only a small boundary area is required. To avoid injury
from accidental exposure to the high laser beam, guard beams
have been proposed to interrupt the beam transmission when
unidentified objects enter into the beam area [79]. Since a dis-
tance of only 12 mm between the guard beam and the high
power beam is required [79], the overall width for a distance of
20 m is only 84 mm. This width is relatively small compared to
the robot’s width. Although the use of guard beams eliminates
the risk of direct exposure to the beam, the electromagnetic radia-
tive effect will still require safety measures and boundaries to
avoid radiative damage to humans. Hence, the deployment of
LPT presents substantial health hazards and would benefit from
having no humans working within the same environment.
5.4. Practical applications
Table 8 summarizes the analysis and discussions, providing a
comparison of the different technologies in air and non-air
mediums, and the relevant design considerations. Alignment
here means that the transmitter and receiver are aligned axially,
while LOS means that the transmitter and receiver are aligned
with no obstacles (only air) between them.
Common to all three technologies is the need for a tracking
mechanism to meet the alignment requirement as the system effi-
ciency decreases with misalignment. A combination of visual and
RF signal tracking in [79] for LPT allows fast and accurate posi-
tioning compared to using only a single approach. This approach
is also suitable for MWPT. For MRPT, a possible approach to the
alignment problem, in addition to frequency or impedance
matching [60], is to characterize the power distribution over a
range of misalignment. The power distribution may then be
used to determine the misalignment vector in which the robot
can then make the necessary motions to achieve alignment.
From Table 8, it can be seen that among the three technologies,
MWPT has the poorest performance as it is only capable of
Table 8. Comparison of WPT technologies transmission distance through various mediums.
System considerations MRPT MWPT LPT
Transmission Mid-range Air Yes Yes Yes
Non-air Yes No No
Far-range Air No No Yes
Non-air No No No
Requires LOS No Yes Yes
Requires alignment Yes Yes Yes
Potentially hazardous Yes Yes Yes
Wireless Power Transfer 11
mid-range transmission in air. MRPT is capable of mid-range trans-
mission through mediums while LPT is capable of far-range trans-
mission in air. Hence, using MWPT in this context has no clear
benefits. For non-air transmission, MRPT provides the highest sys-
tem efficiency hence well suited to such a scenario. However, the
aim of extending the robot’s operational time without compromis-
ing on its mobility is unachievable as the transmission distance is
limited to mid-range. Despite this limitation, a benefit of using
MRPT arises in deployment scenarios where the access is sealed
or there are no feasible docking stations in the operational environ-
ment for the robot, subject to the medium being <1 m thick.
For transmission in air, LPT is the better choice in terms of
transmission distance. However, the primary limitation of this tech-
nology is the LOS requirement. Deployment scenarios are likely to
have obstacles present, thus the use of LPT will require significant
supporting structures (beam relays or overhead laser beaming) to
be put in place. These supporting structures present additional chal-
lenges such as visual servoing, strategic placement position for full
coverage, design, and mounting of the systems. In scenarios
where the deployment of these supporting structures is unavailable,
the robot’s mobility will be severely limited to a straight line only.
With regards to the aim of extending a mobile robot’s oper-
ational time without compromising on its mobility, none of the
current available WPT technologies is able to meet this aim with-
out requiring significant supporting infrastructure. Despite the
limitations of WPT technologies, there are special scenarios
where using WPT presents a clear advantage, such as far-range
LOS or sealed access requiring through-wall power transmission.
The state of the current technology is able to support WPT in
both these scenarios and is possible for immediate deployment.
6. Conclusion
The application of WPT technologies on mobile robots has been
reviewed in this paper and the feasibility of state-of-the-art WPT
technologies for a select case of robots operating in different
environments analyzed. There are six WPT technologies, APT,
CPT, IPT, LPT, MRPT, and MWPT. Among these, LPT, MRPT,
and MWPT are able to achieve mid- to far-range transmission
distance with a smaller receiver area compared to the other
three. The system performance through different transmission
mediums and receiver requirements has been analyzed. LPT
showed the highest system efficiency at far-range in air transmis-
sion while MRPT shows the highest system efficiency for non-air
transmission mediums which include water, glass, concrete, and
reinforced concrete. The safety considerations for mid-power
transmission of these three technologies have been considered,
in which all three present themselves as hazardous even at mid-
power. The practicality of using existing WPT technologies, con-
sidering the challenges of delivering the power required at mid- to
far-range distances for non-stationary mobile robots, is limited as
reviewed in the literature and analysis in this research. Similar to
existing applications of WPT for mobile robots, scenarios which
WPT can be readily used are near- to short-range distances, simi-
lar to the docking station, and long-range with LOS.
Financial Support. This work was supported by UK Research and
Innovation through the Engineering and Physical Science Research Council
under grant number EP/P01366X/1.
Conflict of Interest. None.
Ethical Standards. None.
References
[1] Hutter M et al.(2017) ANYmal –toward legged robots for harsh envir-
onments. Advanced Robotics 31, 918–931.
[2] Nancekievill M et al.(2018) Development of a radiological character-
ization submersible ROV for use at Fukushima Daiichi. IEEE
Transactions on Nuclear Science 65, 2565–2572.
[3] Palomeras N, Nagappa S, Ribas D, Gracias N and Carreras M (2013)
Vision-based localization and mapping system for AUV intervention,
MTS/IEEE OCEANS, Bergen, Norway.
[4] Li B, Zhu Y, Wang Z, Li C, Peng ZR and Ge L et al.(2018)Use of
multi-rotor unmanned aerial vehicles for radioactive source search.
Remote Sensing (Basel) 10, 728.
[5] Connor DT et al.(2018) Application of airborne photogrammetry for
the visualisation and assessment of contamination migration arising
from a Fukushima waste storage facility. Environmental Pollution 234,
610–619.
[6] Bird B et al.(2019) Radiological monitoring of nuclear facilities: using
the continuous autonomous radiation monitoring assistance robot.
IEEE Robotics and Automation Magazine 26,35–43.
[7] West C et al.(2019) Development of a Debris Clearance Vehicle for
Limited Access Environments, UK-RAS Proceedings, Loughborough,
20–23.
[8] West C et al.(2019) A debris clearance robot for extreme environments,
towards autonomous robotic systems, London, UK, 148–159.
[9] Park C, Lee S, Cho GH, Choi SY and Rim CT (2014) Two-
dimensional inductive power transfer system for mobile robots using
evenly displaced multiple pickups. IEEE Transactions on Industry
Applications 50, 558–565.
[10] Choi SY et al.(2015) Advances in wireless power transfer systems for
roadway-powered electric vehicles. IEEE Journal of Emerging and
Selected Topics in Power Electronics 3,18
–36.
[11] Xie L, Shi Y, Hou YT and Lou WJ (2013) Wireless power transfer and
applications to sensor networks. IEEE Wireless Communications 20,
140–145.
[12] Kim HJ, Hirayama H, Kim S, Han KJ, Zhang R and Choi JW (2017)
review of near-field wireless power and communication for biomedical
applications. IEEE Access 5, 21 264–21 285.
[13] Cheah W, Khalili H, Arvin F, Watson S, Lennox B and Green P
(2019) Advanced motions for hexapods. International Journal of
Advanced Robotic Systems 16,1–13.
[14] Tong CH, Gingras D, Larose K, Barfoot TD and Dupuis É (2013) The
Canadian planetary emulation terrain 3D mapping dataset.
International Journal of Robotics Research 32, 389–395.
[15] Leigh A, Pineau J, Olmedo N and Zhang H (2015) Person tracking
and following with 2D laser scanners, IEEE International Conference
on Robotics and Automation, Seattle, WA, USA, 726–733.
[16] Watson SA (2012) Mobile Platforms for Underwater Sensor Networks.
PhD dissertation, The University of Manchester.
[17] Boudergui K et al.(2011) Development of a drone equipped with opti-
mized sensors for nuclear and radiological risk characterization, IEEE
International Conference on Advancements in Nuclear Instrumentation,
Measurement Methods and their Applications (ANIMA), Ghent,
Belgium.
[18] Suarez A, Jimenez-Cano AE, Vega VM, Heredia G, Rodriguez-
Castaño A and Ollero A (2018) Design of a lightweight dual arm sys-
tem for aerial manipulation. Mechatronics (Oxf) 50,30–44.
[19] Beachly E, Detweiler C, Elbaum S, Twidwell D and Duncan B (2017)
UAS-Rx interface for mission planning, fire tracking, fire ignition, and
real-time updating, IEEE International Symposium on Safety, Security
and Rescue Robotics, Shanghai, China, 67–74.
[20] Lu M, Bagheri M, James AP and Phung T (2018) Wireless charging
techniques for UAVs: A review, reconceptualization, and extension.
IEEE Access 6, 29 865–29 884.
[21] Barman SD, Reza AW, Kumar N, Karim ME and Munir AB (2015)
Wireless powering by magnetic resonant coupling: recent trends in
wireless power transfer system and its applications. Renewable and
Sustainable Energy Reviews 51, 1525–1552.
12 W. Cheah et al.
[22] Lu X, Wang P, Niyato D, Kim D and Han Z (2015) Wireless networks
with RF energy harvesting: a contemporary survey. IEEE Communication
Surveys & Tutorials 17,757–789.
[23] Roes MGL, Duarte JL, Hendrix MAM and Lomonova EA (2013)
Acoustic energy transfer: a review. IEEE Transactions on Industrial
Electronics 60, 242–248.
[24] Chen J et al.(2016) A thermoelectric energy harvesting system for
powering wireless sensors in nuclear power plants. IEEE Transactions
on Nuclear Science 63, 2738–2746.
[25] Shinohara N (2011) Power without wires. IEEE Microwave Magazine
12, S64–S73.
[26] Tseng VFG, Bedair SS and Lazarus N (2018) Phased array focusing for
acoustic wireless power transfer. IEEE Transactions on Ultrasonics,
Ferroelectrics, and Frequency Control 65,39–49.
[27] Lu F, Zhang H, Hofmann H and Mi C (2015) A double-sided
LCLC-compensated capacitive power transfer system for electric vehicle
charging. IEEE Transactions on Power Electronics 30, 6011–6014.
[28] Bati A, Luk PC, Aldhaher S, See CH, Abd-Alhameed RA and
Excell PS (2019) Dynamic analysis model of a class E2 converter for
low power wireless charging links. IET Circuits, Devices & Systems 13,
399–405.
[29] Ortabasi U and Friedman H (2006) A photovoltaic cavity converter
for wireless power transmission using high power lasers, IEEE 4th
World Conference on Photovoltaic Energy Conference, Waikoloa, HI,
USA.
[30] Akuzawa Y, Tsuji K, Matsumori H, Ito Y, Ezoe T and Sakai K (2015)
A 95% efficient inverter with 300-W power output for 6.78-MHz mag-
netic resonant wireless power transfer system, IEEE MTT-S
International Microwave Symposium (IMS), Phoenix, AZ, USA.
[31] Wan S and Huang K (2018) Methods for improving the transmission-
conversion efficiency from transmitting antenna to rectenna array in
microwave power transmission. IEEE Antennas and Wireless
Propagation Letters 17, 538–542.
[32] Choi BH, Thai VX, Lee ES, Kim JH and Rim CT (2016)
Dipole-coil-based wide-range inductive power transfer systems for wire-
less sensors. IEEE Transactions on Industrial Electronics 63, 3158–3167.
[33] Becker DE, Chiang R, Keys CC, Lyjak AW, Nees JA and Starch MD
(2010) Photovoltaic-concentrator based power beaming for space eleva-
tor application, in AIP Conference Proceedings, 271–281.
[34] Zhang X, Meng H, Wei B, Wang S and Yang Q (2018) An improved
three-coil wireless power link to increase spacing distance and power for
magnetic resonant coupling system. EURASIP Journal on Wireless
Communications and Networking 1, 131.
[35] Dickinson RM (1975) Evaluation of a Microwave High-Power Reception-
Conversion Array for Wireless Power Transmission, Technical Report.
[36] Chen WX and Chen ZP (2018) Optimization on the Transmission
Distance and Efficiency of Magnetic Resonant WPT System, CSAA/
IET International Conference on Aircraft Utility Systems, Guiyang,
China, 148–154.
[37] Cirimele V, Diana M, Freschi F and Mitolo M (2018) Inductive power
transfer for automotive applications: state-of-the-art and future trends.
IEEE Transactions on Industry Applications 54, 4069–4079.
[38] Bojarski M, Asa E, Colak K and Czarkowski D (2016) A 25 kW indus-
trial prototype wireless electric vehicle charger, in IEEE Applied Power
Electronics Conference and Exposition (APEC), Long Beach, CA, USA,
1756–1761.
[39] Awal MR, Jusoh M, Sabapathy T, Kamarudin MR and Rahim RA
(2016) State-of-the-art developments of acoustic energy transfer.
International Journal of Antennas and Propagation 2016, Article ID
3072528.
[40] Zaid T, Saat S and Yusmarnita Yusop NJ (2014) Contactless energy
transfer using acoustic approach –a review, in International
Conference on Computer, Communication, and Control Technology,
Langkawi, Malaysia, 2055–2061.
[41] Leung HF, Willis BJ and Hu AP (2014) Wireless electric power trans-
fer based on acoustic energy through conductive media, in IEEE
Conference on Industrial Electronics and Applications (ICIEA),
Hangzhou, China: IEEE, 1555–1560.
[42] Bao X et al.(2008) High-power piezoelectric acoustic-electric power
feedthru for metal walls, in Industrial and Commercial Applications
of Smart Structures Technologies, San Diego, California, USA.
[43] Lu F, Zhang H, Hofmann H and Mi CC (2018) A double-sided
LC-compensation circuit for loosely coupled capacitive power transfer.
IEEE Transactions on Power Electronics 33, 1633–1643.
[44] Hu AP, Liu C and Li HL (2008) A novel contactless battery charging
system for soccer playing robot, 15th International Conference on
Mechatronics and Machine Vision in Practice, Auckland, New
Zealand, 646–650.
[45] Mostafa TM, Muharam A and Hattori R (2017) Wireless battery char-
ging system for drones via capacitive power transfer, IEEE PELS
Workshop on Emerging Technologies: Wireless Power Transfer
(WoW) 2017. Chongqing, China: IEEE.
[46] Muharam A, Mostafa TM and Hattori R (2017) Design of power
receiving side in wireless charging system for UAV application, in Int.
Conf. on Sustainable Energy Engineering and Application, Jakarta,
Indonesia, 133–139.
[47] Chen LJ, Tong WIS, Meyer B, Abdolkhani A and Hu AP (2011) A
contactless charging platform for swarm robots, 37th Annual
Conference of the IEEE Industrial Electronics Society, Melbourne,
VIC, Australia, 4088–4093.
[48] Arvin F, Watson S, Turgut AE, Espinosa J, Krajník T and Lennox B
(2017) Perpetual robot swarm: long-term autonomy of mobile robots
using on-the-fly inductive charging. Journal of Intelligent & Robotic
Systems 92 (3–4), 395–412.
[49] Shi JG, Li DJ and Yang CJ (2014) Design and analysis of an under-
water inductive coupling power transfer system for autonomous under-
water vehicle docking applications. Journal of Zhejiang University
SCIENCE C 15,51–62.
[50] Haibing W, Kehan Z, Zhengchao Y and Baowei S (2016) Comparison
of two electromagnetic couplers in an inductive power transfer system
for Autonomous Underwater Vehicle docking application, OCEANS.
Shanghai, China: IEEE, 1–5.
[51] Lin R, Li D, Zhang T and Lin M (2018) A non-contact docking system
for charging and recovering autonomous underwater vehicle. Journal of
Marine Science and Technology 24, 902–916.
[52] Zhong WX, Zhang C, Liu X and Hui SY (2015) A methodology for
making a three-coil wireless power transfer system more energy efficient
than a two-coil counterpart for extended transfer distance. IEEE
Transactions on Power Electronics 30, 933–942.
[53] Kurs A et al.(2007) Wireless power transfer via strongly coupled mag-
netic resonances. Science 317,83–86.
[54] Hui SYR, Zhong W and Lee CK (2014) A critical review of recent pro-
gress in mid range wireless power transfer. IEEE Transactions on Power
Electronics 29, 4500–4511.
[55] Zhang Y (2018) Key Technologies of Magnetically-Coupled Resonant
Wireless Power Transfer. Singapore: Springer.
[56] Liu S, Liu M, Yang S, Ma C and Zhu X (2017) A novel design meth-
odology for high-efficiency current-mode and voltage-mode class-E
power amplifiers in wireless power transfer systems. IEEE
Transactions on Power Electronics 32, 4514–4523.
[57] Kiani M, Jow UM and Ghovanloo M (2011) Design and optimization
of a 3-coil inductive link for efficient wireless power transmission. IEEE
Transactions on Biomedical Circuits and Systems 5, 579–591.
[58] Sun K et al.(2018) An overview of metamaterials and their achieve-
ments in wireless power transfer. Journal of Materials Chemistry C 6,
2925–2943.
[59] Kürschner D, Rathge C and Jumar U (2013) Design methodology for
high efficient inductive power transfer systems with high coil position-
ing flexibility. IEEE Transactions on Industrial Electronics 60, 372–381.
[60] Sample AP, Meyer DA and Smith JR (2011) Analysis, experimental
results, and range adaptation of magnetically coupled resonators for
wireless power transfer. IEEE Transactions on Industrial Electronics
58, 544–554.
[61] Mi CC, Buja G, Choi SY and Rim CT (2016) Modern advances in
wireless power transfer systems for roadway powered electric vehicles.
IEEE Transactions on Industrial Electronics 63, 6533–6545.
Wireless Power Transfer 13
[62] Yamakawa M, Shimamura K, Komurasaki K and Koizumi H (2014)
Demonstration of automatic impedance-matching and constant power
feeding to and electric helicopter via magnetic resonance coupling.
Wireless Engineering and Technology 5,45–53.
[63] Campi T, Cruciani S and Feliziani M (2018) Wireless power transfer
technology applied to an autonomous electric UAV with a small sec-
ondary coil. Energies 11, 352.
[64] Han W, Chau KT, Jiang C, Liu W and Lam WH (2019) Design and
analysis of quasi-omnidirectional dynamic wireless power transfer for
fly-and-charge. IEEE Transactions on Magnetics 55,1–9.
[65] Griffin B and Detweiler C (2012) Resonant wireless power transfer to
ground sensors from a UAV, Proceedings - IEEE International
Conference on Robotics and Automation, Saint Paul, MN, USA,
2660–2665.
[66] Cho I-K, Kim S-M, Moon J-I, Yoon J-H, Jeon S-I and Choi J-I (2013)
Wireless power transfer system for docent robot by using magnetic res-
onant coils, in International Symposium on Microwave, Antenna,
Propagation and EMC Technologies for Wireless Communications,
Chengdu, China, 251–254.
[67] Liu H et al.(2018) Dynamic wireless charging for inspection robots
based on decentralized energy pickup structure. IEEE Transactions on
Industrial Informatics 14, 1786–1797.
[68] Pereira MR, Santos HM, Pessoa LM and Salgado HM (2016)
Simulation and experimental evaluation of a resonant magnetic wireless
power transfer system for seawater operation, OCEANS. Shanghai,
China: IEEE, 1–5.
[69] Foster KR (2013) A world awash with wireless devices: radio-frequency
exposure issues. IEEE Microwave Magazine 14,73–84.
[70] Brown WC (1996) The history of wireless power transmission. Solar
Energy 56,3–21.
[71] Shimamura K et al.(2017) Feasibility study of microwave wireless
powered flight for micro air vehicles. Wireless Power Transfer 4,
146–159.
[72] Kim J, Yang SY, Song KD, Jones S, Elliott JR and Choi SH
(2006) Microwave power transmission using a flexible rectenna for
microwave-powered aerial vehicles. Smart Materials and Structures 15,
1243–1248.
[73] Okuyama T, Noda A and Shinoda H (2012) 30W wireless power trans-
mission to networked robots through 2D waveguide, International
Conference on Networked Sensing (INSS). Antwerp, Belgium: IEEE,
1–5.
[74] Kawasaki S (2011) Microwave WPT to a rover using active integrated
phased array antennas, European Conference on Antennas and
Propagation (EUCAP). Rome, Italy: IEEE, 3909–3912.
[75] Sato I, Shinohara N and Jodoi D (2018) Basic Study for Wireless
Power Transfer to a Pipeline Inspection Robot, IEEE Wireless Power
Transfer Conference. Montreal, QC, Canada: IEEE.
[76] Bett AW, Dimroth F, Lockenhoff R, Oliva E and Schubert J (2008)
III-V solar cells under monochromatic illumination, IEEE
Photovoltaic Specialists Conference, San Diego, CA, USA.
[77] Sahai A and Graham D (2011) Optical wireless power transmission at
long wavelengths, in Int. Conf. on Space Optical Systems and
Applications (ICSOS), Santa Monica, CA, USA, 164–170.
[78] Jin K and Zhou W (2019) Wireless laser power transmission: a review
of recent progress. IEEE Transactions on Power Electronics 34, 3842–
3859.
[79] Iyer V, Bayati E, Nandakumar R and Majumdar A (2017) Charging a
smartphone across a room using lasers. Interactive Mobile, Wearable
and Ubiquitous Technologies 1, 143.
[80] James J, Iyer V, Chukewad Y, Gollakota S and Fuller SB (2018) Liftoff
of a 190 mg laser-powered aerial vehicle: The Lightest Wireless Robot to
Fly, IEEE International Conference on Robotics and Automation,
Brisbane, QLD, Australia.
[81] Achtelik MC, Stumpf J, Gurdan D and Doth KM (2011) Design of a
flexible high performance quadcopter platform breaking the MAV
endurance record with laser power beaming, IEEE International
Conference on Intelligent Robots and Systems, San Francisco, CA,
USA, 5166–5172.
[82] Knight C, Davidson J and Behrens S (2008) Energy options for wire-
less sensors. Sensors 8, 8037–8066.
[83] Park C, Lee S, Cho GH and Rim CT (2015) Innovative 5-m-off-
distance inductive power transfer systems with optimally shaped dipole
coils. IEEE Transactions on Power Electronics 30, 817–827.
[84] Mohan SS, Hershenson M, Boyd SP and Lee TH (1999) Simple accur-
ate expressions for planar spiral inductances. IEEE Journal of Solid-State
Circuits 34, 1419–1424.
[85] Pinuela M, Yates DC, Lucyszyn S and Mitcheson PD (2013)
Maximizing DC-to-load efficiency for inductive power transfer. IEEE
Transactions on Power Electronics 28, 2437–2447.
[86] Shinohara N (2018) Wireless Power Transfer: Theory, Technology, and
Applications. United Kingdom: IET.
[87] Palik ED (1985) Handbook of Optical Constants of Solids, vol. 3. United
States: Academic Press.
[88] Park D, Kwak K, Kim J and Chung WK (2015) Underwater sensor
network using received signal strength of electromagnetic waves, in
IEEE Int. Conf. on Intelligent Robots and Systems., Hamburg,
Germany: IEEE, 1052–1057.
[89] Stone WC (1997) NIST Construction Automation Program Report No.
3: Electromagnetic Signal Attenuation in Construction Materials,
Technical Report 3.
[90] He T et al.(2014) High-power high-efficiency laser power transmission
at 100 m using optimized multi-cell GaAs converter. Chinese Physics
Letters 31, 104203.
[91] Pope RM and Fry ES (1997) Absorption spectrum (380–700 nm) of
pure water II integrating cavity measurements. Applied Optics 36,
8710–8723.
[92] Strassner B and Chang K (2013) Microwave power transmission: his-
torical milestones and system components. Proceedings of the IEEE
101, 1379–1396.
[93] Strassner B and Chang K (2002) 5.8-GHz circularly polarized rectify-
ing antenna for wireless microwave power transmission. IEEE
Transactions on Microwave Theory and Techniques 50, 1870–1876.
[94] Ishida H and Furukawa H (2015) Wireless power transmission through
concrete using circuits resonating at utility frequency of 60 Hz. IEEE
Transactions on Power Electronics 30, 1220–1229.
[95] Inamori M and Morimoto M (2017) Contactless Power Transfer in
Seawater, IEEE European Conference on Power Electronics and
Applications, Warsaw, Poland.
[96] Li ZS, Li DJ, Lin L and Chen Y (2010) Design considerations for elec-
tromagnetic couplers in contactless power transmission systems for
deep-sea applications. Journal of Zhejiang University SCIENCE C 11,
824–834.
[97] Balanis CA (1999) Advanced Engineering Electromagnetics. United
States: John Wiley & Sons.
[98] Bourdi T, Rhazi JE, Boone F and Ballivy G (2012) Modelling
dielectric-constant values of concrete: an aid to shielding effectiveness
prediction and ground-penetrating radar wave technique interpretation.
Journal of Physics D: Applied Physics 45, 405401.
[99] Jiang S and Georgakopoulos S (2011) Electromagnetic wave propaga-
tion into fresh water. Journal of Electromagnetic Analysis and
Applications 3, 261–266.
[100] Lee SH, Kim MY, Lee BS and Lee J (2019) Impact of rebar and con-
crete on power dissipation of wireless power transfer systems. IEEE
Transactions on Industrial Electronics 67, 276–287.
[101] Jang Y, Han JK, Baek JI and Moon GW (2017) Novel Multi-Coil
Resonator Design for Wireless Power Transfer through Reinforced
Concrete Structure with Rebar Array, IEEE 3rd International Future
Energy Electronics Conference and ECCE Asia, Kaohsiung, Taiwan,
2238–2243.
[102] Kim JM, Han M and Sohn H (2015) Magnetic resonance-based wire-
less power transmission through concrete structures. Journal of electro-
magnetic engineering and science 15, 104–110.
[103] Komurasaki K, Mizuno Y, Ishida J, Koizumi H and Yamakawa M
(2014) Wireless power transmission into a space enclosed by metal
walls using magnetic resonance coupling. Wireless Engineering and
Technology 5,19–24.
14 W. Cheah et al.
[104] Rim CT and Mi C (2017) Wireless Power Transfer for Electric Vehicles
and Mobile Devices. United States: John Wiley & Sons.
[105] International Commission on Non-Ionizing Radiation Protection
(ICNIRP) (1998) Guidelines for limiting exposure to time-varying elec-
tric, magnetic, and electromagnetic fields (Up to 300 GHz). Health
Physics 74, 494–522.
[106] Ding PP, Bernard L, Pichon L and Razek A (2014) Evaluation of elec-
tromagnetic fields in human body exposed to wireless inductive char-
ging system. IEEE Transactions on Magnetics 50, 1037–1040.
[107] Christ A et al.(2013) Evaluation of wireless resonant power transfer
systems with human electromagnetic exposure limits. IEEE
Transactions on Electromagnetic Compatibility 55, 265–274.
[108] Ku ML, Han Y, Lai HQ, Chen Y and Liu KJ (2016) Power waveform-
ing: wireless power transfer beyond time reversal. IEEE Transactions on
Signal Processing 64, 5819–5834.
[109] Smalley PJ (2011) Laser safety: risks, hazards, and control measures.
Laser Therapy 20,95–106.
Wei Chen Cheah received his M.Eng. degree in
mechatronics engineering at the University of
Manchester in 2015. He is currently working
toward his Ph.D. and is a research associate
with the Robotics for Extreme Environment
Group at the University of Manchester. His
main research interests are mobile robots and
wireless power transfer.
Simon Watson is a senior lecturer in robotic sys-
tems in the School of Electrical and Electronic
Engineering at the University of Manchester.
He obtained his M.Eng. in mechatronic engin-
eering in 2008 and his Ph.D. in 2012, both
from the University of Manchester. His research
focus is on mobile robots for the exploration
and characterization of hazardous and extreme
environments and active areas of research
include novel platform design, communications and localization and sensing
and navigation. His current research portfolio includes developing submers-
ible, wheeled, and legged robots for the nuclear industry (for the Sellafield
and Fukushima sites) and aerial robots for the power generation industry (off-
shore wind).
Barry Lennox is professor of applied control and
nuclear engineering decommissioning in the
Department of Electrical and Electronic
Engineering at the University of Manchester
and director of the Robotics and Artificial
Intelligence for Nuclear (RAIN) Research Hub.
He is an expert in applied control and its use
in robotics and process operations and has con-
siderable experience in transferring leading edge
technology into industry. He is the research lead for the nuclear engineering
decommissioning theme within the Dalton Nuclear Institute and the Robotics
in extreme environments theme within the School of Electrical and Electronic
Engineering. He leads the EPSRC Programme Grant: Robotics for Nuclear
Environments and the robotics work within the EPSRC TORONE project.
Wireless Power Transfer 15
