Sensors and Actuators A 188 (2012) 2–8

Contents lists available at SciVerse ScienceDirect

Sensors and Actuators A: Physical

journal homepage: www.elsevier.com/locate/sna

Flexible flow sensor for large-scale air-conditioning network systems

M. Shikida a,∗ , K. Yoshikawa b , S. Iwai b , K. Sato b
a
Center for Micro-Nano Mechatronics, Nagoya University, Japan
b
Dept. of Micro-Nano Systems Engineering, Nagoya University, Japan

a r t i c l e i n f o a b s t r a c t

Article history: A patch-type flexible flow sensor was developed to precisely control the supply air in large-scale air-
Available online 9 December 2011 conditioning network systems in buildings. The proposed sensor enables us to reduce fruitless energy
consumption in these systems. The flow sensor was produced by applying photolithography onto a 25-
Keywords: ␮m-thick polyimide film. The four sensors were attached at 90◦ angles inside the surface of an 8-in.
Flow sensor duct, and the obtained outputs were averaged. The relationship between the sensor output and the flow
Air-conditioning network
rate followed the King equation under the 0–3000 m3 h−1 flow condition. The averaged sensor outputs
depended on the distance of the sensor position from the bent duct, and their values were slightly higher
than that obtained in a straight duct. Therefore, the conversion factor, which enables us to calculate the
flow rate values from the obtained sensor outputs in a bent duct, was derived for this paper.
© 2011 Elsevier B.V. All rights reserved.

1. Introduction wires suspended on a polyimide film. They mounted them onto a

Physical sensors, such as pressure, acceleration, and gyro sen-
sors, have been miniaturized and integrated onto Si wafers by
Micro Electro Mechanical Systems (MEMS) technologies since the
1970s [1–4]. They are now widely commercialized into the auto-
motive and amusement industries. The flow sensor fabricated onto
Si wafers also has a long history in MEMS miniaturization, and var-
ious types of MEMS flow sensors have been developed. C. Liu et al.
developed a thermal type of flow sensor to detect shear stress [5].
Zhe et al. used a micro-machined cantilever structure for measur-
ing the shear stress from the flow [6]. Generally, Si-MEMS flow
sensors have excellent space and time resolutions, and therefore,
they are used in semiconductor equipment and the fuel delivery
systems in the automotive industry for precisely controlling the
gas flow. Additionally, Unnikrishnan et al. proposed the MEMS-on-
tube assembly to simplify the packaging process in Si-MEMS [7,8].
They integrated Si-MEMS devices directly onto a glass tube, which
is compatible with a Swagelok® connector, by using a fusion bond-
ing. These and other Si-based flow sensors are summarized in Refs.
[9,10]. Si-MEMS flow sensors are normally assembled onto a rigid
flat board in the packaging process, because of the brittleness of
Si material. This means that difficulties arise when Si-MEMS flow
sensors are applied to curved surfaces.
MEMS flow sensors need to be more flexible in order to be used
in aero-space and medical equipment, and thus, resin material was
used as a substrate. For example, Buder et al. fabricated double hot-
flat surface, and concluded it was able to detect the flow separation
for a specific aerodynamic configuration [11]. Zhu et al. fabricated a
flow sensor directly onto a flexible printed circuit board with elec-
trical circuits, and mounted it onto the curved surface of a wing to
control an aircraft [12]. For medical applications, Li et al. used Kap-
ton film as a substrate, and they integrated pressure, temperature,
glucose, and oxygen sensors onto it, for blood analysis [13]. Naito
et al. fabricated a flow sensor on a 7.5-␮m-thick flexible polyimide
film, and they produced a miniaturized on-wall in-tube flow sensor
[14]. They used it to fabricate a catheter-type flow sensor for the
measurement of aspirated- and inspired-air characteristics in
human beings [15,16]. Ma et al. produced a flexible flow sensor by
using polyimide film for detecting the dynamic wave flow in water
channels [17]. To improve the sensitivity of a flow sensor, Buchner
et al. proposed a novel fabrication process, which enables us to pro-
duce flexible MEMS flow sensors using semiconductor materials
[18,19]. As stated above, the MEMS flow came into use for medical,
aircraft, and water channel applications, and the novel fabrication
process for flexible MEMS flow sensors was also proposed.
The aim of this work is to extend the MEMS flow sensor appli-
cations into new application fields, and thus, we are currently
introducing a flexible MEMS flow sensor into large-scale air-
conditioning network systems. The advantage of a flexible MEMS
flow sensor is that it can reduce the fruitless amount of energy
consumption by precisely controlling the supplied air in the system.
2. Flow control in air-conditioning system in building

∗ Corresponding author at: Froh, Chikusa, Nagoya 464-8603, Japan. Generally, air-conditioned air is produced in a conditioning sys-
E-mail address: shikida@mech.nagoya-u.ac.jp (M. Shikida). tem located outside the building, and delivered to each room via

0924-4247/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.sna.2011.12.002
 M. Shikida et al. / Sensors and Actuators A 188 (2012) 2–8
Fig. 1. Configuration of duct network and position for flow sensing in large-scale
air-conditioning systems.
duct network systems. The duct configuration for supplying air to
the necessary rooms is complicated and the system is composed
of a several bent ducts (Fig. 1). This is because the beam structures
are formed first in the building construction, and then the duct
network is mounted by using the limited amount of empty space
in the ceiling. The air supply to each room has to be controlled at
the ideal sensing point near the outlet port in order to reduce the
fruitless energy consumption of the air-conditioning system. This
means that the flow rate must be precisely measured downstream
of a bent duct.
We took into consideration the following points in the develop-
ment of a flow sensor for measuring the flow rate downstream of
a bent duct.
(1) We developed a patch-type flexible flow sensor to fit onto the
rounded inside surface of a duct. As shown in Fig. 2, it consists
of heaters on a thin polyimide film that can detect the flow
velocity and a flexible printed circuit. The film is supported by
silicone rubber with a hole in it to form a cavity under the heater
element to improve the responsibility.
(2) We decreased the thickness of the sensor structure to less than
525 ␮m to reduce the air flow resistance caused by the sensor
itself, and also put it on the inside surface of the duct. This is
because the flow velocity in the duct is minimized on the inside
surface.
(3) The air flow downstream of a bent duct is generally compli-
cated, because of the secondary flow caused in the bent duct
Fig. 2. Patch-type flow sensors and their in duct mounting position.
3
[20]. As a result, it is difficult to detect the flow rate by conduct-
ing only a single-point measurement. We attached four sensors
on the inside surface of the ducts to overcome this problem, and
averaged the flow rate at these four points to reduce the effect
of the secondary flow (Fig. 2).
Three different types of detection mechanisms, thermal
anemometry, calorimetric flow sensing, and time-of-flight sensing,
are normally used in a miniaturized MEMS flow sensor [9,10]. Ther-
mal anemometry detects the flow rate by determining the cooling
effect on a heated element. It is used in this study, because it has
a relatively large measurement range and the simplest of struc-
tures. The relationship between the electrical energy supplied to
the heater in a constant temperature mode and the flow velocity U
in a low Reynolds number flow can be expressed as [21]
Vh2
= (A + BU n ) · (Th − T0 ), (1)
Rh
where Vh and Rh are the voltage difference and electrical resistance
of the heater, respectively. A, B, and n are the constants depending
on the geometry of the heater element. If the heater were infinitely
long, n would be 0.5. However, n differs by 0.5 when a heater with
a finite length [9] is used. Th and T0 are the temperatures of the
heater and the body of fluid, respectively. Here, the square of the
voltage at the heater is proportional to the nth power of the flow
velocity. The details are explained in Refs. [9,10,21].
3. Fabrication
We fabricated the film-based sensor structure with polyimide
film (25-␮m thick) by using photolithography and thin-film depo-
sition. The details are as follows (Fig. 3):
(a) A polyimide (PI) Upilex film (produced by Ube Industries Ltd.)
was used as the substrate. The film thickness affects the length
of the response time. The heat capacity increases with the
increase in film thickness, and as a result, the length of the
response time increases with the increase in film thickness. To
shorten the response time, we chose a film thickness of about
25 ␮m for the fabrication process.
(b) The film was at first fixed on a glass wafer with thin silicone
rubber at a thickness of 0.5 mm to enable it to be handled in
the process that followed (Fig. 3(a)). A negative-type photore-
sist (ZPN1150-90 produced by Zeon Corporation), which was
specially developed for the lift-off process, was applied to the
film surface (Fig. 3(b)), and patterned with UV light to define
the shape of the thermal sensor (Fig. 3(c)). Then, sputtering
was used to deposit an Au/Cr film (Fig. 3(d) and (e)), and was
patterned by selectively removing the photoresist (lift-off pro-
cess). The Au and Cr thicknesses were 250 nm for the former and
10 nm for the latter. Cr film deposited by sputtering generally
induces tensile stress into thin PI film.
(c) The flexible film was manually peeled from the silicone rubber
on the glass wafer in an acetone solution (Fig. 3(f)). The flexible
film sensors fabricated on a 3-in. wafer are shown in the images
on the right in Fig. 3. The film sensor was 10 mm × 10 mm.
Two heaters were formed on one film sensor. One was for the
flow measurement and the other was for a backup. The typical
electrical resistance of a heater is 55.6 . Each film sensor was
manually cut with a knife.
(d) A flexible printed circuit was used as the electrical connection
from the film sensor (Fig. 4). At first, the film sensor was placed
on the printed circuit, and bonded using an adhesive. Then,
the electrical connection was manually made by using a silver
4 M. Shikida et al. / Sensors and Actuators A 188 (2012) 2–8

Fig. 3. Fabrication process of film sensor.

Fig. 4. Assembled patch-type flow sensor.


R1 R2
R4
Vdif
R3
Rh
Fig. 5. Electrical circuit for driving flow sensor.
paste. The electrical contacted area was finally covered with the
adhesive to increase the mechanical strength.
(e) To form a cavity structure for the thermal isolation, the
film sensor was placed onto a silicone rubber sheet with a
5.0 mm × 5.0 mm hole, and adhesive polyimide film was used
to affix the sensor. The thickness of the sheet was 0.5 mm. The
assembled flexible flow sensor is shown in Fig. 4.
4. Experiment
A flow sensor based on anemometry sensing detects the flow
velocity by measuring the change in electrical resistance caused by
the cooling effect of the flow. Therefore, the relationship between
the sensor temperature and the electrical resistance change was
measured. The resistance linearly increased as the temperature
increased, and a TCR value of 0.0025/K was obtained. As shown
in Fig. 5, a constant temperature circuit was used to shorten the
response time in the experiments.
4.1. Sensor characteristics
As the fundamental characteristics, the relationship between
the applied power to the sensor and the sensor temperature was
investigated by using commercially available infrared thermogra-
phy (TVS-700 produced by Nippon Avionics Co. Ltd.). The sensor
temperature increased with the increase in applied power, as
shown in Fig. 6. A temperature sensor value of 50 ◦ C, which corre-
sponded to the input power of 12.2 mW, was used in the following
flow rate measurements.
The sensor was mounted inside the surface of a tube with an
inner diameter of 3.0 mm to evaluate the response time perfor-
mance. Compressed air was used as the evaluation gas. It was
regulated with a commercially available mass flow controller
(KOFLOC, Model 3200, produced by Kojima Instruments Inc.) and
then introduced into the flow sensors. A three-port solenoid valve
60
55
Temperature (ºC)
50
45
40
35
30
2 4 6 8 10
Power (mW)
Fig. 6. Relationship between applied power and sensor temperature.
M. Shikida et al. / Sensors and Actuators A 188 (2012) 2–8 5
C
+25 V
OPAMP
-25 V
Fig. 7. Film sensor attached inside surface of 8-in. duct.
located between the sensor and the mass flow controller was used
for the response evaluation. We rapidly changed the flow rate by
switching the tube connection and the measured outputs of the
sensor (flow rate: 300 cc/min.). The sensor output rapidly increased
and then saturated, and a response time of 300 ms was obtained.
The obtained value was too short to detect the flow rate in the duct
because of the steady stream measurement.
4.2. Sensor output at duct
A duct with an 8-in. diameter was used, and four sensors were
attached onto the inside surface of the duct at different 90◦ angles
(Fig. 7). They were placed downstream of the bent duct at a dis-
tance L (Fig. 8). A commercially available ultrasonic flow-meter
(GF-2000 produced by SONIC CORPORATION) was placed in the
straight region upstream of the bent duct, and was used to measure
the standard volume flow rate in the duct. The flow-meter detects
the averaged flow rate by using the propagation time period of the
ultrasonic in the duct, and then calculates the volume flow rate
by using the correction factor and the diameter of the duct. The
following two characteristics were evaluated.
(a) Sensor output versus flow rate of straight duct
The relationship between the sensor output and the flow rate
in the straight duct was measured as a standard (Fig. 9(a)). The
applied flow rate ranged from 0 to 3000 m3 h−1 . The square of
the voltage in the heater was proportional to the nth power of
the flow velocity for the anemometer sensing, as shown in Eq.
(1). Therefore, we concluded that the obtained output signals
followed the King equation [21], which was known to show a
typical signal curve for a thermal-type flow sensor. The output
signals obtained form the four sensors were also coincident.
This means the fabricated flow sensor has excellent uniformity.
(b) Sensor output versus flow rate at bent duct
We then applied the sensors downstream of the bent duct.
The averaged sensor outputs depended on the distance of the
sensor position from the bent duct, and their values were
slightly higher than that obtained in the straight duct, as shown
in Fig. 9(b). Therefore, we derived a conversion curve that shows
the conversion factor, which enabled us to calculate the flow
rate values from the sensor output measured in the bent duct.
The detail procedures are as follows.
(1) To derive the conversion factor of the sensor outputs
obtained in between the bent and straight ducts, the out-
puts obtained in the bent duct were normalized by those
in the straight one (Fig. 10). The normalized values were
almost constant if the flow rate was larger than 500 m3 h−1 .
Therefore, we decided to use an averaged value that was
greater than 500 m3 h−1 as the conversion factor in the flow
12 14 16 18 region.
(2) From the derived conversion factor, we plotted the relation-
ship between the conversion factor and the distance to the
sensor position from the bent duct (Fig. 11).
6 M. Shikida et al. / Sensors and Actuators A 188 (2012) 2–8

Fig. 8. Experimental setup.

0.90

0.85

Ubent / Ustraight
0.80

0.75

0.70

0.65
0 500 1000 1500
Distance from end of bent duct to sensor (mm)

Fig. 11. Conversion curve of flow rates between straight and bent ducts.

From these results, we concluded that we can calculate the flow

rate value by using the obtained conversion factor, even if we mea-
sure the flow rate downstream of the bent duct.
We evaluated the rotation angle dependency of the sensor posi-
tion when the sensors are placed on the surface of the duct. The
experiments were performed for two different angles, 0◦ and 45◦ ,
as shown in Fig. 12, and both obtained results were compared. The
measurements were also done using two difference distances from
the end of the bent duct. The relationship between the flow rate
and the averaged sensor outputs from the four sensors is shown in
Fig. 9. Relationship between sensor output and flow rate. Fig. 12. The sensor outputs did not change even if the sensor was
placed onto the duct at different angles. From these results, we con-
cluded that the rotational angle in terms of the sensor attachment
1.0
does not affect the output when averaging the four sensor signals.
1210 mm 810 mm 610 mm 110 mm

0.9
5. Discussion
Flow Ratio

0.8
5.1. Calculation of flow rate for bent duct

0.7
We experimentally derived a conversion curve that shows the
conversion factor, which enabled us to calculate the flow rate values
0.6 from the sensor output measured in the bent duct. In the following
0 500 1000 1500 2000 2500 3000
paragraphs, we derived the equation, which enable us to calculate
Flow Rate (m 3 /h)
the flow rate from the sensor outputs obtained in the bent duct.
Fig. 10. Relationship between flow rate and normalized value of sensor output for The relationship between the electrical energy supplied to the
bent duct. heater in the constant temperature mode and the flow velocity U for
a low Reynolds number flow can be expressed as in Eq. (1). When
 M. Shikida et al. / Sensors and Actuators A 188 (2012) 2–8 7

Sensor Sensor 5.2. Compensation of temperature change

As shown in Eq. (1), the voltage of the heater principally depends

on the temperature of the air. This means that the sensor output
45º
changes with the change in air temperature in air-conditioning net-
work systems. As shown in Ref. [16], a similar type of flow sensor
was used for detecting the characteristics of a person breathing,
and investigated the effect of a change in gas temperature in the
2.2 sensor outputs. The relationship between the sensor output and
the flow rate was largely affected by the change in gas tempera-
2.0
ture change, even if its variation was 10 ◦ C. Therefore, we are now
1.8 trying to integrate a temperature sensor with a flow one to com-
Se nso r o utput (V2)

pensate for temperature variations. A metal pattern working as the

1.6 temperature sensor is formed on the same film as the flow sensor.
The temperature is detected by the change in electrical resistance
1.4
of the metal pattern. Basically, less than 1 mA of electric current
1.2 is used to reduce the self-heating effect on the temperature sens-
0° ing. The compensation for a change in temperature will be our next
1.0 challenge in the near future.
45°
0.8
0 500 1000 1500 2000 2500 3000 6. Conclusion
Flow rate (m3/h)
(a) L= 110 mm We developed a patch-type flexible flow sensor to precisely con-
trol the air supply in large-scale air-conditioning network systems.
2.0 The proposed sensor enables us to reduce the fruitless amount
1.8
of energy consumption in these systems. The following are the
obtained results.
Sensor output (V2)

1.6
(1) The patch-type flexible flow sensor was produced by applying
1.4
photolithography onto a 25-␮m-thick polyimide film.
1.2 (2) The relationship between the sensor output and the flow rate
0°
followed the King equation under the flow condition from 0 to
1.0 3000 m3 h−1 .
45°
(3) The averaged sensor outputs depended on the distance of the
0.8
0 500 1000 1500 2000 2500 3000 sensor position from the bent duct, and their values were
Flow rate (m3/h) slightly higher than that obtained in a straight duct.
(4) We derived the conversion factor, which enabled us to calculate
(b) L= 610 mm
the flow rate values from the obtained sensor outputs in the
Fig. 12. Effects of sensor position at different rotational angles. bent duct.

the gas temperature is constant, the flow velocity in the straight
duct is given by,
2/n
Ustraight = C + DVstraight
where Ustraight and Vstraight are the flow velocity and the voltage
difference in the straight duct, respectively. C, D, and n are the
constants obtained from the experimental results. From the exper-
imental results (Fig. 9), the relationship between the flow velocity
in the straight and bent ducts is simply expressed by the conversion
factor , as follows.
Ubent
= , (3)
Ustraight
where Ubent is the flow velocity in the bent duct. From the obtained
graph in Fig. 10, the conversion factor can be expressed by
 = 0.68 + (1.28 × 10−4 ) · L,
where L is the distance of the sensor position to the end of the bent
duct. From Eqs. (2)–(4), the equation calculating the flow rate in the
bent duct from the sensor output measured in it can be expressed
as
Ubent = {0.68 + (1.28 × 10−4 ) · L} · (C + DVstraight ).
2/n
We concluded that we can roughly estimate the flow rate from
the sensor output measured in the bent duct by using Eq. (5).
Acknowledgement
This research was supported by the COE for Education and
(2) Research of Micro-Nano Mechatronics from MEXT Japan.
References
[1] R.S. Muller, R.T. Howe, S.D. Senturia, R.L. Smith, R.M. White, Microsensor, 1990,
IEEE Press.
[2] H. Yamasaki, Handbook of Sensors and Actuators – Intelligent Sensors, 1996,
vol. 3, Elsevier.
[3] L. Ristic, Sensor Technology and Devices, Artech House, 1994.
[4] N. Maluf, An Introduction to Microelectromechanical Systems Engineering,
Artech House, 2000.
[5] C. Liu, J.-B. Huang, Z. Zhu, F. Jiang, S. Tung, Y.-C. Tai, C.-M. Ho, A micromachined
flow shear-stress sensor based on thermal transfer principle, J. Microelec-
tromech. Syst. 8 (1) (1999) 90–99.
[6] J. Zhe, V. Midi, K.R. Farmer Jr., A microfabricated wall shear-stress sensor with
(4) capacitative sensing, J. Microelectromech. Syst. 14 (1) (2005) 167–175.
[7] S. Unnikrishnan, H.V. Jansen, J.W. Berenschot, B. Mogulkoc, M.C. Elwenspoek,
Microfluidics within a Swagelok® : a MEMS-on-tube assembly, in: Tech. Digest.
MEMS 2009, January, Sorrento, Italy, 2009, pp. 324–327.
[8] S. Unnikrishnan, H.V. Jansen, J.W. Berenschot, B. Mogulkoc, M.C. Elwenspoek,
MEMS within a Swagelok® : a new platform for microfluidic devices, Lab Chip
9 (2009) 1966–1969.
[9] Y. Gianchandani, O. Tabata, H. Zappe, Comprehensive MEMS: 2. Flow Sensor,
Elsevier, 2008, pp. 209–272.
(5)
[10] M. Elwenspoek, R. Wiegerink, Mechanical Microsensors, Springer, 2001.
[11] U. Buder, R. Petz, M. Kittel, W. Nitsche, E. Obermeier, AeroMEMS polyimide
based wall double hot-wire sensors for flow separation detection, Sens. Actu-
ators A 142 (2008) 130–137.
8 M. Shikida et al. / Sensors and Actuators A 188 (2012) 2–8
[12] R. Zhu, P. Liu, X.D. Liu, F.X. Zhang, Z.Y. Zhou, A low-cost flexible hot-film sen-
sor system for flow sensing and its application to aircraft, in: Tech. Dig. IEEE
Micro Electro Mechanical Systems Conference, Sorrento, Italy, January, 2009,
pp. 527–530.
[13] C. Li, P-M. Wu, W. Jung, C.H. Ahn, L.A. Shutter, R.K. Narayan, A novel
lab-on-a-tube for multimodal monitoring of patients with traumatic brain
injury, in: Technical digest of The 15th International Conference on
Solid-State Sensors and Actuators, Denver, USA, June, 2009, pp. 1449–
1452.
[14] J. Naito, M. Shikida, M. Hirota, Z. Tan, K. Sato, Miniaturization of on-wall in-
tube flexible thermal flow sensor using heat shrinkable tube, in: Tech. Dig. IEEE
Micro Electro Mechanical Systems Conference, Tucson, USA, January, 2008, pp.
927–930.
[15] T. Yokota, J. Naito, M. Shikida, T. Kawabe, Y. Hayashi, K. Sato, Catheter type
of flow sensor for trans-bronchial measurement for lung functions, in: Tech-
nical digest of The 15th International Conference on Solid-State Sensors and
Actuators, Denver, USA, June, 2009, pp. 346–349.
[16] M. Shikida, J. Naito, T. Yokota, T. Kawabe, Y. Hayashi, K. Sato, A catheter-type
flow sensor for measurement of aspirated- and inspired-air characteristics in
bronchial region, J. Micromech. Microeng. 19 (2009) 105027, 9 pp.
[17] B. Ma, J. Ren, J. Deng, W. Yuan, Flexible thermal sensor array on PI film sub-
strate for underwater applications, in: Tech. Dig. IEEE Micro Electro Mechanical
Systems Conference, Hong Kong, January, 2010, pp. 679–682.
[18] R. Buchner, K. Froehner, C. Sosna, W. Benecke, W. Lang, Toward flexible ther-
moelectric flow sensors: a new technological approach, J. Microelectromech.
Syst. 17 (5) (2008) 1114–1119.
[19] R. Buchner, C. Sosna, M. Maiwald, W. Benecke, W. Lang, A high-temperature
thermopile fabrication process for thermal flow sensors, Sens. Actuators A
130–131 (2006) 262–266.
[20] Z. Tan, M. Shikida, M. Hirota, Y. Xing, K. Sato, T. Iwasaki, Y. Iriye, Characteristics
of on-wall in-tube flexible thermal flow sensor under radially asymmetric flow
condition, Sens. Actuators A 138 (2007) 87–96.
[21] L.V. King, On the convection of heat from small cylinders in a stream of fluid:
determination of the convection constants of small platinum wires with appli-
cations to hot-wire anemometry, Philos. Trans. R. Soc. Lond. A 214 (1914)
373–432 .
Biographies
Mitsuhiro Shikida received BS and MS degrees in electrical engineering from Seikei
University, Tokyo, in 1988 and 1990, respectively. He received a PhD from Nagoya
University in 1998. From 1990 to 1995, he worked at Hitachi, Ltd., Tokyo. In 1995,
he joined the Department of Micro-System Engineering at Nagoya University as a
research associate. He was an assistant professor from 1998 to 2004 and has been an
associate professor since 2004. He joined the Research Center for Advanced Waste
and Emission Management in 2001, the EcoTopia Science Institute in 2004, and
Department of Micro-Nano Systems Engineering, and is now the center for micro-
nano mechatronics from 2009, at Nagoya University. His research interests include
integration of micro-sensors and actuators for intelligent systems, micro-fabrication
of 3D microstructures for medical applications, and micro-total analysis systems for
biotechnologies. He is a member of the Institute of Electrical Engineers of Japan and
the Japan Society of Mechanical Engineers.
Kazuhiro Yoshikawa received a BS in mechanical engineering from Nagoya Univer-
sity, Japan, in 2010. He is currently working on an MS degree in micro nano systems
engineering at Nagoya University. His research interests include micro-fabrication
and its application for micro-flow sensor.
Satoshi Iwai received a BS, in mechanical engineering and an MS degree in
mechanical science engineering from Nagoya University, Japan, in 2009, and 2011,
respectively. His research interests include micro-flow sensor for medical applica-
tions.
Kazuo Sato received BS degree from Yokohama National University in 1970, and PhD
degree from The University of Tokyo in 1982. He worked with Hitachi Ltd. during
1970–1994. Since 1994, he is a professor of Micromachining and MEMS Labora-
tory, Dept. of Micro/Nano Systems Engineering, Nagoya University. He started MEMS
research in 1983, and published 110 journal papers, 30 review articles, and 11 book
chapters on MEMS. His research areas are micro/nano-physics in anisotropic etching
and mechanical properties of single crystal silicon, as well as applied microsystems
such as sensors and actuators. He co-chaired IEEE MEMS-97. He is the editor in Asia
of Journal of Micromechanics and Microengineering (IOP). He is a fellow of JSME and
JSPE, a senior Member of IEEJ, a member of JSPE and IEEE.