Journal of Physics: Conference Series
PAPER • OPEN ACCESS
Energy Harvesting from Human Motion Using
Footstep-Induced Airflow
To cite this article: H Fu et al 2015 J. Phys.: Conf. Ser. 660 012060
 
View the article online for updates and enhancements.
Related content
Exotic cluster structure in light nuclei
N Itagaki, T Ichikawa, J A Maruhn et al.
-
UCAC5: New Proper Motions Using Gaia
DR1
N. Zacharias, C. Finch and J. Frouard
-
Coherent Optical Control of Molecular
Motion Using Polarized Sequential Pulses
Kohichi Itoh, Yasunori Toda, Ryuji Morita
et al.
-
Recent citations
A methodology for low-speed broadband
rotational energy harvesting using
piezoelectric transduction and frequency
up-conversion
Hailing Fu and Eric M. Yeatman
-
This content was downloaded from IP address 18.51.0.96 on 11/12/2018 at 18:04
      
Energy Harvesting from Human Motion Using 
Footstep-Induced Airflow  
H Fu1, R Xu2, K Seto2, E M Yeatman1 and S G Kim2 
1 Department of Electrical and Electronic Engineering, Imperial College London, 
Exhibition Road, London, SW7 2AZ, UK 
2 Department of Mechanical Engineering, Massachusetts Institute of Technology, 
Cambridge, MA 02139, USA 
 
E-mail: h.fu14@imperial.ac.uk 
 
Abstract. This paper presents an unobtrusive in-shoe energy harvester converting foot-strike 
energy into electricity to power wearable or portable devices. An air-pumped turbine system is 
developed to address the issues of the limited vertical deformation of shoes and the low 
frequency of human motion that impede harvesting energy from this source. The air pump is 
employed to convert the vertical foot-strike motion into airflow. The generated airflow passes 
through the miniaturized wind turbine whose transduction is realized by an electromagnetic 
generator. Energy is extracted from the generator with a higher frequency than that of 
footsteps, boosting the output power of the device. The turbine casing is specifically designed 
to enable the device to operate continuously with airflow in both directions. A prototype was 
fabricated and then tested under different situations. A 6 mW peak power output was obtained 
with a 4.9 Ω load. The achievable power from this design was estimated theoretically for 
understanding and further improvement. 
1. Introduction 
Wearable or portable devices are becoming increasingly popular in our daily lives, including smart 
watches, health monitors, radio frequency identification tags and emergency locators [1, 2]. However, 
they are generally powered by batteries that have a limited lifetime. Replacement or recharging is 
needed regularly, which is inconvenient in some situations. Harvesting wasted energy from human 
motion to power these devices is a potential alternative to conventional batteries, enabling the devices 
to be autonomous and fit-and-forget [3].  
Extracting energy from foot strike has drawn significant interest [4-6], as the energy from foot-
strike during walking is a plentiful and readily tapped wasted energy source [4]. However, it is always 
in a low frequency range and the deformation of shoe heel is quite limited, which exposes power limits 
for foot-strike energy harvesting. One interesting development to address these issues is a rotary arm 
converting the vertical linear motion of footsteps into rotation with a gearbox employed to magnify the 
rotational speed [5]. However, the rotary arm affects the shoe’s comfort, and the gear train increases 
the manufacturing complexity and system unreliability. Moro et al. proposed a piezoelectric shoe 
energy harvester [6]. Although the structure of the device is simpler, its performance is limited due to 
the low operating frequency and the restricted volume of piezoelectric material. In this paper, a 
flexible air pump is employed to convert the vertical footstep motion into airflow, and a miniature 
wind generator is developed to generate electricity from the airflow at high frequencies. This design 
PowerMEMS 2015 IOP Publishing
Journal of Physics: Conference Series 660 (2015) 012060 doi:10.1088/1742-6596/660/1/012060
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
      
overcomes the difficulties in foot-strike energy harvesting mentioned above with a simpler structure 
and an enhanced power output.  
2. Design and Operating Principle 
The schematic of the device is presented in Figure 1. It consists of an air bulb pump and a miniature 
wind turbine. The pump converts the vertical distortion of shoes into airflow, and the turbine generates 
electrical energy from the passing airflow. The pump cover is constructed from latex rubber, and a 
wave spring is adopted to restore the shape of the pump after each strike. In order to enhance the speed 
and pressure of the airflow, an amplification mechanism is achieved by designing a shrunken outlet on 
the pump. Its sectional area is 500 times smaller than that of the pump, which means the airflow 
velocity is magnified up to 500 times on entry to the plastic tube. The airflow then passes through the 
wind turbine and energy is collected from a miniature DC motor as a generator attached to the turbine 
impeller. This design allows the turbine rotor to operate at a higher frequency than that of human 
motion. Thus, frequency up-conversion is achieved by the air pump-wind turbine system, addressing 
the issue of low frequency of footstep and boosting the power output. 
The device is installed in the shoe heel, as shown in Figure 2. Due to the miniaturized size, the 
harvester can be embedded in a shoe heel unobtrusively. The support provided by the wave spring is 
stiff enough to ensure the wearer’s comfort. When a footstep is applied to the air cushion, the pump is 
deformed and air in the pump is pushed forward to the turbine. When the force is removed, the pump 
shape is restored by the spring force, and a reversed airflow is generated in the device, further 
activating the turbine. 
 
               
Figure 1. Schematic of the shoe energy 
harvester. 
Figure 2. Installation and composition of the 
harvester in a shoe. 
 
The inlet and outlet of the turbine is designed specially to ensure the airflow direction to be the 
same during and after each foot strike (Figure 3). The rotational direction of the turbine rotor, 
therefore, remains uniform during the entire motion cycle, which enables the continuous operation of 
the device during human motion.  
3. Prototype and Experimental Set-up 
A prototype of the harvester was built, as illustrated in Figure 4. An air pump was mounted on a sheet 
of stainless steel with the dimension of 120 mm × 60 mm × 13 mm. It is smaller than the foot heel of a 
shoe of size 9 (UK). A miniaturized wind turbine was fabricated using rapid prototyping with ABS 
materials. These two parts were connected by a plastic tube whose cross-sectional area (~Ø4 mm) was 
500 times smaller than that of the pump. The overall dimension of the turbine was Ø12 mm × 6 mm, 
and its transduction was achieved by a DC motor with the dimension of Ø7 mm × 15 mm. The air 
pump was actuated by normal footsteps from a person weighing 65 kg.  
 
PowerMEMS 2015 IOP Publishing
Journal of Physics: Conference Series 660 (2015) 012060 doi:10.1088/1742-6596/660/1/012060
2
      
    
Figure 3.  Design of the wind turbine with two 
generation cycles per foot strike.  
Figure 4.  Experimental set-up with a 25¢ US 
coin as reference.  
4. Results 
The device was tested under different situations. Figure 5 shows the open-circuit output voltage during 
walking and running. In walking situations (1 Hz), the turbine rotates intermittently, whereas it can 
rotate continuously during running (4 Hz). The continuous motion also proves the successful 
implementation of the inlet and outlet design (Figure 3) for the uniform rotational direction of the 
turbine during and after each foot strike.  
 
Figure 5.  Open-circuit voltage output for different motion situations. 
                    
Figure 6. Measured peak power output for a 
single footstep versus different load resistances. 
Figure 7.  Average power output with different 
footstep frequencies.  
Figure 6 illustrates the measured peak power output for a single footstep with varying load resistances. 
A 6 mW power output was obtained with a 4.9 Ω load which is identical to the motor’s internal 
resistance. Figure 7 depicts the power output against footstep frequency. The average power output is 
on the mW scale, which is enough to power certain wireless sensors monitoring the condition of a 
human body. The device was finally embedded into a running shoe as shown in Figure 8. More tests 
will be carried out on a treadmill in future work. 
PowerMEMS 2015 IOP Publishing
Journal of Physics: Conference Series 660 (2015) 012060 doi:10.1088/1742-6596/660/1/012060
3
      
 
Figure 8.  An integrated shoe energy harvester. 
5. Achievable Energy in Theory 
In order to understand the potential achievable power from this mechanism, the upper limit and the 
theoretically achievable power of using this specific method are estimated analytically, showing the 
potential improvement of the power output and the unimpeded nature of the harvester to human 
motion due to the marginal energy extracted.  
5.1. Upper Limit of Power Output 
The upper limit is estimated without considering the specific operating mechanism. It is defined as the 
work done on an enclosed air pump during a foot strike. Assuming it is an isotropic process and the 
temperature, T, is constant, the energy needed to compress the air by a ratio of λ can be calculated 
using the following equation. 
 ln( )W pdv nRT dv v nRT      (1) 
where p is the air pressure, v is the volume of the air, n is the number of moles of gas present and R is 
the ideal gas constant. 
Assuming that the striking happens in a gradual process, the momentum of human motion can be 
neglected. Therefore, for a person weighing m, the compressing ratio can be calculated using 
  
5
0
0 0 3 5
0
1 10 1
1.87637 7.2 10 1 10a
pv v p p
mg A p
          (2) 
where m is the mass of a person, g is the gravitational acceleration, Aa is the sectional area of the air 
pump and p0 is the initial pressure of the air in the air pump. 
In the test, the weight of the person was 65 kg. The maximum power that can be obtained from 
each foot strike by the air pump for a compressing and releasing process is 
 0
0
0.1 1ln 8.31 293.2 ln 6.55 J
22.4 1.87
amg A pW nRT
p
              
 (3) 
where the negative sign means the work done on the air. 
For a walking frequency of 1 Hz, the maximum power obtained by the air pump is  
 6.55 1 6.55 WP W T     (4) 
The estimated upper limit of power output using this transduction is similar to the estimation in [7]. 
5.2. Theoretical Achievable Energy of This Mechanism 
For this specific design, the compressed and accelerated airflow in the pump is used to drive an 
electromagnetic turbine. Assuming the airflow exiting the wind turbine at an increased speed with a 
PowerMEMS 2015 IOP Publishing
Journal of Physics: Conference Series 660 (2015) 012060 doi:10.1088/1742-6596/660/1/012060
4
      
modest pressure, the energy collected from the foot strike can be calculated from the kinetic energy of 
the airflow which is 
 21
2 a a
E m v  (5) 
where ma is the mass of the air and va is the velocity of the airflow exiting the turbine. In practice, 
wind turbines has a theoretical maximum power coefficient Cp of 0.593 [8]. According to the result 
from Figure 5, the working duration of each footstep on the air pump is around 0.1 s and the 
displacement of the air pump is about 10 mm. Hence, the velocity of the airflow entering the turbine 
can be up to 50 m/s considering the amplification factor (δ=500).  
The energy obtained from each foot strike on the specific air pump mentioned above is 
      2 201 1 0.593 0.12 0.06 0.013 1.29 500 0.1 0.0895 J2 2p aE C m v              (6) 
 The theoretically achievable power from 1 Hz foot strikes is 89.5 mW. Therefore, the efficiency of 
the current design at 1 Hz is ~0.28% (0.25/89.5). It indicates that the comfort or feel of shoes is not 
affected due to the marginal energy extracted, and that the power output can be increased by an order 
of magnitude or more with appropriate design. Apparently, the theoretically achievable power highly 
depends on the amplification factor that is adjustable in the design process. Further investigation will 
be conducted to optimize the structure. 
6. Conclusions 
In this paper, an in-shoe energy harvester is developed to harness foot-strike energy. An air-pumped 
turbine system is designed to solve the difficulties of harvesting energy from foot strike, including the 
limited vertical distortion and the low motion frequency. An amplification mechanism for airflow 
speed is achieved by designing the air pump and turbine inlet with different cross-section areas. The 
turbine is activated by airflow and operates with a high frequency, enabling the foot-strike energy to 
be harvested with a higher frequency. The turbine casing is designed specifically to enable the device 
to operate continuously with airflow from both directions. 
A prototype was fabricated and then tested under different situations. A 6 mW peak power output 
was obtained with a 4.9 Ω load. The achievable power from this design was estimated theoretically, 
showing the marginal energy extracted from human motion and the potential improvement of the 
power output. Future work will be on improving the efficiency of the system and exploring potential 
applications. 
Acknowledgements 
The work is supported by the China Scholarship Council (CSC), MIT-SUTD International Design 
Center and the Department of Electrical and Electrical Engineering, Imperial College. The authors 
would also like to thank the Imperial-MIT Global Fellows Programme for the travelling support. 
References 
[1] Mitcheson P D, Yeatman E M, Rao G K, Holmes A S and Green T C 2008 Proceedings of the 
IEEE 96 1457-86 
[2] C -Y Sue and N -C Tsai 2012 Applied Energy 93 390-403 
[3] Pillatsch P, Yeatman E M and Holmes A S 2014 Wearable and Implantable Body Sensor Networks 
(BSN), 2014 11th International Conference on, pp 6-10 
[4] Paradiso J A and Starner T 2005 Pervasive Computing, IEEE 4 18-27 
[5] Hayashida J Y 2000 MIT Media Lab 
[6] Moro L and Benasciutti D 2010 Smart Materials and Structures 19 115011 
[7] Starner T and Paradiso J A 2004 Low-power Electronics Design 1-35 
[8] Briand D, Yeatman E M and Roundy S 2015 Micro Energy Harvesting Advanced Micro and 
Nanosystems (Wiley-VCH) 
PowerMEMS 2015 IOP Publishing
Journal of Physics: Conference Series 660 (2015) 012060 doi:10.1088/1742-6596/660/1/012060
5