Materials Today: Proceedings xxx (xxxx) xxx

Contents lists available at ScienceDirect

Materials Today: Proceedings

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

Design and implementation of bio inspired hexapod for exploration

applications
T. Deepa a, S. Angalaeswari a,⇑, D. Subbulekshmi a, S. Krithiga a, S. Sujeeth a, Raja Kathiravan b
a
School of Electrical Engineering, Vellore Institute of Technology, Chennai 600127, Tamilnadu, India
b
Faculty of Mechanical and Industrial Engineering, Bahir Dar University, Bahir Dar, Ethiopia

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

Article history: With advances in Robotics, robots of different kinds are increasingly being used for scientific discoveries
Received 25 June 2020 and unmanned missions ranging from Under ocean research to Extra-terrestrial exploration. Prominent
Accepted 8 July 2020 missions that heavily used Robots include Exploration and Excavation, Unmanned Mining and Bomb
Available online xxxx
defusal, Complex Terrain mapping and Geo-Surveying, Remote Surveillance and other Military activities.
There are a various type of robots mainly classified on the basis of Control type – Autonomous, Semi-
Keywords: Autonomous and Manual; Locomotion Mechanism – Legged, Rover and Hybrid. While each type has
Robots
its own advantages and drawbacks, the suitability of each type depends on its target application. This
Hexapod
CAD
paper aims to design, develop and test an Autonomous robot, inspired from Biological evolution of
Biological evolution Insects, suitable for unmanned exploration missions.
Machine Learning Ó 2020 Elsevier Ltd. All rights reserved.
Selection and peer-review under responsibility of the scientific committee of the International Confer-
ence on Newer Trends and Innovation in Mechanical Engineering: Materials Science.

1. Introduction Section-2 discusses about Robot Geometry and Gait mecha-

nisms. Section-3 describes the Manufacturing methodologies fol-
A robot tasked with exploration missions has to be capable of lowed to develop the Prototype. Section-4 describes the Electrical
maneuvering through uneven terrains, move through cramped components used and their sizing. Section-5 describes the Machine
spaces and make informed decisions according to situations as Learning techniques used in the prototype.
they arise. Legged Robots are best suited for exploration missions
as require only limited footprint to operate, compared to wheeled
robots and rovers. The number of legs and orientation has to be 2. Robot geometry and Gait
decided based on the geometry of the robot and target application.
The biggest challenge in the design of legged robots is achieving Insects have evolved over the years to achieve body geometry
stability during the process of walking. Also, the gait has to be and gait patterns optimal for their survival by the process of Natu-
altered according to the robot’s surrounding. ral Selection. Parameters like Shape of the body, Number of legs,
Advances in Microprocessors have enabled us to deploy Orientation of legs and Weight distribution across the body deter-
increased processing power on-board the robot required for Path mine the stability and gait pattern. [1-3]. Literature review [4-9]
planning Algorithms. Power consumption and Weight distribution has shown that a 6-legged configuration to be most stable. Various
across the robot have to be considered for choice of Processor mod- types of legs configuration is shown in Fig. 1. The prototype adopts
ule. Also, the Imaging Sensor has to be compatible with the proces- a Hexagonal body structure to facilitate motion in any direction
sor for autonomous path planning algorithms. The use of Camera without turning about its axis. Also, the circular structure helps
mounted on the prototype has been preferred over imaging sen- to maintain Center of Gravity (CG) close to the geometric center,
sors because of reliable performance under fog, mist, smoke and enduring stability. The 6 legs of the robot are oriented uniformly
under water environments. at 60 degrees from each other in Arachnid arrangement with knees
facing outward. Arachnid arrangement helps the robot climb over
obstacles with ease.
⇑ Corresponding author. The effectiveness of this configuration can be proven by the fact
E-mail address: angalaeswari.s@vit.ac.in (S. Angalaeswari). that this configuration is being used by insects like ants.

https://doi.org/10.1016/j.matpr.2020.07.165
2214-7853/Ó 2020 Elsevier Ltd. All rights reserved.
Selection and peer-review under responsibility of the scientific committee of the International Conference on Newer Trends and Innovation in Mechanical Engineering:
Materials Science.

Please cite this article as: T. Deepa, S. Angalaeswari, D. Subbulekshmi et al., Design and implementation of bio inspired hexapod for exploration applica-
tions, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2020.07.165
2 T. Deepa et al. / Materials Today: Proceedings xxx (xxxx) xxx

The legs of the Hexapod have been designed to facilitate bend-

ing of one or more legs simultaneously to climb over obstacles or
reduce its height to maneuver through low-lying areas. The leg
has been designed to be a 3-part structure, going by the leg struc-
ture of most living organisms, including human beings. Such a 3-
part structure allows each leg to be actuated independently, help-
ing the Hexapod bend, stand and sit along different orientations as
shown in Fig. 2. This is particularly helpful to increase or decrease
the CG of the hexapod according to the terrain.
Following the analogy with biological organisms, the 3 parts of
the leg are called Coxa, Femur and Tibia as shown in Fig. 3. Coxa
links the leg with the body of the Hexapod, on the planar axis. Coxa
in analogous to Hipbone in the human body. Femur links the Coxa
and the Tibia and is the largest part of the leg. Femur is analogous Fig. 2. Orientation of legs.
to Thigh bone in the human body. Tibia acts as the end actuator
and is the second largest part of the leg. The Tibia is analogous to
Shin bone in the human body. The Coxa is responsible for forward
and backward motion of the leg along the X-Y axis. The Femur is
responsible for lifting the leg along the Z axis.
The Tibia is responsible for planting the leg firmly and extend-
ing the reach of the leg along the X-Z axis or Y-Z axis. The Coxa,
Femur and Tibia are each called Degrees of Freedom (DoF) as they
control the motion across certain axes. Each leg thus has 3 DoF and
the Hexapod has 6 such legs, giving the hexapod 18 DoF. Different
combinations of DoF being actuated gives different types of config-
uration. A single leg, with its three DoF can give 3! = 6 different
configurations. A total of 18! = 6.402373705728000 are possible,
while only a limited number of those configurations are practically
achievable [10].
Gait is defined as an organism’s manner of walking. The Hexa-
pod needs a certain Algorithm to analyze the surrounding environ-
ment and choose the optimal Gait. There are many types of Gaits
like Tripod Gait, Quadruped Gait, One-by-one Gait and Free Gait,
each being the best fit for geometric body configuration and
intended application. Free gait does not follow any algorithm while
Fig. 3. Parts of the leg.
One-by- one gait actuates each leg in a sequential fashion. Tripod
Gait and Quadruped gaits maintain contact with the surface using
3 alternate legs and 4 legs respectively. Tripod Gait has been cho-
sen for the prototype, given relatively better stability and walking
speed [11,12].
Taking the example of a simple body geometry in Fig. 4, tripod
gait would implement walking by lifting legs 1, 3 and 5 simultane-
ously to reach the new co-ordinates for these legs while legs 2, 4
and 6 remain grounded. Once this is completed, legs 2, 4 and 6
would be lifted and moved to new co-ordinates while legs 1, 3
and 5 remain grounded. This process is repeated to achieve straight
line walking. This sequence has been depicted in Fig. 5. The advan-
tage of this method is that 3 alternate legs remain footed on the
ground at any time, creating a triangle that is always closer to
the ground.

Fig. 4. Tripod gait.

This triangle always encloses the geometric Centre of Gravity of

the body, thus ensuring stability. As the legs adopt a circular orien-
tation, the hexapod gains orthogonal walking capabilities as well,
thus eliminating turning conditions. The significance of orthogo-
nality is that there is no front and rear for the hexapod’s body
[13]. Actuating sequences of different legs gives locomotion along
different directions. This helps improve the hexapod’s responsive-
Fig. 1. Types of legs. ness during walking [5-10].

Please cite this article as: T. Deepa, S. Angalaeswari, D. Subbulekshmi et al., Design and implementation of bio inspired hexapod for exploration applica-
tions, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2020.07.165
 T. Deepa et al. / Materials Today: Proceedings xxx (xxxx) xxx 3

Fig. 5. Tripod gait movement sequence.

3. Manufacturing Also, machining process can be done easily on Aluminium.

3 mm plates were used for the body parts and Aluminium channels
The key mechanical parameters that influence the manufactur- were used for the leg links. Mounts were custom manufactured for
ing process are: Total Weight, Weight-Position distribution and each motor as any unwanted relative motion between the motor
Thickness allowed. The allowed dimensions for the Hexapod have and mount could cause the hexapod to topple and fail. The alu-
been calculated based on the control components to be accommo- minium plates were cut to custom shapes by Waterjet and the
dated. A CAD model has been developed using SolidWorks and channels were welded after grinding. The mounts were screwed
weight analysis performed for different materials at different wherever possible and directly glued to the motor for certain cases.
dimensions and thicknesses. Results of these analyses have been The biggest challenge faced during the manufacturing process was
used to re-iterate the process and the material, dimensions and that each and every part had to be exactly similar to each other as
thickness of different parts optimized through a process of trial any minor difference could cause an uneven net force on the hex-
and error. Once the CAD model has been optimized, the model apod, causing it to topple ad fail.
was simulated in V-Rep Simulation software to validate Dynamic
stability and Gait efficiency as shown in Fig. 6. 4. Electrical considerations
The simulation has been re-iterated for various possible config-
urations of multiple parts and optimal configuration achieved. The point of contact between each leg and the ground needs to
Fresh CAD models have been prepared for the derived optimal con- be at Equilibrium at any given instant. As each leg is identical to
figuration as shown in Fig. 7 and simulated in both software tools each other within limits of manufacturing non-idealities and each
for final Weight Analysis, Gait Analysis and Stability analysis under leg is equidistant from the geometric CG, it can be assumed that
Static and Dynamic conditions. Semi-circular cuts have been made each leg bears one-sixth the net force involved. Taking into consid-
on the body between adjacent legs to improve the Range of Motion eration of all forces involved, the torque equation is written as, T =
(RoM) of each Coxa on either side. A standard Hexagon would have -W1*L1*cos(h1) - W2*(L1*cos(h1) + L2*cos(h2)) - W3*(L1*cos(h1) +
severely restricted the Coxa’s Range of Motion. Also, these cuts L2*cos(h2)) - W4*(L1*cos(h1) + L2*cos(h2)+L3) - 2*W3*(2L3+L1*cos
help in Weight Reduction by a significant margin. According to (h1) + L2*cos(h2)) - 2*W2*(L1*cos(h1) + 2*L2*cos(h2) + 2L3) -
the Weight Analysis, Aluminium 6061 has been chosen as the pre- 2*W1*(L1*cos(h1) + 2*L2*cos(h2)+2L3) + 2N2*(2L1*cos(h1) +
ferred material for manufacturing owing to its low cost, low weight 2L2*cos(h2)+2L3)
and high rigidity for the dimensions under consideration. Where
T- is the net Torque acting on the end of the leg
W1 – Weight of Coxa
W2 – Weight of Femur
W3 – Weight of Tibia
L1 – Length of Coxa
L2 – Length of Femur
L3 – Length of Tibia
h1 – Angle between Coxa and Femur
h2 – Angel between Femur and Tibia
N2 – Normal force acting on Femur
On solving, T = 2.43 kg-cm.
The motors have been chosen such that rated Torque is greater
than 4 times the required torque to account for conditions when
the Hexapod encounters a fall. Servo motors have been preferred
over DC geared motors because of the precise angular control.
Among Servo motors that match requirements, models MG995
and MG996 have been used in the prototype because of cost and
Fig. 6. V-Rep simulation. availability as the specifications mentioned in Table 1.

Please cite this article as: T. Deepa, S. Angalaeswari, D. Subbulekshmi et al., Design and implementation of bio inspired hexapod for exploration applica-
tions, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2020.07.165
4 T. Deepa et al. / Materials Today: Proceedings xxx (xxxx) xxx

Fig. 7. CAD model.

Table 1 Arduino microcontroller because of its ease of use. The servo motor
Specification of motors. has to be controlled using PWM signals from the Arudino GPIO
Motor Stall torque Operating Current
pins. 1 board each of Arduino Uno and Arduino Mega have been
voltage consumption used in the prototype. GPIO pins2, 3, 5, 6, 7, 9 are PWM capable
MG995 10 kg-cm@4.8 V12 kg-cm@6V 4.8 V-6 V Upto 1A
in Arduino Uno whereas digital pins 2–14 are PWM capable in
MG996 9.4 kg-cm@4.8 V11 kg-cm@6V 4.8 V-6 V Upto 1A Arduino Mega.

5. Machine Learning
Both motors are Metal gear type motor, thus reducing chances
of mechanical damage during fall conditions. These motors are The task proposed by this paper is achieved by implementing a
both 3.3 V and 5 V logic compliant, meaning they can be directly highly flexible and efficient algorithm called LSD-SLAM. The algo-
controlled using Arduino or Raspberry Pi. The prototype uses rithm performs a simultaneous localization of the environment

Fig. 8. Path planning and mapping using Quadtree algorithm.

Please cite this article as: T. Deepa, S. Angalaeswari, D. Subbulekshmi et al., Design and implementation of bio inspired hexapod for exploration applica-
tions, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2020.07.165
 T. Deepa et al. / Materials Today: Proceedings xxx (xxxx) xxx 5

Fig 9. 3D Map of the simulated environment.

and maps it in 3D [14]. This enables the user to easily identify hot- References
spots and traverse the map to find possible threats and safe zones
before human activity. Initial task was to make the bot identify the [1] D.D. Esteban, M. Luneckas, T. Luneckas, J. Kriaučiu  nas, D. Udris, ‘‘Statically
stable hexapod robot body construction,” IEEE 4th Workshop on Advances in
environmental parameters and to plan safe paths for navigation. Information, Electronic and Electrical Engineering (AIEEE), Vilnius (2016) 1–4.
Simulations were done using Quadtree algorithm and then later [2] J. Bo, C. Cheng, L. Wei and L. Xiangyun , ‘‘Design and Configuration of a
SLAM was preferred as a better alternative due to the added benefit Hexapod Walking Robot,” Third International Conference on Measuring
Technology and Mechatronics Automation, Shangshai,(2011), pp. 863-866.
of having 3D user controllable point cloud map. LSD SLAM was [3] T. JyhHwa, C.Y. Liang, The development of the hexapod bio-robot system
chosen out of the available methods to map and path plan as this (2008), 27th Chinese Control Conference, Kunming (2008) 642–646.
algorithm uses a camera rather than range finders. The algorithm [4] Z. Lei, D. Xin and J. Hua), ‘‘Motion Simulation of Bionic Hexapod Robot Based on
Virtual Prototyping Technology,” International Conference on Computer
directly works on image intensities to track and map. Geometry
Network, Electronic and Automation (ICCNEA), Xi’an, (2017),pp. 426-430.
is estimated in the form of semi-dense depth maps, obtained by fil- [5] Y. Isvara, S. Rachmatullah, K. Mutijarsa, D. E. Prabakti and W. Pragitatama,
tering over many pixel wise stereo comparisons. ‘‘Terrain adaptation gait algorithm in a hexapod walking robot,” 13th
Further simulations were performed to analyze and calculate International Conference on Control Automation Robotics & Vision (ICARCV),
Singapore, (2014), pp. 1735-1739.
the performance of path planning algorithms and to plot a map. [6] E. H. Hasnaa and B. Mohammed, ‘‘Planning tripod gait of an hexapod robot ,
Quadtree algorithm was used to simulate the mapping and path ”14th International Multi-Conference on Systems, Signals & Devices (SSD),
planning as shown in Fig. 8. The robot in its testing phase was used Marrakech, (2017),pp. 163-168.
[7] C. Zhang, X. Jiang, M. Teng, J. Teng, ‘‘Research on Gait Planning and Static
to capture the video of the environment using a camera mounted Stability of Hexapod Walking Robot,” 8th International Symposium on
on the bot. LSD-SLAM algorithm was used to create a map of the Computational Intelligence and Design (ISCID), Hangzhou (2015) 176–179.
environment. The result was a 3D map of the environment shown [8] Baoling Han, Lin Huang, Qingsheng Luo, Xu. Jia, ‘‘The biomimetic research on
new-style hexapod robot’s locomotion planning,” 7th World Congress on
below in Fig. 9 and the algorithm also calculated possible path Intelligent Control and Automation, Chongqing (2008) 3264–3268.
between two points in the environment and can move to any point [9] A. Konthong, A. Monprajuck, R. rattanavorragant and Y. Jewajinda, ‘‘An
on the map without GPS. Intelligent Locomotion Control Architecture for Hexapod Robot ,” 15th
International Joint Conference on Computer Science and Software
Engineering (JCSSE), Nakhonpathom, (2018), pp. 1-6.
6. Conclusion [10] G. Jianhua, ‘‘Design and Kinematic Simulation for Six-DOF Leg Mechanism of
Hexapod Robot ,” IEEE International Conference on Robotics and Biomimetics,
Kunming, (2006),pp. 625-629.
A Hexapod robot has been designed and manufactured using [11] G. Zhang, G. Yu and You Bo, ‘‘Hexagonal hexapod robot foot-modeled
Aluminium 6061 as the base material. The design involves exten- dynamics simulation analysis,” Proceedings of 2013 2nd International
sive CAD modelling and simulation. Motor and Battery calculations Conference on Measurement, Information and Control, Harbin, (2013), pp.
1107-1110.
have been performed to employ proper Electrical components. The [12] L. Chen, G. Zhong, Z. Liu and H. Deng, ‘‘A novel hexapod robot: Design and
Hexapod employs advanced Mapping algorithms to map its envi- mobility analysis,” IEEE International Conference on Mechatronics and
ronment and simultaneously perform Path-planning to eliminate Automation, Harbin,(2016), pp. 2094-2098.
[13] H. Gao, M. Jin, Y. Liu, L. Ding, H. Yu and Z. Deng, ‘‘Turning gait planning and
the need for manual intervention. This autonomous decision mak- simulation validation of a hydraulic hexapod robot ,” International Conference
ing allows the robot to be used in challenging, inaccessible terrains on Fluid Power and Mechatronics (FPM), Harbin, (2015),pp. 842-847.
without the need for human presence. [14] Y. Endo, K. Sato, A. Yamashita and K. Matsubayashi, ‘‘Indoor positioning and
obstacle detection for visually impaired navigation system based on LSD-
SLAM,” International Conference on Biometrics and Kansei Engineering
Declaration of Competing Interest (ICBAKE), Kyoto, (2017), pp. 158-162.doi: 10.1109/ICBAKE.2017.8090635

The authors declare that they have no known competing finan-

cial interests or personal relationships that could have appeared
to influence the work reported in this paper.

Please cite this article as: T. Deepa, S. Angalaeswari, D. Subbulekshmi et al., Design and implementation of bio inspired hexapod for exploration applica-
tions, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2020.07.165