Literature Review: Addressing Research Gaps in Humanoid Robotics

This report presents a literature review on the field of robotics, specifically focusing on the research gaps that persist in the development of humanoid robots. The review examines the evolution of robotics, highlighting the advancements in robot design and functionality. It identifies key areas of concern, including the discrepancies between simulated robot models and real-world performance, the challenges associated with integrating artificial intelligence, and the potential societal impacts of widespread robot adoption, such as job displacement. The analysis draws on various research papers and studies to pinpoint these gaps and evaluate the potential consequences of these issues. The report emphasizes the need for further research to address the limitations of current simulation models, the ethical considerations surrounding the development of AI, and the broader implications for human employment and societal structures. It concludes by underscoring the importance of a balanced approach to robotics development, one that prioritizes both technological advancement and responsible implementation.

STUDENT NAME - 42908 - Assignment 2
07 September 2016 1

STUDENT NAME - 42908 - Assignment 2
Literature Review Template
By
Abstract
Robotics is the field of study that deals with the research, design and development of robots –
automated machines that can perform duties as instructed through programming. Throughout the years,
researchers have developed more and more advanced designs of robots with latest technology.
However, there are several gaps that are persisting with the existing research results on the humanoid
robots. In this report, a literature review has been conducted to find the gaps that are still remaining in
the research of robots and robotics and these gaps have been addressed.
1. Introduction
A robot is referred to as a programmable machine that can perform different types of activities based
on the command sent to it through wireless media. These robots are programmed to do many
mechanical activities using the movement of the robotic arms that are analogous to the movement of
human arms. Humanoid robots are the ones whose appearance is similar to a human being. Humanoid
robots have been designed in order to reduce the mechanical work load of the human beings. The types
of work that a robot can do include moving an object from one place to another, sweep the floor using a
brush or cleaner, interaction with mechanical tools or even drive a car (Madhavan, Prestes & Marques
2015). The robot is made to perform these activities with the help of electronic commands sent from an
electronic controller. Since its first inception in the late 19th century, robotics has come a long way and
now, a robot can perform a wide range of activities. A robot consists of a human shaped metallic body
with arms and a head, which can rotate in any direction and its legs are fitted with wheels that help it to
move from one location to another (Aloimonos 2013). Robots are considered to be the future of
humanity as it is expected that robots will be able to perform about 70% work done by humans
currently within the next 100 years. However, there are several gaps in the research activities that have
been performed till now on the robots (Anderson 2016). Most of the robotics scientists have proposed a
number of work activities that can be performed using robots in the future and how robots will take
over the world dominated by human in the next century but have not sufficiently emphasized on other
07 September 2016 2

STUDENT NAME - 42908 - Assignment 2
important factors like the costs of making such advanced robot, insertion of artificial intelligence in
robot and others.
In this report, a literature review has been conducted to find the gaps that are still remaining in the
research of robots and robotics.
2. Literature Review
Different researchers have worked on robotics and proposed many theories and hypothesis. Most of
their works have considerably promoted the field of robotics and over the years, many new and more
advanced robots have been developed. Currently, robots are even made to football and there are
tournaments existing for robotic football matches (Bruce et al. 2017). All these on the bright side of the
field of robotics and people can expect even more advanced and functioning robots in the near future.
However, some gaps have still remained in the research on robotics that has not yet been addressed by
the researchers. Some researchers have analyzed some of the gaps in recent times and proposed some
feasible solutions to address the solutions.
According to Ervin et al. (2015), one main research gap the humanoid robotics is that the researchers
are developing robot models based on the simulation results that are being done using some softwares.
Although simulation results can provide a proper standard of the actual design, there is significant gap
between the simulation and the actual robot design. Using simulation, the software created robot model
can literally do anything – from playing soccer to performing complex mechanical activity. Actual
humanoid robots on the other hand have significant limitations as of the currently available designs.
Working of the current designs of the humanoid robots depend solely on the placement of round pivots
at the arm joints and the wheels fixed on the bottom of the robot legs and the functions are limited to
light mechanical activities. Moreover, simulated robots can be made do anything using simple
programming. On the other hand, in reality, the commands are limited as the coding of these
commands is complex and it is hard to make the robot understand the actual meaning of the commands.
Xu, Mellmann and Burkhard (2010) have mainly emphasized their work on one particular research gap
in robotics and that is the gap between the simulation and reality. They have also used a particular case
study to identify the gap. According to them, SimSpark is a simulation software that provides a good
platform for simulating software created robot models. They also researched on the Robot World Cup
(RoboCup) that had been created earlier for making the robots play soccer with the main targets of
fostering the research on robotics as well as develop a full robot team that will be able to defeat a full
human team in football by the year 2050 (Colorado et al. 2015). In addition to the research, the
07 September 2016 3

STUDENT NAME - 42908 - Assignment 2
researchers have also built a unified robot team that will be able to play both Simulation League and
Standard Platform League in RoboCup. Based on the researchers’ findings, the following are the details
of the two different leagues:
3D Simulation League – Initially, the simulation league was conducted in 2D simulation environment
as the 3D technology was not yet developed. In this 2D simulation, sphere shaped agents ran across the
field based on the provided commands. Later this simulation technique was utilized in the development
of sports games like football, NBA and others that could be played in the computer. After 2007,
humanoid robots were introduced for playing in the 3D simulation league. The researchers based their
research on these humanoid robots in the 3D simulation and found that as a consequence of the changes
to the 3D simulator, a (temporary) shift of problem solving to the more basic problems of body control
could be observed (Ienca et al. 2016). The researchers observed restrictions in this particular
environment regarding the application of tactics among the robots for decision making (e.g. making a
good pass, shooting from long range, etc.). They also found that simulations allow more robot players
than it is possible in real scenario. They found that the simulation can easily play 6 a side robots
efficiently or sometimes even 11 v 11 if the system is of very high configuration. On the other hand, in
reality, only 5 v 5 matches could be held at a time.
Standard Platform League – Standard Platform League is where real humanoid robots are played in
soccer matches. Due to the limitation of space and size of the ground, generally 3 v 3 games are played.
These robots are not controlled by any external remote controls as they play on their own based on the
programming that has been preset on their operating systems. Since these robots are biped, the motion
is much less flexible and balanced that the quadruped robots.
From the study of these two, the researchers came to the conclusion that the simulation results are
much more efficient that the actual models mainly due to the geometry of the actual robot models
(Imaida & Senda 2015). Hence, this gap can be filled by further developing the robot models by adding
from flexibility to the entire structure and development of more complex codes to program more
complex movements.
Rösch et al. (2015) talked about another research gap in the field of robotics and development of
humanoid robots. They stated that current robot developers are trying to introduce more and more
artificial intelligence into the robots so that they are more efficient and will not require any external
programming or coding to make them perform any activity. Few years ago, this was merely a
hypothesis. However, this has become a reality and scientists have successfully able to implement
07 September 2016 4

STUDENT NAME - 42908 - Assignment 2
artificial intelligence in the robots to a certain extent. According to an estimation by the researchers,
complete implementation of artificial intelligence in the robots will be possible within the next 50-70
years (McDowell & Gunkel 2016). However, in this matter, there is a significant gap or void – nobody
can answer to the question that what will happen if the robots are completely intelligent artificially one
day. Unless this question is properly addressed, it should not be a very good idea to implement too
much artificial intelligence within a humanoid robot.
3. Evaluation
Based on the analysis of the literature review on the research gaps in the development of humanoid
robots, several points have been noted. The first gap refers to the huge gap between the simulation and
reality. As most scientists prefer to use simulated models for preparing the design of the real robots,
there is a significant gap as the simulation models and real models are entirely different. Simulation
models are only visually recreated and these models can do whatever instruction is given to them. On
the other hand, the actual model cannot just “do” anything (Murata et al. 2013). This is due to the fact
that the movement of the robot is limited by the flexibility of its metallic body as well as the mode of
the robotic movement. In spite of this particular gap, scientists are still using simulation of robot
models to develop even better models. Another main gap discussed is the after-effects of artificial
intelligence inside the robots (Saicharan, Tiwari & Roberts 2016). Scientists are developing artificially
intelligent robots so that that can some specific tasks like driving a car, run software programs or even
defense (military). However, the following questions arise regarding this particular gap:
How much intelligent can a robot become using artificial intelligence?
What if a robot becomes so intelligent that is able to replicate itself?
What if robots take over the world controlled by humanity?
What if some powerful countries use intelligent robots as military?
Most of the researchers of robotics seem to ignore all of these issues while developing more and more
intelligent models of robots (Srinivasan 2015). While all of the above questions seem to be fictional or
far-fetched, researchers have no answers to them if they actually happen in the next 50-75 years.
However, for the time being, artificial intelligence should only be implemented in a robot in a
controlled environment and should be limited to only some particular functions.
Another issue that has been found from the analysis is that due to the rapid advancement of robots,
increasing numbers of people are losing their jobs around the world (Venturini & Verbano 2014).
Robots were built to aid human activities and reduce the workload from an individual whereas the
07 September 2016 5

STUDENT NAME - 42908 - Assignment 2
robots are actually taking over the jobs of many people as they are much efficient and one robot can
actually perform the duties of 20-30 people at a time. According to an estimate, in the next 50 years,
robots will take over almost 50% of the jobs that are currently available for human (Yang 2016). As
mentioned by some researchers, the world does not need robots to replace humans; rather it needs
robots that will be able to enhance the performance an individual.
4. Conclusions
In this report, a literature review has been conducted to find the gaps that are still remaining in the
research of robots and robotics. Based on the research on the topic, it has been found that different
researchers have worked on robotics and proposed many theories and hypothesis. Some researchers
have analyzed some of the gaps in recent times and proposed some feasible solutions to address the
solutions. The first research gap that has been identified is that the researchers are developing robot
models based on the simulation results that are being done using some softwares. Although simulation
results can provide a proper standard of the actual design, there is significant gap between the
simulation and the actual robot design. Using simulation, the software created robot model can literally
do anything – from playing soccer to performing complex mechanical activity. Another research gap
that has been identified from the study is that current robot developers are trying to introduce more and
more artificial intelligence into the robots so that they are more efficient and will not require any
external programming or coding to make them perform any activity. Few years ago, this was merely a
hypothesis. However, this has become a reality and scientists have successfully able to implement
artificial intelligence in the robots to a certain extent. According to an estimation by the researchers,
complete implementation of artificial intelligence in the robots will be possible within the next few
decades. Hence, before further development of the humanoid robots, these gaps must be covered and
addressed by the researchers so that all the issues are solved and the robots are developed according to
the needs and not on the wants.
07 September 2016 6

STUDENT NAME - 42908 - Assignment 2
5. References
Aloimonos, Y. ed., 2013. Active perception. Psychology Press.
Anderson, G., 2016. The Economic Impact of Technology Infrastructure for Advanced Robotics. NIST,
Gaithersburg, MD) NIST Economic Analysis Briefs, 2.
Bruce, C.D., Davis, B., Sinclair, N., McGarvey, L., Hallowell, D., Drefs, M., Francis, K., Hawes, Z.,
Moss, J., Mulligan, J. & Okamoto, Y., 2017. Understanding gaps in research networks: using “spatial
reasoning” as a window into the importance of networked educational research. Educational Studies in
Mathematics, 95(2), pp.143-161.
Colorado, J., Rossi, C., Zhang, C. & Barrientos, A., 2015. Towards efficient flight: insights on proper
morphing-wing modulation in a bat-like robot. Advanced Robotics, 29(24), pp.1599-1610.
Ervin, A.M., Taylor, H.A., Meinert, C.L. & Ehrhardt, S., 2015. Evidence gaps and ethical review of
multicenter studies. Science, 350(6261), pp.632-633.
Ienca, M., Jotterand, F., Vică, C. & Elger, B., 2016. Social and assistive robotics in dementia care:
Ethical recommendations for research and practice. International Journal of Social Robotics, 8(4),
pp.565-573.
Imaida, T. & Senda, K., 2015. Performance improvement of the PD-based bilateral teleoperators with
time delay by introducing relative D-control. Advanced Robotics, 29(6), pp.385-400.
Madhavan, R., Prestes, E. & Marques, L., 2015, May. Robotics and Automation Technologies for
Humanitarian Applications: Where we are and where we can be. In Full-day Workshop, IEEE
International Conference on Robotics and Automation (ICRA).
McDowell, Z.J. & Gunkel, D.J., 2016. Introduction to" Machine Communication". communication+
1, 5(1), pp.1-5.
Murata, S., Konagaya, A., Kobayashi, S., Saito, H. & Hagiya, M., 2013. Molecular robotics: A new
paradigm for artifacts. New Generation Computing, 31(1), pp.27-45.
Rösch, S., Ulewicz, S., Provost, J. & Vogel-Heuser, B., 2015. Review of Model-Based Testing
Approaches in Production Automation and Adjacent Domains—Current Challenges and Research
Gaps. Journal of Software Engineering and Applications, 8(09), p.499.
Saicharan, B., Tiwari, R. & Roberts, N., 2016, July. Multi Objective optimization based Path Planning
in robotics using nature inspired algorithms: A survey. In Power Electronics, Intelligent Control and
Energy Systems (ICPEICES), IEEE International Conference on (pp. 1-6). IEEE.
Srinivasan, A., 2015. Stop the Robot Apocalypse: the New Utilitarians. London Review of
Books, 37(18), pp.3-6.
Venturini, K. & Verbano, C., 2014. A systematic review of the Space technology transfer literature:
Research synthesis and emerging gaps. Space Policy, 30(2), pp.98-114.
Xu, Y., Mellmann, H. & Burkhard, H.D., 2010, November. An approach to close the gap between
simulation and real robots. In International Conference on Simulation, Modeling, and Programming
for Autonomous Robots (pp. 533-544). Springer, Berlin, Heidelberg.
Yang, J., 2016. The robot apocalypse is here: and it's not what we expected. ACM SIGCAS Computers
and Society, 46(1), pp.33-35.
07 September 2016 7