Cyborg Intelligence Recent Progress and

CYBORG INTELLIGENCE:RECENT
PROGRESS AND FUTURE DIRECTIONS
SEMINAR REPORT
submitted by
AJAYGOSH P.S.
Reg. No. REANEEC007
Eighth Semester
BACHELOR OF TECHNOLOGY
in
Electronics and Communication Engineering
UNIVERSITY OF CALICUT
Department of Electronics and Communication Engineering

Royal College of Engineering and Technology
Akkikavvu , Thrissur -680 604
( www.royalcet.ac.in)
March 2016
( www.royalcet.ac.in)
Bonafide Certificate
This is to certify that the seminar report entitled “CYBORG INTELLIGENCE:RECENT
PROGRESS AND FUTURE DIRECTIONS”, is a bonafide record of work carried
out by AJAYGOSH P.S. (Reg.No.REANEEC007) of Eighth Semester Electronics and
Communication Engineering (2013 admissions) of Royal College of Engineering and
Technology, as a requirement for the award of Degree of Bachelor of Technology in
Electronics and Communication Engineering of UNIVERSITY OF CALICUT dur-
ing the academic year 2016-2017.
SUPERVISOR Head of Department
Mr.GNANAJEBADAS Dr.V.Vijayarangan
Asst. Prof., Dept. of ECE Dept. of ECE
This report was evaluated on ..../..../........

(Department Seal)
CYBORG INTELLIGENCE:RECENT PROGRESS AND FUTURE DIRECTIONS
Acknowledgement
I wish to record my indebtedness and thankfulness to all those who helped me to

prepare this report titled “ CYBORG INTELLIGENCE:RECENT PROGRESS AND
FUTURE DIRECTIONS ” and present it in a satisfactory way.
First and foremost I thank God Almighty for His providence and for being the
guiding light throughout the seminar.
I want to start expressing my thanks to my seminar guide Mr.GNANAJEBADAS
, Asst. Prof., Dept. of Electronics and Communication Engineering, because of her
valuable advice and guidance towards this work. I received motivation, encouragement
and hold up from her during the course of work.
I am thankful to Dr.V.Vijayarangan, Head of Electronics and Communication

Engineering Department, and our Principal Dr. S. P. Subramanian, for their sole co-
operation.
Finally, I would like to extend my sincere gratitude to friends who have always
been helpful, in preparing and presenting the report and in the discussion following the
presentation.
AJAYGOSH P.S.
Reg. No. REANEEC007
Eighth Semester
Electronics and Communication Engineering (2013 Admissions)
March 2016
Department of ECE i RCET

ABSTRACT
The interconnection and tight integration of biological information processing

units and man-made computing components provides extensive information exchange
between biological organisms and computing devices, which spawned cyborg intelli-
gence. Artificial and biological intelligence begin to share common territory in provid-
ing sensation, perception, cognition, and locomotion. Many amazing results have been
achieved in various areas, including animals as sensors and actuators, mind-controlled
machines, neurochips, intelligent prosthesis, and neural rehabilitation. Recently, the
open Cyborg Olympics(CO) were held to drive the realistic combination of biological
and artificial intelligence and help disabled people better acclimate to daily life.
Department of ECE ii RCET

Contents
ACKNOWLEDGEMENT i
ABSTRACT ii
LIST OF FIGURES iv
LIST OF SYMBOLS AND ABBREVIATIONS v
1 INTRODUCTION 1
1.1 Report Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2 LITERATURE SURVEY 3
3 Cyborg Intelligence: Recent Progress and Future Directions 14

3.1 Conceptual Framework of Cyborg Intelligence: A Revisit . . . . . . . . 16
3.2 Cyborg Intelligence: Research Progress . . . . . . . . . . . . . . . . . . 17
3.2.1 Sensation-Augmented Rat Cyborgs . . . . . . . . . . . . . . . . 18
3.2.2 Monkey Hand-Gesture Decoding . . . . . . . . . . . . . . . . . 19
3.2.3 Mind-Controlled Rock-PaperScissors Game . . . . . . . . . . . . 20
4 RESULT AND DISCUSSION 22
5 CONCLUSION 24
5.1 Advantages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.2 Disadvantages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.3 Future scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
REFERENCES 26
Department of ECE iii RCET

List of Figures
2.1 Three main components of our rat cyborg system. The electrode picture
is taken under a microscope. The rat-mounted pack includes a miniature
camera, a wireless module, and a stimulator. . . . . . . . . . . . . . . . . 7
3.1 Hierarchical conceptual framework for cyborg intelligence.1 Three lay-

ers are abstracted to describe the interconnection between the biological
organisms and computing machines. . . . . . . . . . . . . . . . . . . . . 17
3.2 Illustration of research progresses in the framework of cyborg intelligence. 18
3.3 Vision-augmented rat cyborg. The ratbot can recognize image indicators
and navigate complicated environments guided by commands from its
computing components. . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.4 Robotic hand control with neural decoding. We used motor control com-
mands decoded directly from neural spikes of the primary motor cortex
to manipulate the robotic hand to demonstrate four gestures. . . . . . . . 20
3.5 Rock-paper-scissors game controlled by human electrocorticogram (ECoG)
signals. Hand gestures and their corresponding ECoG signals. And live
demonstration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Department of ECE iv RCET

List of Abbreviations
RCET Royal College of Engineering and Technology
AI Artificial Intelligence
FMRI Functional magnetic resonance imaging
ECoG Electrocorticogram
CO Cyborg Olympics
Department of ECE v RCET

Chapter 1
INTRODUCTION
Typical and effective approaches to implementing engineering systems and ex-

ploring research problems in cyborg intelligence are based on brain-computer (or neural-
computer) integration methods. Using these methods, computers can record neural activ-
ity at multiple levels or scales, and thus decode brain representation of various function-
alities and precisely control artificial or biological actuators.
In recent decades, there have been continuous scientific breakthroughs re-

garding the directed information pathway from the brain to computers. Meanwhile, be-
sides ordinary sensory feedback such as visual, auditory, tactile, and olfactory input,
computers can now encode neural feedback as optical or electrical stimuli to modulate
neural circuits directly. This forms the directed information pathway from the computer
to the brain. These bidirectional information pathways make it possible to investigate
the key problems in cyborg intelligence. Here focused on the computational architec-
ture of cyborg intelligence, especially for sensorimotor processes in the brain-computer
integration system. First proposed a conceptual illustration and constructed a cognitive
computational model for brain-computer integration systems.
To understand neural representation in the brain, we also explored encod-

ing and decoding principles underlying the sensorimotor loop, and then the computers
implemented novel AI algorithms to enhance sensation and motor control functions of
the overall brain-computer integration system. From our theoretical and technological
results, we implemented cyborg rats, monkey mind control, and a rehabilitation demon-
stration as evaluating systems. Extensive experiments show that the philosophy concept
Department of ECE 1 RCET

and computation architecture of cyborg intelligence is promising for enhancing, repair-

ing, or extending the intelligent capacity of both biological and computing units.
1.1 Report Organization

The seminar report is organized in such a way that, chapter 1 is the introduc-
tion, chapter 2 is the literature survey. chapter 3 deals with Cyborg Intelligence: Recent
Progress and Future Directions. The fourth chapter deals with the experimental results.
The last chapter includes the conclusions reached from this study. The last section is the
references.

Chapter 2
LITERATURE SURVEY
1. Author: Yueming Wang, Minlong Lu, Zhaohui Wu, Liwen Tian,Kedi Xu and Xi-
aoxiang Zheng,Gang Pan
Title: Visual Cue-Guided Rat Cyborg for Automatic Navigation
Published in: , IEEE Computational Intelligence, vol. 10, no. 2, 2015, pp. 4252.
A rat robot is a type of animal robots, where an animal is connected to a

machine system via a brain-computer interface. Electrical stimuli can be generated
by the machine system and delivered to the animal’s brain to control its behaviour.
The sensory capacity and flexible motion ability of rat robots highlight their poten-
tial advantages over mechanical robots. However, most existing rat robots require
that a human observes the environmental layout to guide navigation, which limits
the applications of rat robots. This work incorporates object detection algorithms
to a rat robot system to enable it to find humaninteresting objects, and then use
these cues to guide its behaviours to perform automatic navigation.
A miniature camera is mounted on the rat’s back to capture the scene in

front of the rat. The video is transferred via a wireless module to a computer and we
develop some object detection/identification algorithms to allow objects of interest
to be found. Next, we make the rat robot perform a specific motion automatically
in response to a detected object, such as turning left. A single stimulus does not
allow the rat to perform a motion successfully. Inspired by the fact that humans
usually give a series of stimuli to a rat robot, we develop a closed-loop model that
issues a stimulus sequence automatically according to the state of the rat and the

objects in front of it until the rat completes the motion successfully. Thus, the rat
robot, which we refer to as a rat cyborg, is able to move according to the detected
objects without the need for manual operations. The object detection methods and
the closed loop stimulation model are evaluated in experiments, which demonstrate
that our rat cyborg can accomplish human-specified navigation automatically.
An animal robot is an animal that is connected to a machine system, usu-

ally via a brain-computer interface, which allows electrical stimuli to be delivered
to specific brain areas, thereby driving the animal to take actions that are specified
by humans. The stimuli are delivered to specific brain areas via implanted elec-
trodes. In particular, animal robots can be controlled by humans to navigate along
a specified path. Because of the specific motion and perceptual abilities of animals,
animal robots have great potential for use in rescue and search applications.
A rat robot is a typical animal robot, which can navigate along a hu-
manspecified route. A major disadvantage is that humans need to identify the ar-
rangements of objects in the environment before giving appropriate instructions to
facilitate navigation in the environment. This limits the possible applications of rat
robots in environments that humans cannot observe. In some applications, only a
few objects are of interest to a rat robot, such as human faces or indication signs.
If the rat robot system can find these objects and a motion action is specified for
each object, this would allow the rat robot to perform human-specified navigation
automatically. In this preliminary study, we attempt to address this problem. We
construct a rat robot where the rat can accept stimuli and perform a few basic ac-
tions, such as turning left, turning right, and walking forward. Our novel system
makes two major contributions, as follows.
1) To allow the rat robot to find human-interesting objects, i.e., the objects that eas-
ily attract humans attentions such as human faces and indication signs, a miniature
camera is mounted on the back of the rat robot and the video captured by the cam-
era is transferred to a computer. Interesting objects in the video, such as human
faces and arrow signs, are then identified by object detection Visual algorithms and

the detection results are used to control the rat robot.
2) To allow the rat robot to navigate automatically while being guided by the iden-
tified objects/cues, we develop a stimulation model that drives the rat robot to per-
form a unique motion action in response to the detection of an object. A problem
with automatic control is that a single stimulus, e.g., a stimulus for turning left,
does not allow the rat to perform a successful turning left motion. Humans usually
give a series of stimuli to the rat for this purpose, according to the state of the rat
and the objects in front of it. Inspired by this manual control process, we develop
a closed-loop stimulation model that mimics the human control procedure, which
issues a stimulus sequence automatically according to the state of the rat and the
objects detected until the rat completes the motion successfully.
Existing animal robots can be categorized according to three classes:

aerial, aquatic, and terrestrial animal robots.
The first aerial animal robot was described in 1997, where the locomo-
tory reaction of a Periplaneta americana in response to various electrical stimuli
was analyzed. Using two photosensors as inputs, an electronic backpack was used
to drive the insect to walk along a black line. The success of the Periplaneta amer-
icana robot greatly inspired research into animal robots based on insects. Micro-
probes were implanted in Manduca sexta during its early metamorphosis to allow
direct control of its wing motions, thereby controlling the flight direction in Man-
duca sexta.
Stimuli were delivered to the brains of beetles to elicit, suppress, or modu-

late wing oscillations, thereby controlling flight initiation, cessation, and elevation.
Turns were triggered by direct muscular stimuli. In the flight initiation and ces-
sation behaviors of honeybees were generated in a reproducible manner by using
electrical pulses between two wire electrodes implanted in the honeybees brain.
The effects of different stimulus patterns on the honeybees behavior were com-

pared.
The second category of animal robots is based on aquatic animals. In

locomotion control in the horizontal plane was accomplished in goldfish by stim-
ulating the Nflm region in the midbrain. When the stimulation site was closer to
the Nflm region, a lower stimulus intensity was required to evoke movements in
goldfish. It was shown that turning left, turning right, moving forward, and moving
backward behaviors could be induced in adult carp via electric stimulation of the
cerebellum.
Terrestrial animal robots such as Gekko gecko and rats, have also been
investigated. A rat robot was first developed in, where applications of electrical
stimuli to the somatosensory cortices (SI) and medial forebrain bundle (MFB) were
used as cues and rewards, respectively. The rat could easily be guided through pipes
and across elevated runways, and it could even be instructed to climb or jump from
trees. A new method for rat robot navigation was proposed in, which was based on
virtual punishment.
Electrical stimuli were delivered to the thalamic ventral posterolateral nu-

cleus and amygdala nucleus of the rat brain. In response to these virtual sensations,
rats would change directions and escape in an active manner. In, the immobility
behavior triggered by dorsolateral periaqueductal gray (dlPAG) stimulation was in-
vestigated. After stimulating the MFB or dlPAG during navigation, a rat could be
switched between active and motionless states, where the motionless period was
controlled to a certain extent.
A new locomotion control scheme was developed for rat navigation. Al-
though animal robots have been studied extensively, few efforts have been made to
develop an animal robot that is capable of performing automatic navigation. In this
study, we incorporate object detection algorithms in a rat robot and the objects iden-
tified are used to guide automatic navigation via a closed-loop stimulation model.

Thus, the present study comprises neuroengineering and a preliminary study in

the new field of cyborg intelligence, i.e., incorporating biological intelligence and
machine intelligence to obtain more powerful capacities in a system.
Figure 2.1: Three main components of our rat cyborg system. The electrode picture is
taken under a microscope. The rat-mounted pack includes a miniature camera, a wireless
module, and a stimulator.

2. Author: Zhaohui Wu, Gang Pan,Jose C. Principe,Andrzej Cichocki

Title: Cyborg Intelligence: Towards Bio-Machine Intelligent Systems
Published in: IEEE Intelligent Systems,vol. 29, no. 6, 2014.
Artificial intelligences goal is to design computers that process informa-

tion and act like human brains. Several milestones in specific applications have
been achieved in the last two decades. For example, the chess-playing computer,
Deep Blue, defeated the world chess champion Garry Kasparov in 1997.
The unpiloted vehicles by the Stanford Racing Team successfully com-

pleted a 212-km off-road course with more than 100 sharp left and right turns and
won the 2005 DARPA Grand Challenge. The computer system of answering ques-
tions in natural language by IBM competed on the quiz show Jeopardy! against two
former winners and won the prize in 2011. These advances highlight the growing
importance of Al.
However, human-level Al remains impossible to achieve. All b iolog-

ical organisms display specifi c skills showcasing biological intelligence that is
difficult to imitate. They possess special sensory capabilities (such as dogs, with
their heightened olfactory senses), adaptability to living environments, and special
cognitive capabilities (such as reasoning and planning). Naturally, biological and
machine intelligence have their own intrinsic merits, and integration of both kinds
of intelligence is emerging. From the biological side, machines can assist and aug-
ment biological beings, while from the machine side, biological beings may help
solve machine tasks.
As research continues, the line between humans and machines begins to

blur. In fact, Al is extending to include biological intelligence. For instance, in
collective intelligence or crowdsourcing, the machine asks humans to solve a prob-
lem, then collects and integrates their solutions. However, in these approaches, the
connection between humans and machines is loosely coupled in the traditional hu-

man computer interaction manner.
Cyborg intelligence is dedicated to integrating AI with biological intel-

ligence by tightly connecting machines and biological beings, for example, via
brain-machine interfaces (BMIs). BMIs bring the promise to enhance strengths
and compensate for weaknesses by combining the biological perception/cognition
ability with the machines computational ability. Here, cyborg refers to a symbi-
otic bio-machine system, consisting of both organic and computing components.
Tightly-coupled connection between the organic and computing parts is a key fea-
ture of cyborg intelligence.
There are two critical emerging technologies pushing the convergence of

machine and biological intelligence. The first one is BMIs, often with real-time
neurofeedback (NF). BMIs aim at a communication pathway in bridging the work-
ings of machines and the brain. BMIs operate at the nexus of thought and action,
exploiting the brains electrical signals to maneuver external machine actuators and
feeding the machine-coded neural information back to the brain to regulate the
brains behaviors. Bidirectional BMIs with real-time NF are promoting the connec-
tion of machines and the brain at multiple levels.
The second technology is neuromorphic computing, which further envi-

sions the closely-coupled connection of the organic and computing components
and machines. Neuromorphic computing uses digital, analog, and mixed-mode
digital/analog circuits and software systems to mimic biological models of neural
systems (such as perception, motor control, and multisensory integration). It of-
ten attempts to incorporate detailed behavior models of an individual neuron, even
including realistic neural conductance and ion channel models. The fine-grained
electric mimicry of the nervous system makes the machine side closer to the bio-
logical side. Thus, it reduces the difficulty in integrating machine and biological
intelligence.

The future of cyborg intelligence may lead towards many promising ap-
plications such as neural intervention, rehabilitation, medical treatment, and early
diagnosis of some neurological and psychiatric disorders. It may replace, repair,
assist, and augment human sensory-motor or cognitive functions. For example,
neuroprosthetics can replace a missing body part and still interface with the hu-
man nervous system and brain to increase precision and comfort of movements, or
memory chips for restoring and enhancing memory function. Cyborg intelligence
has the potential to make the bionic man reality. While cyborg intelligence has
many potential exciting applications, research in this area is still in the preliminary
stages.

3. Author:Daniel Zeng, Zhaohui Wu

Title: From Artificial Intelligence to Cyborg Intelligence
Published in: , IEEE Intelligent Systems, vol. 29, no. 5, 2014, pp. 24.
One of the primary and utilitarian goals of artificial intelligence research

is to develop machines with human-like intelligence. Great progress has been made
since the start of AI as a field of study. Generations of AI thinking, AI schools of
thoughts, and AI engineering have given us expert systems, artificial neural net-
works, outstanding chess-playing programs such as Deep Blue, autonomous ve-
hicles such as Stanley, and human-level performance question-answering systems
such as Watson. However, realizing human-like intelligent behavior, such as un-
guided learning, high-level reasoning and sense-making, and adaptability, still has
a long way to go.
One dominating research paradigm in AI has been based on the assump-

tion that various aspects of human intelligence can be described and understood
well enough to the extent that it can be simulated by computer programs through
smart representational frameworks and generic reasoning mechanisms. Despite
great progress enabled by this paradigm, its limitations have been well-recognized
by the research community. An alternativeor to a large extent, a complementary
paradigm (which has almost-as-deep roots and history)is gaining tremendous mo-
mentum lately and has attracted much attention. This perspective is based on the
realization that varying kinds and degrees of intelligence reside in humans, ani-
mals, and other kinds of biological systems. Mimicking and making use of such
biological intelligence at different levelshardware design and algorithmic princi-
ples, among othersin a more direct manner, could greatly influence the design of
AI systems, opening fresh pathways and application areas for AI.
Biological systems possess all kinds of sensory abilitiesvision, hearing,

olfactory, haptic, and gustatory senses, to name a few. They also adapt to changes
in external environments, and are capable of a range of cognitive functions. AI

systems could greatly benefit from biological intelligence, solving problems that
are still beyond the capabilities of the state of the art. For instance, image under-
standing is a relatively easy job for humans, yet it still challenges even the most
sophisticated AI algorithms.
The recapcha approach, as an example of collective intelligence, has demon-

strated the power of integrating biological intelligence and machine intelligence,
helping to digitize old printed material by asking users to decipher scanned words
from books that computerized optical character recognition failed to recognize.1
In such approaches, however, the linkage between human intelligence and ma-
chine intelligence is loose, in the traditional sense of human-computer interaction.
Recent years have seen quantum leaps in research dedicated to this linkage and the
enormous potential enabled by deeply connecting and integrating biological and
machine intelligence.
Biological beings and computer systems share some common physical

foundations. Communication in both biological nervous systems and computer
systems, for example, depends on electrical signals. Yet, the gap between these two
classes of vastly different systems is obvious. Thanks to new developments in neu-
roimaging technologies, such as functional magnetic resonance imaging (fMRI),
magneto encephalography (MEG), and positron emission tomography (PET), how-
ever, the gap is no longer insurmountable. These technologies allow us to observe,
in increasing levels of resolution and fidelity, the brains inner workings, and reveal
the brains structure and function. Furthermore, progress in brain-machine inter-
faces (BMIs) in the last decade has made possible direct communication pathways
between the brain and man-made systems at the signal level.
An-made systems at the signal level. These new developments represent

significant advances in cyborg intelligence.2 Cyborg intelligence aims to integrate
AI with biological intelligence closely and deeply by connecting computer sys-
tems and biological systems via BMIs, enhancing strengths and compensating for
weaknesses of both systems by combining the biological systems perceptive and

cognitive abilities with the computer systems computational power. The term cy-
borg was coined by Manfred Clynes and Nathan Kline in 1960,3 to describe a
being with both organic and synthetic parts. More broadly, cyborgs refer to sym-
biotic biological-machine systems, consisting of both organic and computing com-
ponents. Cyborg intelligence is a new research paradigm, aiming to combine the
best of both machine and biological intelligence.
At the core of cyborg intelligence is the closely-coupled connection of the

organic and computing parts. BMIs offer a communication pathway in bridging
this gap between the two. Such technology is helping us decode thinking-related
signals from the scalp, the dural cortex, and even subcortical areas. It also helps
connect the brain directly to the outside world. Neural signals can control machine
actuators, and machine-coded sensory information can be delivered into specific
areas of the brain. Through bidirectional BMIs, we can connect biological com-
ponents to machine components at multiple levels, building a hybrid intelligent
system of great promises.

Chapter 3
Cyborg Intelligence: Recent Progress

and Future Directions
The interconnection and tight integration of biological information processing

units and man-made computing components provides extensive information exchange
between biological organisms and computing devices, which spawned cyborg intelli-
gence. Artificial and biological intelligence begin to share common territory in provid-
ing sensation, perception, cognition, and locomotion. Many amazing results have been
achieved in various areas, including animals as sensors and actuators, mind-controlled
machines, neurochips, intelligent prosthesis, and neural rehabilitation. Recently, the open
Cyborg Olympics were held to drive the realistic combination of biological and artificial
intelligence and help disabled people better acclimate to daily life.
Typical and effective approaches to implementing engineering systems and

exploring research problems in cyborg intelligence are based on brain-computer (or neural-
computer) integration methods. Using these methods, computers can record neural ac-
tivity at multiple levels or scales, and thus decode brain representation of various func-
tionalities and precisely control artifi cial or biological actuators. In recent decades, there
have been continuous scientifi c breakthroughs regarding the directed information path-
way from the brain to computers. Meanwhile, besides ordinary sensory feedback such
as visual, auditory, tactile, and olfactory input, computers can now encode neural feed-
back as optical or electrical stimuli to modulate neural circuits directly. This forms the
directed information pathway from the computer to the brain. These bidirectional infor-
mation pathways make it possible to investigate the key problems in cyborg intelligence.

Although related research challenges cover the most fundamental problems

in AI, physiology, and psychology, we focused on the computational architecture of cy-
borg intelligence, especially for sensorimotor processes in the brain-computer integration
system. We first proposed a conceptual illustration and constructed a cognitive computa-
tional model for brain-computer integration systems. To understand neural representation
in the brain, we also explored encoding and decoding principles underlying the sensori-
motor loop, and then the computers implemented novel AI algorithms to enhance sensa-
tion and motor control functions of the overall brain-computer integration system. From
our theoretical and technological results, we implemented cyborg rats, monkey mind
control, and a rehabilitation demonstration as evaluating systems. Extensive experiments
show that the philosophy concept and computation architecture of cyborg intelligence is
promising for enhancing, repairing, or extending the intelligent capacity of both biologi-
cal and computing units.

3.1 Conceptual Framework of Cyborg Intelligence: A

Revisit
From the systematic perspective, a critical problem in cyborg intelligence re-
search is how to merge the brain with the computer at various scales. On the basis of
the similarity between brain function partition and corresponding computing counter-
parts, we presented a hierarchical and conceptual framework for cyborg intelligence.
The biological part and computing counterparts are interconnected through information
exchange, and then cooperate to generate sensation, perception, learning, memory, emo-
tion, and other cognitive functions. We argue that this involves two key aspects: first, the
cooperation between biological and AI units will output the functional units of cyborg in-
telligence, and second, the final form and paradigm of cyborg intelligence is determined
not only by this interconnection and cooperation but also by the merging of biological
and AI units.
For the sensorimotor process, our previous work abstracted the biological
component of cyborg intelligence into three layers: perception and behavior, decision
making, and memory and intention. We also divided the AI functional units into three
corresponding layers: sensor and actuator, task planning, and knowledge base and goal
layers. We also defined two basic interaction and cooperation operations: homogeneous
interaction (homoraction) and heterogeneous interaction (heteraction). The former repre-
sents information exchange and function recalls occurring in a single biological or com-
puting component, whereas the latter indicates the operations between the function units
of both biological and computing parts. Homoraction is also modeled as the relation-
ship between units within the same part. In the case of a single part in a braincomputer
integration system, it will reduce to a biological body or computing device just with ho-
moraction inside. Consequently, verifying the existence of heteraction is necessary for
cyborg intelligent systems.
We believe that learning and memory units are fundamental for problem
solving in the computational framework of cyborg intelligent systems. Biological learn-

Figure 3.1: Hierarchical conceptual framework for cyborg intelligence.1 Three layers are
abstracted to describe the interconnection between the biological organisms and comput-
ing machines.
ing paradigms (such as classical conditioning, operant learning, and insight learning) and
learning rules (such as sequential and recursive patterns) are essential to generate adap-
tive behaviors, whereas AI algorithms enable computing devices to be smart to perform
intelligent tasks. Cyborg intelligent systems interact with the environment to achieve
better behavior performance, which requires the biological and AI units to adapt to each
other within the system and adjust their actions according to changing environmental sit-
uations, thus possessing enhanced learning, memory, and problem-solving capabilities.
Neural plasticity and machine learning cooperate and merge to represent the cyborg in-
telligent systems learning capacity, usually in a semisupervised or unsupervised manner.
3.2 Cyborg Intelligence: Research Progress

THere, we summarize our teams recent advances in cyborg intelligence. We
denote these advances as numbers in the conceptual framework of cyborg intelligence in
Figure 2, roughly indicating their positions in the framework. We denote the neural mech-
anism of the primary somatosensory cortex and primary motor cortex as 1 in the figure
and the cross-model sensation fusion as 2.. The novel cognitive model for the machine

parts is labeled as 3. To bridge the biological and machine components, we presented

robust neural decoding algorithms and encoding approaches . Furthermore, to enhance
the capability and performance of the complete cyborg system, we explored both biolog-
ical and artificial reconstruction methods. This promoted the deep integration of neural
circuits and computing components. On the basis of this computational architecture, we
presented three demonstrative systems: sensation-augmented rat cyborgs, monkey hand-
gesture decoding, and mind-controlled scissor games.
Figure 3.2: Illustration of research progresses in the framework of cyborg intelligence.
3.2.1 Sensation-Augmented Rat Cyborgs
As typical brain-computer integration systems of animal as the actuators, rat cy-

borgs, or ratbots were developed to validate how animals can be enhanced by AI. Ratbots
are based on the biological platform of the rat with electrodes implanted in specific brain
areas, such as the somatosensory cortex and reward area. These electrodes are connected
to a backpack fixed on the rat that delivers electric stimuli to the rats brain. For vison-
enhanced ratbots, a minicamera is connected to the backpack to capture movement or the

Figure 3.3: Vision-augmented rat cyborg. The ratbot can recognize image indicators
and navigate complicated environments guided by commands from its computing com-
ponents.
surrounding environment. A computer analyzes video stream input and generates stim-
ulation parameters that are then wirelessly sent to the backpack stimulator to control the
rats navigation behavior by manipulating virtual sensation or reward. As Figure shows,
vision enhanced ratbots can precisely find human-interesting objects that is, human faces
and arrow signs, identified by object-detection algorithms.
We also explored the automatic training method, using computers only to

learn the intelligent training procedures and complete the ratbot training. Furthermore,
with a speech-recognition interface, ratbots can be enhanced with speech understanding
capacity and navigate according to human words.
3.2.2 Monkey Hand-Gesture Decoding
To verify the brain-to-computer neural information path, we implemented a

mind-reader demonstrating system in which a monkey brain controlled four gestures of
a robotic handgrabbing, hooking, holding, and pinching. We implanted two 96-channel
microelectrode arrays into both the premotor and primary motor cortex to record neuron
population spiking activity. We identified the spatiotemporal representation of the differ-
ent gestures and extracted the signal features using a fuzzy k-nearest-neighbor algorithm
that mapped the spikes to a specific gesture with high accuracy.

Figure 3.4: Robotic hand control with neural decoding. We used motor control com-
mands decoded directly from neural spikes of the primary motor cortex to manipulate the
robotic hand to demonstrate four gestures.
Furthermore, we designed a centerout behavior paradigm to investigate the

neurons responsible for kinematic positions, acceleration, and velocity. A dual-sequential
Monte Carlo adaptivepoint process-filtering algorithm exploited the time-variant neu-
ral tuning function from spiking events. The short time tuning of individual neurons
marginally contributed to the stable performance of brain-to-computer systems, validat-
ing the continuous trajectory control of the robotic arm by the monkey brain.
3.2.3 Mind-Controlled Rock-PaperScissors Game
Figure 3.5: Rock-paper-scissors game controlled by human electrocorticogram (ECoG)

signals. Hand gestures and their corresponding ECoG signals. And live demonstration.
We also evaluated the brain-to-computer neural information path in the hu-

man mind-control case. We implemented an electrocorticogram (ECoG)-based brain-
machine integration system to control a prosthetic hand, using brain signals from human
participants who underwent invasive ECoG monitoring for seizure localization. We an-

alyzed the spatiotemporal patterns of ECoG signals. We extracted the power spectrum
of high gamma frequency components(80 to 120 Hz) to decode rock-paperscissors hand
movements, and we then controlled the prosthetic hand to perform the gestures.

Chapter 4
RESULT AND DISCUSSION
By developing the demonstrative systems discussed here, we verified the basic

concept and simplified the computational architecture of cyborg intelligence. Integrating
biological and machine intelligence into a unified mode of intelligence is a challenging
but promising goal. We have identified three future directions for cyborg intelligence.
The first direction is to develop a multimodal sensation integration mecha-

nism and computational model in a large-scale association cortex. For neural representa-
tion in large-scale sensorimotor networks, casual and sequential relationships among the
association cortex remain mostly unexplored. How crossmodal sensations are integrated
in the temporary sequence in the association cortex and what their exact role is in motor
planning are two of the most important scientific problems for sensorimotor research in
cyborg intelligent systems. The network-based source imaging method using electroen-
cephalogram and fMRI is promising to provide efficient tools to help researchers collect
finer evidence to identify network dynamics.
From a holistic view of sensorimotor functions, it is essential to build a com-
plete circuit model to explain the information processing pathway for cross-modal sen-
sation integration and the association motor cortex. Such research would promote new
mechanisms and computing architectures in cognitive architectures. In terms of decod-
ing algorithms, there is an emerging research trend to utilize prior knowledge of neu-
roscience principles to formulate neural information analytics. Furthermore, statistical
machine learning methods leverage brain representation spatiotemporal structures to de-
code neural signals stably and accurately.

Another concern is the creation of a biologically plausible cognitive model.

We presented the awarenessbelief-goal-planning model to describe the coordination, co-
operation, and merging relationship between biological and computing counterparts. This
model was inspired by the multiagent framework and based on symbolic reasoning. Bio-
logically plausible architecture combing neural networks with logical reasoning will rev-
olutionize the cognitive model, especially in cyborg intelligent systems. For example, we
need novel feature representation and integration mechanisms to generate awareness and
other internal mental states, which requires end-to-end processing of both artificial sen-
sor inputs and neural-spiking activity information. Deliberation mechanisms and policy-
driven reasoning approaches should consider the implementation of the neural network
and generalize to various cognitive tasks in complicated environments.
Finally, the co-adaption and colearning between biological components and

computing components play a critical role in merging cyborg intelligence. Neural cir-
cuits have remarkable plasticity at various levels,from the firing rates of single neurons
to the dynamic structure of largescale networks. Apart from exploring robust biological
reconstruction of physiological functions, it is important to develop machine learning al-
gorithms to extract neural representation patterns and build correct associations between
cognitive goals. These mapping associations might expand knowledge frontiers on how
neural circuits and computing components can be unified as a single cyborg intelligent
component.

Chapter 5
CONCLUSION
It seem that in future we may have more and more artificial body parts like
eyes,nose,hands and leg.Also Cyborg system can act as a human body by improving the
quality and the life.
5.1 Advantages
1. There are also brain implants based on neuromorphic modelling .A deaf man has
had his inner ear replaced so he can now engage in telephone conversation. (this in
time will be upgraded so that he can hear music)
2. PARKINSON DISEASE: there are brain implants that help reverse the most dev-
astating symptoms of the disease.
3. CEREBAL PALSY: implants that help control tremors.
4. Using ratbots Simple commands were sent to them to manipulate their brains and
into following a pre-programmed route/command. This was successful and now
researchers are proposing to use them to search for land mines and buried victims
of earthquakes as they can move more quickly then human rescuers or existing
robots.
5. Prolongs life
6. Give a part of the body back
7. Improve the quality of life
8. It enable extra sensory input and to communicate in much richer way

9. Enables to lead a normal life.
5.2 Disadvantages
1. Robots can sense the world in ways that humans cannot-ultraviolet, x-rays, in-
frared and ultra sonic perception. So basically there is more dependence on cyber
technology.
2. They out perform humans intellectually in aspects of memory and mathematical

processing .
3. Cyborg have no problem thinking of the world around them in wider dimensions
(multiple) where as human beings are more restricted in that sense.
4. Cyborgs have physical limitations. Cyborgs do not heal Body damage normally,
but instead must be repaired. For example, broken limbs and damaged armour
plating must be replaced which can be expensive and time consuming.
5.3 Future scope

By the development of cyborg intelligence that, direct interface between the
brain and the Internet, like the brain-machine interface (BMI) described in the science
fiction manga Ghost in the Shell, has been successfully tested in laboratory experiments.
That extreme level of human-machine fusion aside, people’s lives would already be com-
promised by not having a mobile phone or smartphone, and in some cases, when we think
about those who lead a life that is dependent on such devices.

References
[1] Zhaohui wu, ”Cyborg Intelligence:Recent Progress and Future Direction”, IEEE
Intelligent Systems 2016.
[2] Z.Wuetal, ”Cyborg Intelligence:Towards Bio-Machine Intelligent Systems”,IEEE

Intelligent Systems,vol. 29, no. 6, 2014.
[3] Y. Wang et al, ”Visual Cue-Guided Rat Cyborg for Automatic Navigation”, IEEE
Computational Intelligence, vol. 10, no. 2, 2015, pp. 4252.
[4] D. Zeng and Z. Wu, ”From Artificial Intelligence to Cyborg Intelligence”, IEEE
Intelligent Systems, vol. 29, no. 5, 2014, pp. 24.

(www.royalcet.ac.in)

Cyborg Intelligence Recent Progress and

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Cyborg Intelligence Recent Progress and

Hochgeladen von

Copyright:

Verfügbare Formate

CYBORG INTELLIGENCE:RECENT

PROGRESS AND FUTURE DIRECTIONS

Department of Electronics and Communication Engineering

SUPERVISOR Head of Department

This report was evaluated on ..../..../........

I wish to record my indebtedness and thankfulness to all those who helped me to

I am thankful to Dr.V.Vijayarangan, Head of Electronics and Communication

Department of ECE i RCET

The interconnection and tight integration of biological information processing

Department of ECE ii RCET

LIST OF SYMBOLS AND ABBREVIATIONS v

3 Cyborg Intelligence: Recent Progress and Future Directions 14

4 RESULT AND DISCUSSION 22

Department of ECE iii RCET

3.1 Hierarchical conceptual framework for cyborg intelligence.1 Three lay-

Department of ECE iv RCET

RCET Royal College of Engineering and Technology

FMRI Functional magnetic resonance imaging

Department of ECE v RCET

Typical and effective approaches to implementing engineering systems and ex-

In recent decades, there have been continuous scientific breakthroughs re-

To understand neural representation in the brain, we also explored encod-

Department of ECE 1 RCET

and computation architecture of cyborg intelligence is promising for enhancing, repair-

1.1 Report Organization

Department of ECE 2 RCET

A rat robot is a type of animal robots, where an animal is connected to a

A miniature camera is mounted on the rat’s back to capture the scene in

Department of ECE 3 RCET

An animal robot is an animal that is connected to a machine system, usu-

Department of ECE 4 RCET

the detection results are used to control the rat robot.

Existing animal robots can be categorized according to three classes:

Stimuli were delivered to the brains of beetles to elicit, suppress, or modu-

Department of ECE 5 RCET

The second category of animal robots is based on aquatic animals. In

Electrical stimuli were delivered to the thalamic ventral posterolateral nu-

Department of ECE 6 RCET

Thus, the present study comprises neuroengineering and a preliminary study in

Department of ECE 7 RCET

2. Author: Zhaohui Wu, Gang Pan,Jose C. Principe,Andrzej Cichocki

Artificial intelligences goal is to design computers that process informa-

The unpiloted vehicles by the Stanford Racing Team successfully com-

However, human-level Al remains impossible to achieve. All b iolog-

As research continues, the line between humans and machines begins to

Department of ECE 8 RCET

man computer interaction manner.

Cyborg intelligence is dedicated to integrating AI with biological intel-

There are two critical emerging technologies pushing the convergence of

The second technology is neuromorphic computing, which further envi-

Department of ECE 9 RCET

Department of ECE 10 RCET

3. Author:Daniel Zeng, Zhaohui Wu

One of the primary and utilitarian goals of artificial intelligence research

One dominating research paradigm in AI has been based on the assump-

Biological systems possess all kinds of sensory abilitiesvision, hearing,

Department of ECE 11 RCET

The recapcha approach, as an example of collective intelligence, has demon-

Biological beings and computer systems share some common physical

An-made systems at the signal level. These new developments represent

Department of ECE 12 RCET

At the core of cyborg intelligence is the closely-coupled connection of the

Department of ECE 13 RCET