DOC - iitk.ac.in - Indian Institute of Technology Kanpur

kneewastefulAI and Robotics

Oct 29, 2013 (3 years and 7 months ago)

203 views


Asia
-
Link Programme: Applicati on Form, 5
th

Call for
Proposals,
2006

Page
1



EUROPEAN COMMISSION




Asia
-
Link Programme

Phase III

Grant Application Form

O
pen Call for Proposals


2006



Budget line
:

19.1002


Reference:
EuropeAid/123738/C/ACT/Multi


Deadline for receipt of applications:
19 October 2006




Na
me of applicant:

Indian Institute of Technology, Kanpur,
INDIA




Dossier No


(for official use only)



Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


2




NOTICE





All personal data (such as names, addresses, CVs, etc.) mentioned in your application form will be
processed in accordance with Regulati
on (EC) No 45/2001 of the European Parliament and of the Council
of 18

December

2000 on the protection of individuals with regard to the processing of personal data by the
Community institutions and bodies and on the free movement of such data.
Your repli
es to the questions in
this form are necessary in order to assess your grant application and they will be processed solely for that
purpose by the department responsible for the Community grant programme concerned. On request, you
may be sent personal data

and correct or complete them. For any question relating to these data, please
contact the Commission department to which the form must be returned. Beneficiaries may lodge a
complaint against the processing of their personal data with the European Data Pr
otection Supervisor at
any time (Official Journal L

8, 12.1.2001).





Please read and complete this form with all due care, in accordance with the guidelines for applicants
.


Please note that the procedure has changed
. T
he evaluation of your application w
ill only be performed if
your concept note
is

provisionally selected. Your app
lication will then undergo the
evaluation. The
eligibility conformity check will only be performed for the proposals that have been provisionally selected
according t
o the score
obtained after the
evaluation, on the basis of the supporting documents requested by
the
European Commission

and the Declaration by the applicant signed and sent
together with the
application
.


Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


3

CONCEPT NOTE

1. Summary of the action

1.1

Brief description of

the proposed action.


In the last decade robotics has stepped out from industrial applications into human society as helpers,
rehabilitation devices, welfare supervisors, and entertainment aids. Out of all these applications, human assistive
robots are li
kely to play a major role in improving the lives of persons with disability due to old age, disease or
injury. While ageing society of European countries need customized assistive robots, both European and Asian
countries need assistive robots for severely

disabled people whose proportion is uniform across the globe.
Although European countries have already made visible progress in the field of assistive robotics, the field is still
relatively new. It is needless to say that Asian countries like India are
la
gging

behind in this very important field of
research.
However a collaborative

project between Indian Universities and European Universities in the field of
assistive robotics will not only produce a large scale specialized man
-
power that will support bo
th the societies, but
also EU can take advantage of such a collaborative effort to win over other competing nations such as US and
Japan. Furthermore, assistive robotics is highly multi
-
disciplinary and involves many different research fields. This
is idea
l for linking universities from EU and Asia together since the broad coverage of different disciplines enables
us to collaborate better at the university level and generate higher impact. With this back
-
drop, the main objective
of our

proposal

entitled

A
Collaborative Academic Program for Innovation in Intelligent Assistive
Robotics


is to develop specialized man
-
power through rigorous basic and experimental research in major
aspects of assistive robotics that includes rehabilitation robots, wheelchair ro
bots and other mobility aides, and
manipulator arms for the physically disabled. Simultaneously the proposal initiates a novel method of interactions
between European and Asian Universities through synchronous e
-
learning, tele
-
robotics and exchange visits

so
that strength of each partner can be shared. In this regard, the project has a bold vision such as a student from an
ill
-
equipped lab can perform remote experiments on a costly physical set
-
up in another laboratory. Following
objectives are outlined i
n this project:



To develop highly skilled human resources


graduates with PhD and Masters’ degrees
-

in assistive
robotics through collaborative basic and experimental research.



To develop a long vision in assistive robotic research through two internati
onal workshops so that a far
-
reaching impact can be made in this direction. In this regard the help of experts from EUROP, EURON and
such bodies will be sought.



To
bring in a synchronism in assistive robotics education between EU and Asia through synchrono
us e
-
learning and tele
-
robotics. This would enable to
augment
graduate

level teaching courses
through live e
-
classes by respective experts in partner
-
institutes. So also students can perform remote experimentation
via internet using the concept of tele
-
rob
otics.




To conduct road
-
shows on assistive robotic systems in EU and India to increase awareness among
potential users.



To facilitate participants of this proposal to learn foreign culture in industrial and academic sectors that
would lead to mutual coope
ration between Europe and Asia.



To assist Institute of Engineering
(IOE), Tribhuvan

University, Nepal in the field of industrial automation
and robotics so that they can start doing indigenous research in the field of robotics. It is planned that IOE
will
be helped to set up a robotics lab as well as to develop course curriculum in robotics. IOE will then act
as a seed in Nepal to further disseminate the outcomes from this project to other universities in Nepal in
the future.

As part of the
human resource d
evelopment program, lead researchers in partner institute will jointly supervise
Ph.D. and Masters’ students
in following areas that has a common theme of
development of

a
low
cost visually

navigated smart wheel
-
chair equipped with intelligent robotic arm
s

including brain
-
operated assistive devices
:



a brain
-
computer interface (BCI) that allows a disabled person to control a smart wheelchair and robotic
manipulator combination;



a visual tracking system for operating the wheelchair as an automated guided veh
icle (AGV) to give
mobility;



an integrated navigation system with both embedded and off
-
board sensors for safe navigation of the
wheel chair;



a self
-
organising control architecture for a redundant manipulator for natural execution of actions desired
by th
e disabled user;



a self
-
adaptive learning system that evolves multiple neural networks incrementally from simple to complex
tasks for robotic control;



an evolvable hardware system for controlling a
ssistive robots and wheelchairs

adaptively
;



development of
an exo
-
skeletal robot for human hand support


mechanical design, sensor fusion, and
control
.

Fifteen

faculty members from six institutes
/Universities situated in four counties

(
India, Germany, UK
and Nepal
)
will supervise
22 Ph.D. and M.Sc.

students joint
ly for successful completion of the project.



Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


4

2. Relevance:

2.1

How relevant is your proposal to the needs and constraints of the target country(ies) or region?

As per a conservative estimate, 1 in 3500 of the world population may suffer from a neuro
-
musc
ular disorder

such
as motor neuron disease (MND)
. Such people may have no means to communicate with the external world at all.
They may be provided mobility and enhanced autonomy through an appropriately designed
practical brain
-
operated assistive robotic
system
. European countries such as UK and Germany are facing problems such as
ageing of society, lack of carers and high labour cost of caregivers which can be adequately solved using
customized assistive robotic systems. There are about 70 million people
with disabilities in the EU. However only
two rehabilitation robots have reached noticeable sales numbers on the market: the
Handy1
, and the
MANUS
.
Based on the 2001 census, 27.87% of the dis
abled people which amounts to 6
1 million suffer from movement
dis
ability in India alone. Additionally India needs low cost assistive robots for the blind, deaf and dumb people.
Above all, assistive and rehabilitation robots are urgently needed in health
-
care institutions of both EU and India to
take care of needy patien
ts.

Since the field of assistive
-
robotics is relatively new, both basic and experimental researches are
necessary to develop low
-
cost prototypes such that disabled people can afford them. This would require
specialized man
-
power in the field of assistive
robotics. India and Nepal are good sources of high quality man
-
power and European Universities have best research facilities. Thus EU
-
India
-
Nepal

network at the university
level can easily tap young man
-
power of India

and Nepal

to educate them in this fie
ld through a collaborative
research and teaching which allows exchange visits, advanced research under joint supervision and provision for
high quality experimental facilities.

Moreover robotic research is yet to take a shape in Nepalese context for which

the principal cause lies in
the lack of academic curriculum regarding robotics and automation. These courses are vital for overall research
interest as well as industrial development. Furthermore there is practically no infrastructure for development of
r
obotics and automation in the country apart from a small student group, Robotics Club at
the Institute of
Engineering (
IOE
)
,
Tribhuvan University, Nepal
. Thus the proposal aims to provide logistic support to IOE so that
young Nepalese students will be grea
tly benefited and will be a great source of specialized man
-
power. We are
very much convinced that EU, India and Nepal collaboration in assistive robotics will play a vital role to serve the
humanity and is a step forward for EU to compete with Japan and U
S in this cutting
-
edge technology.

2.2

What are the problems to be resolved and the needs to be met?

There is an urgent need for developing affordable assistive robotic systems for providing independence to ageing
and disabled people.
The primary focus o
f this proposal is
to develop high quality engineering graduate
researchers
to perform basic research as well as experimental research necessary

for generating

expertise to
build assistive robotic systems.

Research objective is not only to develop robust a
nd intelligent algorithms but also
to find answer
s

to build
ing

lost
-
cost prototypes. Simultaneously the need of IOE in terms of minimal infrastructure
and specialized curriculum will be

taken care of. Through joint supervision
, we will train young graduate
s from
Asia, set up potential common curricula and strengthen the links among all partner universities.

2.3

Who are the actors involved (final beneficiaries, target groups)?

Target groups are the partners associated with this project where the knowledge
frontier in the field of assistive
robotics will be pushed much forward. The project will ultimately benefit a large number of graduates from EU and
Asia as a result of our project, which would have laid down the foundation and mechanisms for sustained EU
-
Asia
collaboration. Both European and Asian disabled community will be final beneficiaries for which such long term
collaboration will yield low
-
cost health
-
care assistances.

2.4

What are the objectives and expected results?

The natural outcome of this endeav
our would be sp
ecialized manpower. Fifteen

lead researchers from six
institutes across four co
untries will guide 22

Ph.D. and Masters’ students. The research will be carried out in a
collaborative fashion creating a strong scientific network between Europe
an and Asian Universities. Research
results will be disseminated through scientific publications in major journals and conference procee
dings. Efficacy
of a low cost
brain
-
operated visually navigated smart wheel
-
chair equipped with intelligent robotic ar
ms will be
demonstrated through road
-
shows in EU and Asia. Remote experimentation facility through tele
-
robotic system
and distance education through synchronous e
-
learning will bring in s
ynchronization in education between EU and
ASIA. The overall objecti
ve will be the strengthening of EU
-
Asia links, especially among graduates and
universities. There will be better understandings among EU and Asia partners.

2.5

What is the added value of the action (what adds the action by reference to (central or local) g
overnment
action and actions implemented by non state actors)?

All partners have already attracted substantial national funding to carry out research in some aspects of assistive
robotics. Further, if Asia link programme supports this proposal, then it is
possible for us to get additional funding
from the respective
research support institutes of the individual nations such
as

DAAD
, Germany, Department of
Science and Technology (DST), India, Engineering and Physical Sciences Research Council (EPSRC), UK. Su
ch
additional funding can be used to recruit additional PhD students as well as for equipment support to develop low
cost proto
-
types. The local governments, e.g., Birmingham City Council in UK, are all very keen on establishing
close ties with Asian count
ries in higher education.
Recently a UK India Education and Research Initiative
(UKIERI) has been launched by UK and India governments to support collaborative projects between UK and
India.
The proposed project
thus
fits the strategy of

central and

local
governments very well and is likely to obtain
additional support from local governments for certain activities, e.g., dissemination of project outcomes.


Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


5

3. Methodology and Sustainability:

3.1

What are the main project activities?

Following are the main ac
tivities:



Human resource development

-

Fifteen

faculty members from six institutes
/Universities situated in four
counties

(
India, Germany, UK
and Nepal
) will jointly supervise 22 Ph.D. and M.Sc.

students

in the broad
area of intelligent assistive robotics.

These students will perform basic and experimental research to
develop necessary expertise to build low cost assistive robotic devices for disabled people.




Synchronism in a
ssistive robotics education between EU and ASIA
will be brought out
through
synchr
onous e
-
learning and tele
-
robotics.



Logistic support to IOE
Nepal
will be provided
for robotics education.



Two international workshops


one in India and other in Europe


will be organized. The project will kick off
with a work
-
shop in India in the first
month.
Ten

experts will be invited to help the collaborative team
s

to
properly project realistic goal. Also a course module on
Soft
-
computing applications in Assistive Robotics

will be formulated which will be floated as a graduate level 4
-
unit course. Ano
ther work
-
shop will be held
after 2.5 years in Europe for critical evaluation of the project outcome with the help of same set of experts.
Besides a popular graduate level course module on
Assistive Robotics

will be formulated at the end of this
second wo
rkshop.



Visibility
of
the outcome of this EU
-
Asia collaborative effort

will be achieved
t
hrough road shows as well as
through
disseminati
on of results in

conference proceedings and journal publications
.




FWBI will act as transfer unit for

exchange of ide
as from this consortium

to SMEs. So also Invacare UK
Limited will closely monitor the research outcome for translating to feasible technology.


3.2

Who will be your main implementing partners, what is the length of your relationship with them and how will
they be involved in the project?

We are six partners:
Indian Institute of Technology, Kanpur

(IITK), India
; International Institute of Information
Technology, Hyderabad(IIITH), India
; Institute of Automation
, University of Bremen, Germany

(IAT); University

of
Ulster, Magee Campus(UUM), UK
;

the University of Birmingham(UB), UK

and the

Institu
te of Engineering (IOE),
Tribhuvan

University,

Nepal.

This partnership was kick
-
started when
Dr. L. Behera (
LB
) from IITK

visited UUM in
the summer 2006 for collaboratio
n on brain
-
computer interface (BCI) ba
sed communication system using Quantum
Neural N
etworks. The team further expanded when LB contacted Prof. Graeser from IAT, the group known for their
research in Assistive Robotics and
Dr. G. Prasad (
GP
) from UUM

conta
cted Prof. Xin Yao from UB. IAT group has
already a research link with IIITH. Incidentally, IOE has a long term connection with IITK where key technological
information exchange takes place through specialized school
s

and quality improvement programme.
IIT
K has

been
interacting with IAT for

last two years in regard to assistive robotic research.
The
IIITH group have a very strong
research relationship with IITK group as well. UB has been collaborating with the Mechanical Engineering
Department of IITK since

2001, when an Asia
-
ITC grant was awarded to them. In a nut
-
shell, all six partners have
a coherent view about the proposed plan and are determined to work as a team to achieve the end
-
result. Lead
researchers of all six partner
-
institutes will share and
supervise various modules of research objectively
and
jointly
through exchange visits. The collaborative effort will be strengthened by involving senior Ph.D. students through
exchange visits where they interact with their fellow Ph.D. scholars for joint r
esearch and experimentation.
A total
of
5 new Ph.D. students
, 14 senior Ph.D. students and 3 M.Sc. students will carry out their thesis work through this
proposal.

T
heir thesis works

will be jointly supervised by lead researchers from six institutes. Altho
ugh IITK will
take the responsibility for overall coordination, all six partners will jointly work as a team to make the project
successful.

3.3

How will the project achieve sustainability?

The duration of the present proposal is three years. The strong n
etworking structure among European and Asian
Universities would enable each institute to be self
-
sufficient in assistive robotics in general. If necessary, we will
continue our collaboration through joint international project schemes such as DST, DAAD, As
ia
-
link and UKIERI.
Because the project’s overall goal is to strengthen the links between universities from EU and Asia and all the
partner universities have strong interests in long
-
term cooperation, we expect the partner universities will invest in
our f
uture activities
,

after
we have built up a sound infrastructure and mechanism for cooperation with the support
of this project.

FWBI, Bremen, a technology transfer organization, will assist the partner
-
institutes to develop
industry linkage so that this co
llaboration can be sustained in long term. In addition, Invacare UK Limited, a leading
manufacturer of smart wheel chairs and assistive robotic products will closely work with the partner
-
institutes to
take the research output from this joint venture to te
chnology development.

3.4

Will it have multiplier effects?

Yes, specialized man
-
power developed from this project will spread across the societies in both Europe and Asia
through specialized jobs, entrepreneurship and academic and industrial research.

C
ontinuous interaction with
FWBI Bremen and Invacare UK limited will result in a long
-
term collaboration in a framework of academia
-
industry
linkage which will have a greater dissemination effect of the outcome of this project.





Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


6


4. Expertise and operatio
nal capacity:

4.1

What is the experience of your organisation in project

management?

IITK

has been consistent
ly rated as number one tech
-
school

both in teaching and research in India over the years.
Every year the institute gets funding
of
more than Rupees

500 million for execution and implementation of various
projects and consultancies.

4.2.


What is the experience of your organisation and your
partner(s) of the issues to be
addressed?

IIT, Kanpur, India

IITK research group consists of four lead research
ers, namely, Dr. Laxmidhar Behera (LB), Prof. P.K. Kalra (PKK),
Dr. Ashish Dutta (AD) and Dr. Y.N. Singh(YNS); five senior Ph.D. students, and two new Ph.D. students

to be
recruited in this project
. The group has extensive research experience in intellige
nt control, visual motor
coordination, advanced neural learning algorithms, advanced control architecture, visual tracking, obstacle
avoidance, humanoid robots and data modelling using soft
-
computing approaches.
The group is also well
equipped with advance
d experimental set
-
ups such as a IRB140 6DOF manipulator with stereo vision system,
SGI Prism server for visualization, a 7 DOF power
-
cube robot (redundant manipulator), Patrolbot for
experimentation as automated guided vehicle (AGV) and many prototypes o
f humanoid robots such as static and
dynamic walkers.

Simultaneously the group is investigating sponsored research with a funding of Rupees 25
million.

IIIT, Hyderabad, India

IIIT, Hyderabad research group consists of Dr. Madhav Krishna, Dr. Bipin Indur
kh
ya, Dr. C. V. Jawahar, two senior
PhD stud
ents, one M.Sc. student, and one

n
ew Ph.D. students

to be recruited in this project
. The research focus
involves multi robotic systems, mo
bile robotics and hybrid legged
-
wheeled vehicles. Research projects include
probabilistic localization and mapping, SLAM, planning under constraints, multi sensor surveillance systems for
multi target detection, cooperative localization and navigation


and force feedback control for hybrid legged
wheeled vehicles and visual inf
ormation processing. A micro robot is being developed with informal collaboration
with CAIR as a part of platform development to get insight into hardware and embedded system related issues as
well as to jump start swarm robotics research. Previously a fo
ur wheeled steered
ro
bot was developed in
collaboration with CMU.

Institute of Automation (IAT), University of Bremen, Germany

IAT is
one of the premier research institutes in Bremen and Germany in the field of automation and robotics.
IA
T
research group
consists of
Prof
. Axe
l Gräser,

two senior Ph.D. students

and one new Ph.D student to be
recruited in this project
.

The institute is a pioneer in the field of assistive robotic research while it has strong
expertise in areas such as
Computer Vision, Augm
ented Reality, Signal Processing, Robust Image processing
methods and Human Bra
in Interface. There are
eigh
teen international PhD students

and several master and
diploma s
tudents as research assistants who are dedicated to research full
-
time.


University

of Ulster, Magee, UK

The UUM research team consisting of three academic staff Dr. G. Prasad (GP), Prof. T. M. McGinnity (TMM), and
Dr
. Liam
P. Maguire
(LPM) and t
hree

existing senior PhD students Pawel Herman (PH)
,

Alan Browne (AB),

and
Philip Vance (PV)

is part of Intelligent Systems Engineering Laboratory (ISEL) research group based at the Magee
Campus of the University of Ulster (UU) which is one of the largest universities in the UK. The university has also
committed to recruit a new PhD student for in
clusion in the team. The research areas are Brain
-
Computer Interface
(BCI), assistive robotics, intelligent embedded systems, bio
-
inspired systems, evolvable hardware, self
-
organising
and self
-
adapting computational systems, self
-
repair of complex embedded

systems, and intelligent monitoring and
control of industrial systems. The ISEL group has several EU collaborations; are board members of
EURON:European Robotics Network, (an EU

funded network of excellence)
and as a result has a range of
international li
nks in the European robotics community. The research work on the BCI is being undertaken in
collaboration with Royal Victoria Hospital, Belfast and Guger Technologies, a spin off company from the Graz BCI
group of the Technical University of Graz, Austria,

involved in manufacturing of BCI systems.

University of Birmingham, UK

The Natural Computation Group led by Prof. Xin Yao at UB has more than 50 full
-
time researchers
.

Prof.. Xin Yao,
Dr. Xiaoli Li and two
senior
PhD students will participate in this pr
oject. The Group, including the Centre of
Excellence for Research in Computational Intelligence and Applications (CERCIA), has currently more than £4
million external funding. It completed a very successful Asia
-
ITC project in 2002, where partner countries

include
India and Germany. In terms of UB, it is one of the leading European universities. It has more than £70 million
external funding each year. It has rich experience of managing large EU projects, more than £10 million each year.
UB also has right re
search expertise needed for the project, especially in computational intelligence techniques for
intelligent systems (including evolutionary computation, neural networks and intelligent robotics). UB
’s

standing in
the world is partially indicated by the ed
itorship of major IEEE Transactions and many other journals, and
competitive grants from a number of different sources. It is also worth noting that UB has an active EU grant on
intelligent robotics at the moment. In terms of industrial links,
UB

ha
s

worke
d with over 100 companies in the last
few years, including both large multinationals and local SMEs.

IOE, Tribhuv
an

University, Nepal

The IOE research team consists of
two lead researchers

Dr. Arabinda Mishra,
and Mr.
AC
Mahendra and

2

M.Sc.

students
. IOE
needs logistic support to embark upon robotic education in a full swing.


Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


7

FULL
APPLICATION FORM


Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


8

I.

THE ACTION

1.

DESCRIPTION


1.1

Title
:
A Collaborative Academic Program for Innovation in Intelligent Assistive
Robotics




1.2

Location
(s)




Indian Institute of Technolo
gy, Kanpur

(IITK)
, India



International Institute of Information Technology, Hyderabad

(IIITH)
, India



Institute of Automation (
IAT
)
, University of Bremen, Germany;



University of Ulster, Magee Campus(UUM), UK



University of Birmingham(UB), UK.



Institute of E
ngineering

(IOE)
, Tribhuvan

University,

Kathmandu,
Nepal




1.3

Cost of the action and amount requested from
the European Commission


Total eligible cost of the action

Amount requested from the
European Commission

% of total eligible cost of action

EUR
1 069

666


EUR

748 852


%

70.
00


NB: The % of total eligible cost of the action is calculated by dividing the Amount requested from the
European
Commission

by the Total eligible costs of the action and multiplying by 100.


Please note that the cost of the a
ction and the contribution requested from the
European
Commission

have

to be
expressed in EURO





























Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


9

1.4

Summary

(
Maximum 1 page
)


Duration of the
action

36

months

Objectives of
the action

Overall objective(s)



Human resource developm
ent:

15 faculty members from si institutes (UK, Germany,
India and Nepal
) will supervise 22

Ph.D. and masters students jointly
in the field of
assistive robotics.



Curriculum Development: Institute of Engineering, IOE, Nepal will be assisted in robotic

ucation.

Specific objective



To develop specialized man
-
power in assistive robotics through collaborative basic and
eperimental research.



To develop a long vision in assistive robotic research through two international
workshops so that a far
-
reaching im
pact can be made in this direction. In this regard the
help of eperts from EUROP, EURON and such bodies will be sought.



To augment

graduate level teaching courses in Asian Universities through synchronous
e
-
learning system from European Universities and
vice
-
versa



To conduct practical eperiments on assistive robotic systems in European University
from an Asian university through tele
-
robotic concept and to duplicate this facility among
Asian Universities.



To conduct road
-
shows on assistive robotic system
s in EU and India to increase
awareness among potential users.



To assist Institute of Engineering
, Tribhuvan

University, Nepal in the field of industrial
automation and robotics so that they can start doing indigenous research in the field of
robotics. It

is planned that IOE will be helped to set up a robotics lab as well as to
develop course curriculum in robotics.

Partner(s)



Indian Institute of Technology, Kanpur

(IITK)
, India



International Institute of Information Technology, Hyderabad

(IIITH)
, India



I
AT, University of Bremen, Germany;



University of Ulster, Magee Campus

(
UUM), UK



University of Birmingham

(UB), UK.



Institute of Engineering

(IOE)
, Tribhuvan

University, Nepal



FWBI, an organization for transfer of
technology


an associate



An SME
Invacare
UK Limi
ted


an associate

Target group(s)

Partner Universities of this proposal will directly benefit in terms of sci
entific knowledge gained,
suffi
cie
ncy attained in assistive robotics

and specialized man
-
power.

Final
beneficiaries

Health
-
care system
will be benefitted so also disabled community. Specialized man
-
power
developed through this proposal will contribute further to research in assistive robotics, so also
they may play a significant role

in

industrial sectors such as car
-
manufacturing units
in India.
EU can take benef
it of such a scenario to attain a competitive edge
.

Estimated
results



19 Ph.D. students and 3 masters
’ students

wi l l be benef i t
t ed t hr ough j oi nt s uper vi s i on
i n an i nt er nat i onal c ol l abor at i ve ambi enc eK



S y nc hronis m in E U
-
A s ia edu
c at ion

s y s t em in t he field of as s is t ive robot ic s

t hrough
s y nc hronous e
-
learning and t elerobot ic s will be brought out.



Through FW B I t rans fer of t ec hnology t o S ME s will be ens ured.



Fur t her pr omot i on of EU
-
As i a l i nk.


Ma in a c t ivit ie s



Human res ourc e developme




Synchronism in Assistive robotics education between EU and ASIA through
synchronous e
-
learning and tele
-
robotics.



Logistic support to IOE Nepal for robotics education.



Visibility of EU
-
Asia collaborative efforts through international workshops, dissemin
ation
of results through conference proceedings and journal publications as well as road
shows.



FWBI will act as
technology
transfer unit for echange of ideas from academia to SMEs.

So also

Invacare UK Limited will closely monitor the research outcome for

translating to
feasible technology.



Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


10


1.5


Objectives

(Maximum 1 page)

The main objectives of the proposal
A Collaborative Academic Program for Innovation in Intelligent Assistive
Robotics

are in the area of human resource development and Curriculum Deve
lopment. Design of innovative
intelligent assistive robots for human assistance is necessary for EU and
Asia

and main thrust is given to joint
supervision of Ph.D. and Masters’ students in basic and experimental research in this cutting edge
-
technology. 15

faculty members will jointly supervise 22 Ph.D. and Master’s students for carrying out high
-
quality technological
innovation in intelligent assistive robotics. Since the transfer of results to new or enhanced products is a core
element in engineering rese
arch, in this proposal Invacare UK limited, a manufacturing sector in rehabilitation
product development and FWBI Bremen, a transfer organisation are included. Faculty members will interact with
industry in Asia and Europe to learn from first hand experien
ces
of
the different methods and approaches in
European and Asian Companies. In a cooperating world, such industry
-
academia interactions will enhance EU
-
Asia scientific link that will have long
-
term sustainability.


Intelligent assistive robotics is highly

multi
-
disciplinary and involves many different research fields. This is an ideal
co
-
theme for linking universities from EU and Asia together since the broad coverage of different disciplines
enables us to collaborate better at the university level and gen
erate higher impact. In this regard, the Ph.D. thesis
topics are so selected that expertise of partner
-
institutes
complement

each

other resulting
in
a win
-
win situation for
both EU and Asia. Ph.D. and Master’
s

students
will be

trained under joint supervisi
on while availing the best
facilities that are there in partner institutes and associated manufacturing sector. In addition, specialized man
-
power development in smart wheelchairs and redundant manipulator systems as envisioned in this proposal will
have a
n additional impact in industrial robotics given the fact that India has a large market in automobile industrial
products where EU is going to play a major role.


Another key aspect of this proposal is that partner
-
Universities will associate with an orga
nization FWBI which is
specialized in the transfer of research results from academia to SMEs and large companies which are active in the
field of intelligent industrial and assistive robots.


The
present
consortium will enhance their existing curricula wi
th the goal to complement each other much better
and define a standard in the education for assistive robots. New methods for long distance learning and learning in
different cultures will be developed.
In this context, two important objectives of the prop
osal are:




Development of a synchronous e
-
learning scheme among partner
-
institutes to augment graduate level
courses.



Partner
-
institutes will be connected through tele
-
robotic systems to create one virtual laboratory so that
students can verify their algo
rithms through remote experimentation on a physical set
-
up available in
another laboratory.


The proposal is so planned that each new Ph.D. student recruited in
an
Asian University will spend one year in EU
partner
-
institutes while a master

s student will
carry out his/her master

s thesis in one of the EU partner
-
institute
s
.
Similarly new Ph.D. student recruited in EU will visit India for one year. Besides senior students from EU and Asia
will have exchange visit to get the maximum benefit of the joint supe
rvision. Given the fact that out of 22, 15
students are from Asia, it is hoped that the maximum benefit of the Asia
-
link program will be derived by properly
training the great human
-
potential available in India and Nepal while maximally utilizing the facil
ities in EU and
Indian Universities. Simultaneously the provision of 11 exchange visits of lead researchers for joint supervision will
lead to unique exchange of ideas and vibrant research atmosphere. If this proposal is funded by Asia
-
Link, then
award of

Joint Ph.D. degrees among Asia
-
EU partners will be pursued. Such a provision will usher a new era in
Asia
-
EU scientific network and will have multiple effects.


The technology transfer institute FWBI will gather information about the research results of t
he different partner
-
institutes and actively search for the transfer in new products and the upgrade of existing products. The transfer
institute will evaluate the different approaches in Asia and Europe and will feed this information back to the
researche
rs of the partner
-
institutes. It is expected that such a method will generate a consortium from academia
and industry to design new or enhanced products that would sustain the present collaboration.


Final outcome of project work will be demonstrated thro
ugh road
-
shows in IAT, Bremen, UUM, UK, IITK, India and
IIITH, India where SMEs will be invited.
Collaborative deals

with interested
SMEs will

be made who would like to
adopt our technological innovations in assistive robotic systems. Two international
wor
kshops

will be organized to
take further guidance from international experts as well as for the critical assessment of the progress.


Finally logistic support to IOE Nepal will be given so that IOE will act as a seed in Nepal to further disseminate the
o
utcomes from this project to other universities in Nepal in the future.



Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


11


1.6

Justification
(Maximum 3 Pages)

1.6.1

R
elevance of the action to the objectives
and priorities
of the programme

The relevance of the proposal lies in the area of human resource development
and curriculum development.
Fifteen

experienced faculty members will jointly supervise 14 senior Ph.Ds,
5 new PhDs and
3 senior

master’s
students
.

Th
us a total of 37

skilled persons in the field of assistive robotics will be benefited whose impact at the
g
lobal level will surely be felt after the successful implementation of the project. Two international workshops will
help this collaborative team to closely interact with the best experts in this field. While the project strongly
connects EU and Asia Univ
ersities through research collaboration, simultaneously synchronism in education in the
field of assistive robotics between EU and Asia will be achieved through e
-
learning and tele
-
education.
A special
school for IOE, Nepal will be conducted to foster the
robotics education there while formulating
up
-
to
-
date course
-
curricula

in robotics.


1.6.2

I
dentification of perceived needs and constraints in the target countr
y/
ies
, in particular in
the region(s) concerned
.

As per a conservative estimate, 1 in 3500 of the w
orld population may suffer from a neuro
-
muscular disorder. Such
people may have no means to communicate with the external world at all. They may be provided mobility and
enhanced autonomy through an appropriately designed
practical
brain
-
operated assistive

robotic system
.
European countries such as UK and Germany are facing problems such as ageing of society, lack of carers and
high labour cost of caregivers which can be adequately solved using customized assistive robotic systems. There
are about 70 millio
n people with disabilities in the EU. However only two rehabilitation robots have reached
noticeable sales numbers on the market: the
Handy1
, and the
MANUS
. Based on the 2001 census, 27.87% of the
disabled people which amounts to 6.1 million suffer from mo
vement disability in India alone. Additionally India
needs low cost assistive robots for the blind, deaf and dumb people. Above all, assistive and rehabilitation robots
are urgently needed in health
-
care institutions of both EU and India to take care of ne
edy patients.

Since the field of assistive
-
robotics is relatively new, both basic and experimental researches are
necessary to develop low
-
cost prototypes such that disabled people can afford them. This would require
specialized man
-
power in the field of
assistive robotics. India and Nepal are good sources of high quality man
-
power and European Universities have best research facilities. Thus EU
-
India network at the university level can
easily tap young man
-
power of India to educate them in this field thr
ough a collaborative research and teaching
which allows exchange visits, advanced research under joint supervision and provision for high quality
experimental facilities.

Moreover robotic research is yet to take a shape in Nepalese context for which the p
rincipal cause lies in
the lack of academic curriculum regarding robotics and automation. These courses are vital for overall research
interest as well as industrial development. Furthermore there is practically no infrastructure for development of
robotic
s and automation in the country apart from a small student group, Robotics Club at IOE, TU. Thus the
proposal aims to provide logistic support to IOE so that young Nepalese students will be greatly benefited and will
be a great source of specialized man
-
po
wer.

Our proposed project can leverage on the existing infrastructures that some partners have and as a result
save project effort and cost. For example, UUM has a very good facility for BCI based experiment, IAT has been
pioneer in the smart
-
wheel chair
technology, and IITK has very good facility for experimentation in assistive
robotics. IAT runs a master’s course in robotics and automation
,

UUM runs several MSc courses in the broad area
of intelligent systems

and the University of Birmingham has special
ised MSc in natural Computation that can
provide many useful modules to the students.

1.6.3

Description of the

target group
(
s
)

and final beneficiaries
and estimated number

IIT, Kanpur, India

IITK research group consists of four lead researchers, namely, Dr. L
axmidhar Behera (LB), Prof. P.K. Kalra (PKK),
Dr. Ashish Dutta (
AD) and Dr. Y.N. Singh; and five senior Ph.D. students, and two

new Ph.D. students
.
The group
has extensive research experience in intelligent control, visual motor coordination, advanced neur
al learning
algorithms, advanced control architecture, visual tracking, obstacle avoidance, humanoid robots and data
modelling using soft
-
computing approaches. This proposal has been prepared taking into account of the area of
expertise necessary to develo
p an assistive robotic systems.
The group is also well equipped with advanced
experimental set ups such as a IRB140 6DOF manipulator with stereo vision system, SGI Prism server for
visualization, a 7 DOF power
-
cube robot (redundant manipulator), Patrolbot

for experimentation as automated
guided vehicle (AGV) and many prototypes of humanoid robots such as static walker and dynamic walkers.

Simultaneously the group investigating sponsored research with a funding of Rupees 25 million.



IIIT, Hyderabad, Indi
a

IIIT, Hyderabad research group consists of Dr. Madhav Krishna, Dr. Bipin Indur
khya and Dr. C. V. Jawahar and 2
senior students, one New Ph.D. students and one senior

Masters’ students. The research focus is diverse and
involves multi robotic systems, m
obile robotics and hybrid legged wheeled vehicles. Research projects include
probabilistic localization and mapping, SLAM, planning under constraints, multi sensor surveillance systems for
multi target detection, cooperative localization and navigation


and force feedback control for hybrid legged

Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


12

wheeled vehicles and visual information processing. The center's activities include algorithms for off
-
the
-
shelf
mobile platfo
rms as well as platform
devel
opment. A micro robot is being
developed with informa
l collaboration
with CAIR as a part of platform development to get insight into hardware and embedded system related issues as
well as to jump start swarm robotics research. Previously a four wheeled steered bot was developed in
collaboration with CMU. The

research is being funded by DST, CAIR, DRDO, MCIT, HP Labs among others.


Institute of Automation (IAT), University of Bremen, Germany

IAT is one of the premier research institutes in Bremen and Germany in the field of automation and robotics. IAT
has th
ree research groups: "Robotics and Process Automation", "System Dynamics and Control" along with "Real
-
Time Systems". Currently, the institute has two professors; Prof. Axel Gräser and Prof. Thiele Georg, both
experienced and actively participating in cur
rent research projects of the institute. Apart from these professors, the
institute has three other associate professors,
Prof. Rüdiger

Kutzner, Prof. Kai

Müller and Prof. Jörg

Schultz. The
institute also has four experienced researchers Dr. Dorin

Aite
anu, Dr. Bernhard

Graimann, Dr.

Ola

Friman, and Dr.
Ivan

Volosyak each of them concentrating their research on
several
topics namely; Computer Vision, Augmented
Reality, Signal Processing, Robust Image processing methods and Human Brain Interface. There ar
e also,
eighteen international PhD students; and several master and diploma students as research assistants.

The institute
is currently pursuing active research in dealing with different aspects of automation and
robotics. It covers topics like computer

vision, robust image processing, reactive sensor based architecture, brain
-
computer interface, human machine interface, augmented reality, mapped virtual reality, motion planning, software
framework for service robotics, and control systems. There have be
en different successful projects in the
respective areas which are either completed or currently being pursued in collaboration with different research
partners around Europe and Asia.
To name a few from the many project partners, the institute has strong
colloborations with Bien’s System Control Laboratory (BSCL),
KAIST, Korea,
Bremen Zentrum für Mechatronik
(BCM), Forschungsverbun
d Logistik (FOLO), Zentrum für K
ognitionswissenschaften (ZKW)
.

In teaching the IAT is responsible for the international Master
program on ‘Information and Automation
Engineering’, a 2 year course program completely taught in English. IAT is also responsible for the specialization in
‘Automation and Robotics’ for the B.Sc. and M.Sc. studies in Systems Engineering.

On an academic

note, each year several students visit the institute as exchange students under the
umbrella of exchange student programs like ERASMUS, SOCRATES, DAAD, etc.

The institute boasts
excellent laboratory setups, hardware, and software required in the fields o
f robotics,
process automation, and control systems. Students are trained to attain theoretical and practical knowledge on
cutting edge technologies over the available range of laboratorial facilities.

The institute gets funding from several research proje
cts that are currently being carried out. The
undertaken research projects are supported by various funding institutes and industries ranging from the local
government based research funding institutions and the European Commission.
Presently
the EU
-
FP6 re
search
programs BRAINROBOT is coordinated by IAT. On a national level the research programs AMAROB and MLK are
coordinated by IAT. Prof. Graeser is a visiting advisory professor at KAIST, Korea and member of the FWBI board.


University of Ulster, Magee,
UK

The UUM research team consisting of three academic staff Dr. G. Prasad (GP), Prof. T. M. McGinnity (TMM), and
Dr. Liam
P. Maguire
(LPM) and three existing senior PhD students Pawel Herman (PH), Alan Browne (AB) and
Philip Vance (PV), is part of Intellig
ent Systems Engineering Laboratory (ISEL) research group based at the
University of Ulster, Magee Campus (UUM) which is one of the largest universities in the UK. The

group research
addresses

complex problems in several areas such as Brain
-
Computer Interfa
ce (BCI), assistive robotics,
intelligent embedded systems, bio
-
inspired systems, evolvable hardware, self
-
organising and self
-
adapting
computational systems, self
-
repair of complex embedded systems, and intelligent monitoring and control of
industrial sys
tems. Senior PhD students associated with this proposal are researching in the area of BCI , optical
flow
-
based navigation and wireless network
-
based localisation of a motorized wheelchair . The ISEL group has
several EU collaborations; are board members o
f EURON2:European Robotics Network, (EU funded network of
excellence) and as a result has a range of international links in the European robotics community. The research
work on the BCI is being undertaken in collaboration with Royal Victoria Hospital, Be
lfast, UK and Guger
Technologies, a spin off company from the Graz BCI group of the Technical University of Graz, Austria, involved in
manufacturing of BCI systems. The group has been successful in attracting research funding from a range of
sources such a
s UK’s Engineering and Physical Sciences Research Council (EPSRC), EU and local government.
More recently, a 5 year Research Council UK (RCUK) Fellowship has been secured to research on biologically
-
inspired hardware systems.

Some relevant recent grants a
re:



SenseMaker
-

a multi
-
sensory, task
-
specific adaptable perception system, (IST
-
2001
-
34712) Funded by
the European Commission under IST research programme on Life
-
like Perception, (0.5M Euros), 2002
-
05.



Neurofeedback for Enhanced Use of Mental Practice
in Stroke Survivors and EEG
-
based Brain
-
Computer
Interface, funded by Department of Learning, Govt. of N. Ireland, (£5k), 2006.

In terms of experimental facilities, BCI projects are supported by a state
-
of
-
the
-
art EEG
-
based BCI system bought
from Guger Tec
hnologies OEG, Austria. The system is installed in a screened room, which is part of the advanced
experimental facility of ISEL at the School of Computing and Intelligent Systems. It has 56 EEG channels and 8

Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


13

electromyogram (EMG) channels interfaced to a h
igh power industrial PC for Bio
-
signal processing under the
MATLAB environment. The system also has an additional stimulation unit for interfacing both analogue and digital
devices such as prosthetic limbs. A mobile BCI experimental setup is also available

for on
-
site experimentation
outside the lab in normal working environment. ISEL also has a well
-
resourced mobile robotics laboratory. It has a
computer
-
controlled power wheel
-
chair and a robotic manipulator, Peoplebot, a fleet of Khepera robots. The group

has recently procured a high specification FPGA platform (with an associated cost of £250K) that will enable the
implementation of very large scale neural networks that would have performance comparable to sub
-
regions of the
human brain.


The University o
f Birmingham, UK

The Natural Computation Group led by Prof. Xin Yao at UB has more than 50 full
-
time researchers (including PhD
students). Prof.. Xin Yao, Dr. Xiaoli Li and two PhD students will participate in this project. The Group, including
The Centre
of Excellence for Research in Computational Intelligence and Applications (CERCIA), has currently
more than £4 million

of

external funding. It completed a very successful Asia
-
ITC project in 2002, where partner
countries include India and Germany. In terms

of UB, it is one of the leading European universities. It has more
than £70 million external funding each year. It has rich experience of managing large EU projects, more than £10
million each year. There is a well
-
established infrastructure in UB for man
aging externally funded projects and for
ensuring their success.
UB

also ha
s

right research expertise needed for the project, especially in computational
intelligence techniques for intelligent systems (including evolutionary computation, neural networks a
nd intelligent
robotics).
UB’s

standing in the world is partially indicated by the editorship of major IEEE Transactions and many
other journals
, and
competitive grants from a number of different sources. It is also worth noting that
the UB has

an
active E
U grant on intelligent robotics at the moment. In terms of industrial links, we have worked with over 100
companies in the last few years, including both large multinationals and local SMEs.


IOE, Tribhuvan

University, Nepal

The IOE research team consists
of
two
lead researchers


Dr. Arab
inda Mishra
and Mr. Mahendra and 2 master’s
students. IOE which was established in 1972 under the Tribhuvan University, Nepal has become not only a leading
institute in Nepal but a reputed institute
at

regional level. In

this journey it has received many academic
development projects and implemented successfully. Among them the institute has able to receive and manage
the biggest engineering project of 20 million US Dollar
during

1989
-
1999 to enhance the engineering educ
ation in
Nepal as a whole with the assistance of donor agencies CIDA (Canada), SDC (Swiss), IDA (World Bank) etc.

Although there is practically no infrastructure for development of robotics and automation in Nepal, the Institute of
Engineering, T.U. has be
come the foremost leading institution devoted
to

develop
ing

and strengthening
the field of

robotics in Nepal.
A
Robotics Club
was

established in 2001 in the institute. Since th
en

it has achieved following;



Has built some robots mainly for sports and partic
ipated in some international competitions held in
countries like, Japan (2002), Thailand(2003),

and
China (2005).



In two international robot competitions in year 2004 and 2005, organized by IIT Bombay, India, it achieved
third position with 500 dollars p
rize in 2004 and ‘Best Idea and Implementation Award’ in 2005.



Also presently it is engaged in developing some prototypes of ‘industrial robot’, ‘obstacle sensing robot’,
‘escalator’,
and
‘solar car’.

1.6.4

R
easons for the selection of the target group
(
s
)

and i
dentification of their needs and
constraints.
How does the Action contribute to the needs of the target group(s) and final
beneficiaries?

Interestingly 5 partner institutes have been working in intelligent systems design for application in robotics. IITK
a
nd UUM have been collaborating
for

last one year while UB has been collaborating with the Mechanical
Engineering Department of IITK since 2001, when an Asia
-
ITC grant was awarded to them. IIITH and IAT Bremen
have been collaborating in assistive robotics a
s well. Although IAT, Bremen is ahead of other partners in assistive
robotics research, coincidentally all partners have shown their interests to work united in this field of assistive
robotics as everybody’s expertise is complementary to each other. UB g
roup who are very strong in evolutionary
algorithms will take the lead in guiding research in evolutionary ensemble neural networks for performing complex
tasks as well as
in
evolvable hardware design for fault
-
tolerant control. IAT Bremen will lead the re
search in visual
navigation while UUM will lead the research in BCI based natural communication between disabled and robotic
system. IITK will lead the research in robot control and exo
-
skeleton while IIITH will lead in integrated navigation
system. All th
ese aspects of research is complementary for all partners who will gain expertise in this multi
-
disciplinary field of assistive robotics. Simultaneously all five partners will assist IOE, Nepal to come up in robotic
research so that the seed in Nepal is pr
operly groomed to further disseminate the outcomes from this project to
other universities in Nepal in the future.







Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


14


1.7

Detailed description of activities
(Maximum 9 Pages)


Title:

A Collaborative Academic Program for Innovation in Intelligent Assistive
Ro
botics


The project proposal
includes activities in four main areas
as
discus
sed below:


1.7.1
Specialized Man
-
power Building Through Basic and Experimental Research
in

Assistive Robotics


Common Theme for the Ph.D. thesis work
:


As per a conservative est
imate, 1 in 3500 of the world’s population may suffer from a neuro
-
muscular disability
such as motor neurone disease (MND), and may have no means of communicating with the external world. This
implies nearly 0.3 million people with severe movement disabili
ty in India alone.
In addition, out of 21.9 million
disabled people (
Sensus 2001), 6.1 million suffer from movement disability.
For providing independence and
assisted living to such people,
it is intended to

investigate

intelligent
systems for devising
an autonomous
wheelchair and robotic
arm

combination which allows the user to interact through a brain
-
computer interface (BCI)
as and when needed, while the combination performs activities of daily living autonomously through visual
interaction with the e
nvironment. In this proposal we identify Master’s and PhD thesis topics that would
investigate
the critical issues involved in the development of

a low cost brain
-
operated visually navigated smart wheel
-
chair
equipped with intelligent robotic arms.



The
first
area to be investigated

is an EEG
-
based
brain
-
computer interface (
BCI
)

operated asynchronously by the
user, based on mapping of imagined
-
tasks related mental states to respective computer commands. Existing BCI
systems are mainly driven synchronousl
y under computer control and lack sufficient accuracy and robustness.
Based on promising results obtained from the
UUM

partner’s recent BCI work, we believe, it is timely to investigate
a user driven asynchronous BCI. Quantum neural network (QNN), pioneere
d by the Indian partner (IITK), has the
unique ability to track time
-
varying probability
-
density
-
function associated with a stochastic signal. The project will
therefore investigate modelling of stochastic EEG signals related with imagined
-
tasks using a QN
N based feature
extraction technique for enhanced classification accuracy and robustness. Simultaneously an enhanced type
-
2
fuzzy logic based classifier design for improved uncertainty handling for the EEG signal will be investigated, which
will build on t
he

UUM
partner’s current promising work. The intended actions as interpreted by the BCI will be
validated through a specially designed virtual keyboard, before forwarding to activate the wheelchair or the
manipulator. Simultaneously a scene reader for read
ing visual map of the environment will be designed, using
which the disabled person can point to the desired target.


The visual tracking of
a wheelchair acting as
an
automated guided vehicle (AGV) re
quires very accurate target
position prediction (TPP). T
PP of a moving target needs a thorough investigation since a real
-
time algorithm that
would demand less computation while providing a greater accuracy is still elusive in the current literature. The
position estimation algorithm has to be succeeded by a mo
tion control algorithm that would ensure accurate target
following of the AGV. The project will investigate an optimal path following algorithm using intelligent techniques
which requires minimum examples during training. German partner (IAT) has recently
proposed an algorithm that
involves feedback structure in image
-
processing. Enhancement of this idea for TPP while comparing with methods
such as unscented Kalman filter as well as quantum neural network pioneered by Indian partner will be
investigated in
detail. Safe navigation of the smart wheel chair in a dynamically changing environment is still a
challenge. Traditionally the AI or algorithmic aspect of mobile robotics consists of five pillars: localization, mapping,
exploration, planning and collision
avoidance. A completely autonomous robot would want all these modules
working together in tandem seamlessly performing their tasks. An integrated navigation system consisting of these
modules will be developed. The Indian partner (IIITH) has developed som
e landmark algorithm
---

while UK
partner (UUM) have been successful using optical
-
flow based obstacle avoidance scheme.


Self
-
organised control architecture for redundant manipulator system interfaced with a visual
-
feedback system will
be designed in suc
h a way that it is robust, stable and smart. Model based control systems for redundant
manipulators without visual feedback has already been tested in real
-
time. Visual motor coordination based
manipulation schemes already exist for non
-
redundant systems i
n which the Indian Partner (IITK) has necessary
expertise in both basic and experimental aspects. However, a visual motor coordination of a redundant
manipulator system that simultaneously avoids obstacles while performing smart manipulative jobs is a chal
lenge
since the learning problem involves constrained optimization with multiple objective functions and a real
-
time
execution capability. The expertise of the IITK group will be combined with novel methodology of feedback
structure in image processing by
the German partner to come up with a practical and efficient solution to the
problem.


Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


15


In general, control algorithms for smart wheel chair and redundant manipulator are
very

challenging given the
back
-
drop of
time
-
varying
complex
system characteristics

re
quiring on
-
line adaptation and

constrained
optimization. One novel way to solve this problem is to apply evolutionary techniques. Ensemble of neural
networks can be trained to perform complex tasks through evolutionary approaches, so also evolvable hardwar
e
can be designed to develop fault
-
tolerant controls systems for the smart
-
wheel chair and redundant manipulator
combination. In this regard, expertise of the UK partner (UB) will be shared by other partners to develop suitable
evolvable control algorithm
and control hardware for assistive robotics.



In some cases of
disability due to old age, disease or injury
, people prefer
muscular function

augmentation through
exo
-
skeletal devises

so that they could perform the activities of daily living (ADL) using t
heir own hands rather than
an external robotic manipulator. Building on the expertise of the IITK partner, a PhD

topic is identified to
deal with
the development of an exo
-
skeletal robot
ic device

for supporting the hand of disabled persons.

Its

basic
mecha
nical structure
will
consist of a mechanical hand that has three to four degrees of freedom at the wrist, and
is attached to the arm below the elbow. The hand
will
consist of a two finger gripper that is force or position
controlled. The two fingers would

have
two degrees
-
of
-
freedom (
DOF
)

each and will be designed so that it can
hold objects of daily use. The wearer will communicate with the hand by an advanced man machine interface
consisting of sensors, actuators and a controller.


In a nutshell, the t
hesis topics identified in this proposal is based on mutual expertise available among partners.
14
senior Ph.D. students and three
master’s

students who are already working on the various aspects of assistive
robotics will be given further assistance of jo
int supervision and complementary laboratory facilities for
experimentation.
Five

new Ph.D. students will be funded through this Asia link program
to provide

joint supervision.
It is envisioned that successful completion of these thesis topics will go a lo
ng way in assisting the disabled
community while the outcome will further cater to various other generic needs in the field of assistive robotics and
industrial automation. FWBI will act as an associate
partner

for transfer of technology from this joint co
llaborative
effort to SMEs.

We further elaborate
below
each thesis topic with detail description of joint supervision and
exchange visits
.


Project Details:


1. A Brain
-
Computer Interface (BCI) for natural communication

Supervisor
s
:

Dr. G. Prasad, Prof. T.
M McGinnity, UUM
.


Collaborator
External

Supervisor:

Dr. L. Behera, IIT
K.

Senior Ph.D. student
s
:


P. Herman,
UUM
spends 4 months at II
TK, Ms. Indrani Kar,IITK

spends 6 months at
UUM
.

New Ph.D. Student:

1(UUM)
student
spends 12 months at IITK
.

For monito
ring and joint supervision, GP will visit IITK for four months and LB will visit UUM for two months.


This project will build on the expertise of the U
UM

partner to classify the brain imagery pattern
s

so as to generate
action commands as desired by the dis
able person. UUM’s work on BCI design based on modelling of movement
imagery (MI) related EEG time
-
series data using neural networks
[1]

and self
-
organising fuzzy neural networks
(SOFNNs)
[5]

and classifier design using interval type
-
2 fuzzy logic

[2]

will be revisited. Quantum Neural Network
(QNN)
approach
[18]
[19]
[20]
[21]
to time
-
series data modelling as pioneered by the Indian Partner (IITK) will be
further investigated. An algorithm using QNN to model MI
-
related EEG time
-
series data and obtain appropriate
feature patterns will be developed.

The performance of the QNN based predictive algorithm will be evaluated
against other approaches
[1]
[7]
. Simultaneously an enhanced type
-
2 fuzzy logic (T2FL) classifier will be devised
and its perfo
rmance will be evaluated against existing approaches
[1]
[4]
[6]
[7]

as well as QNN based algorithms.
Once a robust classifier is designed, the
n action commands have to be generated. An enhanced predictive virtual
keyboard (VK) extending the current work
[3]

will be designed

to create a system for validating the action
commands before their eventual execution
. A scene

reader will be designed that will facilitate reading the details
from a captured scene, so as to point to the desired target using BCI. Finally experiments with both able
-
bodied
and disabled subjects to evaluate the capability of QNN/T2FL
-
based BCI config
ured to operate asynchronously the
VK with scene reader and feedback for running the AGV and a manipulator will be conducted in UUM

and

IITK.


2.
Visual tracking of a smart wheel
-
chair operated as an automated guided vehicle (AGV)

Supervisor:

Dr L. Behera,

IIT, Kanpur

Collaborator

External Supervisors:

Prof A. Graeser, IAT, Bremen, Dr. G. Prasad, UUM, UK and Dr. C.V.
Jawahar, IIIT
H.

Senior Ph.D. students:

Swagat Kumar,

IITK spends 4 months at UUM,
A Browne,
UUM

stu
dent spends 4
months at IAT.

New Ph.D stu
dent:

1(IITK) spends 6 months at IAT and 6 months at UUM
.


Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


16

For joint supervision LB visits IAT and GP visits IAT for two months.


The IAT group has developed great expertise
[9]

in this area and collaborating partners will take
advantage of this
expertise through exchange visits. Visual tracking (VT) algorithms including target position prediction (TPP) and
motion control algorithms for AGVs
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]

will be critically reviewed. An efficient visual
tracking algorithm using unscen
ted Kalman Filter

[17]

will be developed and will be implemented on Patrolbot at
IITK where senior Ph.D. students from IAT and UUM will also be involved. Simultaneously this algorithm will be
tested on Friend II and Peopl
e
bot i
n IAT and UUM respectively. A TPP algorithm using QNN
[18]
[19]
[20]
[21]

will be
developed and its performance will be evaluated with existing

algorithms. The successful implementation of the
algorithm on Patrolbot at IITK will be shared with collaborating partners. Feedback structure in image processing
[22]
[23]

is a closed
-
loop digital

image processing which has been successfully implemented by IAT group for visual
tracking. This novel concept will be further investigated by the collaborative partners to improve the performance of
the visual tracking algorithms. A combined AGV motor dy
namics and camera model will be developed
[24]
[25]

for
efficient implementation of the motion control algorithm of the smart wheel chair. A novel
adaptive
motion control
algorithm using SOFNN
[8]

will
then
be developed. Real
-
time test and verification of BCI operated AGV on
Peoplebot and wheelchair at UUM, Patrolbot at IITK and Friend II at IAT will be carried out.


3. Development of an integrated navigation system for

a wheel chair like assistive robot

Supervisors:

Prof. Bipin Indurkhya, Dr. M. Krishna, IIITH

Collaborator

External Supervisors:

Prof. A. Graeser, IAT Bremen, Prof. PK Kalra, IITK India

Senior Ph.D. students:

AK Pandey
,
IITH, a senior Ph.D

student
spends
4 months at IAT
.

New Ph.D. Student:

1(IIITH)

student

spends 12 months at IAT
.

For monitoring and joint supervision M. Krishna and Bipin Indurkhy
visit IAT

for two months.


Traditionally the AI or algorithmic aspect of mobile robotics consists of five pill
ars: localization, mapping,
exploration, planning and collision avoidance. A completely autonomous robot would want all these modules
working together in tandem seamlessly performing their tasks. As a part of this project we would develop an
integrated nav
igation system consisting of these modules. Based on a particular task at hand currently some or all
of these modules will come into play. The research component of this effort include
s

actions taken by the robot
based on visual and other sensory cues from

patients, localization as well as obstacle avoidance in dynamic
worlds especially if the wheel chair is moving on crowded floors or corridors traversed by other patients and
visitors. In this regard the experience of IAT team
[26]
[27]
[28]

is complementary to the experiences of IIITH team
[29]
[30]
[31]
[32]
[33]
[34]
[35]

and IITK team
[36]
[37]
[38]
[39]
. Thus a co
llaborative research in this regard will lead to
development of efficient sensor based architecture and algorithms for a mobile robot for safe navigation. Although
IIITH group will take benefit of the infrastructure facility available at IAT and IITK, it w
ould like to procure a wheel
chariot from Mobile Robots Inc
that will cost USD 12,824.



4. Intelligent control of a redundant manipulator system using visual feedback

Supervisors:

Dr. Laxmidhar Behera, Dr. Ashish Dutta, IITK, India

Collaborator

External S
upervisors:

Prof. Axel Graeser, IAT,
and
Dr. G. Prasad, UUM
.

Senior Ph.D. students:

A
K

Ray
and S

Kumar from IITK

spend 6 months at UUM
and
IAT

respectively
, 1 IAT
student visits IITK.

Prof Graeser visits IITK for two months.


Current research projects in r
edundant manipulators are addressing novel control techniques for redundancy
resolution so that the manipulator can perform extra tasks in addition to the basic task of end
-
effector trajectory
control, thus increasing the versatility of the manipulators in

both assistive and industrial robotics. Both IAT and
IITK groups have been working on manipulator control using visual feedback and both groups have 7DOF
redundant manipulator systems. Manipulator control using visual feedback (VF)
[40]
[41]
[42]
[43]

schemes will be
critically analyzed as applicable to redundant manipulators. The redundancy resolution schemes such as task
-
based configuration c
ontrol
[44]
[45]

will be investigated.

Adaptive control schemes
[46]
[47]

as applicable to a
redundant manipulator will be derived. A visual
feedback control scheme using fuzzy TS
models
[48]
[49]

will be
developed. These algorithms will be implemented on a 7DOF power
-
cube robot both at IITK and IAT. UUM team
will be involved
in

developin
g an embedded control system using FPGA for hardware implementation of the neural
control algorithms. Self
-
organized map (SOM) based visual motor coordination schemes KSOM

[50]
[51]
[52]
,

PSOM
[53]
,

and QSOM
[54]

as extended to redundant manipulator system where redundancy is resolved through
constrained optimization will be investigated for developing an efficient alg
orithm. The feedback structure in image
processing

[22]
[23]

will be introduced to make visual
-
motor coordination scheme more robust and fast. These
algorithms will be implemented in real
-
time on a 7
DOF power cube robot both at IITK and IAT. The performance
will be compared with Visual motor coordination using SOFNN algorithm
[8]

and type
-
2 fuzzy system based
learning algorithm
[2]
.





Asia
-
Link Programme: Applicati on Form, 5
th

Call for Proposals,
2006

Page


17

5. Neur
al Network Ensemble Learning for Robotic Control

Supervisors:

Prof. Xin Yao, UB, UK
.

Collaborator

External Supervisors:


Dr. L. Behera, IITK, and Dr. AK Mishra, IOE Nepal
.

Senior Ph.D. student
s
:
Mr
Awhan Patnaik, IITK will spend 6

months in UB, 1(UB

studen
t
) will spend 4 months in
IITK.

Senior Master’s student
:

Ramesh Chaudhary,
IOE, Nepal
, a senior M.Sc.
student

will spend 6 months

at
UB

during his/her thesis work.

For joint supervision LB visits UB for two m
onths and XY visits IITK for two

month
s
.


Many r
eal
-
world problems are too large and too complex for a single neural network (NN) to solve alone. An NN
ensemble consisting of several individual NNs has been shown to be able to improve NN's generalization
performance
[55]
[56]
[57]
. There have been much work in training NN ensembles
[55]
[56]
[57]
, in mixtures of
experts, and in various b
oosting and bagging methods. However, all these methods are used to adapt weights in
an ensemble. The structure of the ensemble, e.g., the number of NNs in the ensemble, and the structure of
individual NNs, e.g. the number of hidden nodes, are all designed

manually and fixed during the training process.
While manual design of NNs and ensembles might be appropriate for problems where rich prior knowledge and an
experienced NN expert exist, it often involves a tedious trial
-
and
-
error process for many real
-
wor
ld problems
because rich prior knowledge and experienced human experts are hard to get in practice. This project
will
stud
y

constructive ensemble learning algorithms, where the ensemble structure, NN structure and NN weights are
trained automatically. In p
ractice, a robot will often encounter new environments that it has never seen before, it is
essential for the robot to learn incrementally and continuously. NN ensembles are well
-
suited to continuous
incremental learning due to its population
-
based structu
re. This project will investigate this aspect. So
there is a

need

for

two

graduate

students
in this project
. In particular, the objectives of this project include:



to implement and compare experimentally selected existing ensemble training algorithms on di
fferent
robotic tasks;



to analyse the above experimental results and study in detail the impact of different ensemble and NN
structures on the results;



to propose and study constructive ensemble learning algorithms, which can learn the structures as well
a
s weights;



to develop continuous incremental learning algorithms based on neural network ensembles;



to adapt the new learning algorithm
[58]

developed by the IITK partner for ensemble networks;



to evaluate the proposed algorith
ms using real robots both at UB and IITK.


6. Evolutionary Design of Fault
-
Tolerant Control Systems

Supervisor
:

Prof. Xin Yao, UB

Collaborator

External Supervisors:



Dr. Laxmidhar Behera, IITK, Dr. AK Mishra, IOE Nepal.

Senior PhD students
:

Mr. Ashutosh
Dwivedi, IIT
K

will spend 6

months in UB, 1 (UB

student
) will spend 4 months
at IOE, Nepal.

Senior Master’s student:

Suresh Gautam,
IOE, Nepal
, a senior M.Sc.
student

will
spend 6 months in UB for his

thesis work.

For joint supervision, AKM visits UB for t
wo months.


Evolvable hardware (EHW)
[59]
[60]

refers to one particular type of hardware whose architecture/structure and
functions change dynamically and autonomously in order to improve its perform
ance in performing certain tasks.
The emergence of this new field in recent years has been influenced profoundly by the progresses in
reconfigurable hardware and evolutionary computation. Traditional hardware is notorious for its inflexibility. It is
impos
sible to change the hardware structure and its functions once it is made. However, most real world problems
are not fixed. They change with time. In order to deal with these problems efficiently and effectively, different
hardware structures are necessary.

EHW provides an ideal approach to make hardware "soft" by adapting the
hardware structure to a problem dynamically.


The primary aim of this project is to study the evolutionary design of fault
-
tolerant control system using our idea of
negative correlatio
n and artificial speciation
[61]
[62]