The use of virtual learning environments to aid teaching of heat transfer and Artificial Neural Network modelling in Bioprocess Engineering

sciencediscussionAI and Robotics

Oct 20, 2013 (3 years and 9 months ago)

92 views

The use of virtual learning environments to aid teaching of heat transfer and
Artificial Neural Network modelling in Bioprocess Engineering


Brian Freeland, Sumsun Naher, Greg Foley and Dermot Brabazon

School of Mechanical and manufacturing Engineering &

S
chool of Biotechnology, Dublin City University, Ireland.

Email: brian.freeland
2
@
mail.
dcu.ie


Paper Type: Practitioner


Keywords:

Virtual instruments, engineering laboratory practicals, labview, artificial
neural networks
.


Abstract

High quality laboratory
practical’s for undergraduate students require extensive
demonstrator resources to implement on a week to week basis. This can be difficult to
maintain over the course of a semester. This paper presents work in which an
alternative technique to the traditi
onal approach was developed. A virtual learning
environment was employed to implement the entire lab, reducing demonstrator
involvement and ensuring a constant quality of explanation and demonstration of
concepts was provided to each group of students. Thi
s new laboratory practical was
developed to demonstrate heat transfer and artificial neural network modelling to bio
process engineering students. The use of this virtual environment approach allowed
the inclusion of the application of industrial process m
onitoring and data acquisition
in the teaching process. A full user interface was constructed that the students
navigate through; this interface guided students through the entire lab, teaching heat
transfer and artificial neural network concepts along wit
h the general data collection
and machine setup procedures. The virtual environment was constructed, with
students learning styles in mind, providing information in both a global and sequential
manner. This approach was seen to useful in terms of enabling
student engagement.


National Instruments LabView software was selected as the programming
environment as it allowed easy integration with data acquisition and analysis with
high quality graphical user interfaces. This work shows how it is possible to att
ain
low cost multifunctional data acquisition and device control to develop educational
resources. Safety features were inbuilt into the program to ensure students could not
damage the heat transfer rig, or injure themselves from the rig overheating, or
ma
lfunctioning. Trajan was used to simulate the artificial neural network models. The
lab practical has been run quite efficiently over the course of the semester and it was
seen that the use of a virtual learning environment engages the students more than
t
raditional techniques, reduced the demonstrator workload, and provided increased
student interactivity with industrial equipment. Evidence of this is presented in
student and staff surveys as well as student learning outcome results.








1 Introduction


Interest in teaching techniques for engineering has increased over the years.

There is
c
onstant work

being
carried out

in
researching how best to present classical theories
to

students
. There is an aim to try and match their learning styles w
ith teaching
styles
and

questions have been posed of how best

universities

can

equip students for
industry. More and more practical based teaching

method are

employed either using
virtual laboratories
[2]

or using full laboratory practical’s
[3]
.
In Biote
chnology and
the processing industries the requirement for graduates to be able to problem solve
and work with processing equipment is critical. Therefore
, hands
-
on experience on
processing equipment during
their undergrad
uate studies is a requirement.
To
this end
the BSc degree program in Biotechnology in Dublin City University offers a
considerable amount of its teaching
time to laboratory practicals. H
owever due to
resource restrictions it is not practical to offer undergraduate students full access to
t
he schools bioreactors.
It was decided to develop a new
third year engineering
laboratory to teach heat transfer

and artificial neural networks while demonstrating
the use
of
bio processing equipment

to biotechnology students
.
This new laboratory
utilises
a virtual learning environment that can integrate with low cost data acquisition
and control equipment to offer a hands on fully interactive computer program and lab
equipment driven practical.


The aim of teaching through the use of a virtual environment
was to provide a visual
based
lab that

would reflect the systems in use in industry while also offering a
consistent and quality teaching platform

that would guide students

sequentially
through the lab using
a custom built user interface.

It was expected t
hat using a
graphical user interface should reduce the require demonstrator involvement
, while
providing experience with industrial specification user interfaces
.


1.1 Student assessment

In order to assess the learning styles of a significant sample of the

Biotechnology
student population, the 2
nd

year students were given Fielders learning styles
questionnaire

[4]
. The outcome of
this questionnaire
would be used to
evaluate

the

type of teaching
that
best suits biotechnology stude
nts. The survey was conducted by

86% of the total class. The survey

results showed that 80% of student’s surveyed are
strongly visual learners,
i.e.

they learn better by reading charts and diagrams to
explain concepts rather than presented in a verbal manner. Most of students are
balanced between sequential and global, me
aning that they like to be presented the
full picture of concepts at the start of a lesson and then, introduced to the concept
piece by piece. 75% of the class are active learners, with the rest balanced and a small
percentage reflective. Its surprising th
at some students are reflective as studies have
show that the engineering and science students are heavily active based learners
rather than reflective.


Figure
1
: BT2 Visual/Verbal learning style



Figure
2
: BT2

Sequential/Global learning style



Figure
3
: BT2 Active/Reflective learning style



Figure
4
: BT2 Sensing/Intuitive learning style


From the results of the fielder surveys it can be determined
that laboratory practic
al’s
and teaching using problem

based learning, are techniques that would best suit,
biotechnology students. These results also ties in well with general engineering
student
s
, who

have been seen to be

more active and visual learners

[5]
.
As a result of
these

surveys
,

the
new laboratory
practical

was designed to provide information in a

sequential manner, u
sing a lot of visual aids to explain

the main

concepts. Students
should have a lot of input into the practical, and have a sense of ownership over the
practical and their work, this was seen to be a beneficial motivating factor with the
design and build he
at exchanger project in previous years
[6]
.


2
.

Implementation


Labview was chosen for the user interface, data acquisition and machine control
programming environment because of its straight forwa
rd integration with data
acquisition equipment and its easy to use graphical user interface. It utilises a
graphical programming language using icons rather than calling text. It operates in
parallel rather than sequentially this offers operational speed a
dvantages as functions
can execute as soon all their terminals have received information. Programs
developed in Labview are called vi’s or “virtual instruments”.


2.1
Bioreactor operation

The completed rig seen in
Figure
5

is composed of a stainless steel jacketed vessel
(1), agitator (2), coolant tank (3), coolant pump (4), thermocouples (5a), coolant flow
meter (5b), tachometer (5c) and ancillary pipe fittings.


Figure
5
: Completed Rig


The pump, valves and agitation speed can be set manually by the students, from the
instruction given in the laboratory interface, for safety reasons the steam line
s

are
automatic
ally controlled by the program. A low cost national instruments
multifunctio
n data acquisition card (USB 6008) interfaces with the sensors

and
controls the steam valves
. A low cost RS232 Pico Technology
thermocouple data
logger relays

temperature readings to the computer. Both devices integrated
seamlessly into the labview progra
mming environment and provided reliable
information and control of the equipment.

All the
process variables; coolant flow
rate, agit
ation speed, process fluid bulk
temperature and coolant jacket input
temperature
were

monitored and recorded automatically.

These variables along with
calculated values are

automatically

stored into a .txt file for analysis.

While designing
the bioreactor and labview program there was
a lot of thought given to

allowing
students “hands on” access to as much of the equipment as

possible
. This was made
possible by writing “fail safe
” systems

into the labview program.


2.2

Lab structure

The laboratory practical
is

run
over

a full day every week in semester 2.

One group of
four
students

participate each week. The Labview

user inte
rface drives the lab, almost
all the required information is provided to students via the interface, a

small
supplementary manual
to
demonst
rate the use of the
Trajan

neural network simulator
software

is provided
.
The basic structure of the lab is as follo
ws
;

1.

From the main welcome screen, students

can

study the relevant theory.

2.

The rig operation is guided through in a sequential manner.

3.

Heat transfer experiments are performed for varying process parameters
(selected by the students).

4.

The experimental resul
ts can be reviewed via an “Analys
is screen”, students
perform heat transfer calculations
.

5.

The experimental data
gained by the students is automatically added to
previous groups collected data. This
is

then

exported to Trajan artificial neural
network simul
ation software.

6.

An appropriate artificial neural network model is generated

by the students
,
based on the theory provided

to them

earlier.

7.

Instead of a laboratory report the students filled in a workbook. This
was used

along with an online version of the
virtual environment

as study for a
laboratory exam given at the end of semester.


2.3
Virtual environment implementation

The lab is started from the welcome screen in the labview interface; all options can be
accessed from there, except the ANN model devel
opment, which uses Trajan neural
network simulator software.

The students are offered all relevant theory, to aid the
global learners and demonstrated the rigs operation in a sequential manner. Screen
shot of various parts of the program can be seen in the

following figures. An
emphasise was placed on making the experimentation screens as close to industrial
specification machine status and control displays as possible as seen in figures 9 and
10. This was in keeping with the projects aims to offer students

experience with
industrial specification industrial equipment.

As a whole the Labview programming
environment proved a straight forward platform to produce a high quality interactive
user interface.

The

finished

executable program and code can be download
ed from the
web
[9]
.





Figure
6
: Welcome Screen



Figure
7
: Theory



Figure
8
:
Rig operation and current

state



Figure
9
: Data Analysis


3.
Results


The
exams

are currently being graded and will be displayed in the presentation.
Using
the

virtual
learning
environment dramatically reduced the
demonstrator’s

workload,
as students w
ere guided through every aspect of the practical.
All student groups were
able to successfully navigate the program and set up the
bioreactor

using

the step by
step instructions. Students were ab
le to grasp a complicated subject matter such as
Artificial n
eural networks

within a single day’s demonstration, and they seemed
genuinely enthusiastic about using a different, more “Hi
-
tec” learning technique.
Based on the success of this project, it has been decided to develop more virtual
learning environment bas
ed laboratory practicals.


References

1.

CEE
. [cited 17/08/08]; Available from:
http://cee.che.ufl.edu/
.

2.

R.Greene, A.I., P.Smyth, D.Brabazon, and E.McLoughlin,
Developmenty of
interactive an
d remote learning instruments for engineering education
, in
International Symposium for engineering education
. 2007: DCU.

3.

C.O'Sullivan.
Teaching Heat Transfer to Engineering Students
-

a course of
computer
-
based hands
-
on activities
. in
International Con
ference on
Engineering Education
. 2007. Portugal.

4.

R.M. Felder, B.A.S.

Index of Learning Styles Questionnaire
Volume
,

5.

R.M.Felder, J.S.,
Applications, Reliability, and Validity of the Index of
Learning Styles.

Intl. Journal of Engineering Education,
2005(21(1)): p. 103
-
112.

6.

B. Freeland, J.T., G. Foley
A biotechnology student project and competition to
design and build a simple heat exchanger
, in
International Symposium for
Engineering Education
. 2007: Dublin City University.

7.

Mohan, P., A. Nichol
as Emery, and T. Al
-
Hassan,
Review heat transfer to
Newtonian fluids in mechanically agitated vessels.

Experimental Thermal and
Fluid Science, 1992.
5
(6): p. 861
-
883.

8.

Triveni, B., B. Vishwanadham, and S. Venkateshwar,
Studies on heat transfer
to Newtoni
an and non
-
Newtonian fluids in agitated vessel.

Heat and Mass
Transfer, 2008.

9.


Available from:
http://student.dcu.ie/~freelab2
.