noc19-mg15 - Lecture 51: GERT

IIT Kanpur July 201
20 followers
1 Video View·Dec 3, 2023
Graphical Evaluation Review Technique (GERT), Q-GERT, Examples of GERT
[Music]
[Music]
good morning everybody good afternoon
and good evening to all of you welcome
back to this da DM 2 lecture series on
NPTEL MOOC and as you know this is the
data necessary session making to course
and we are in the 11th week and as you
are aware which I do repeat before the
starting of any class for this didm 2
this is a 12 week course that means we
have the learn week and the 12th week to
complete before you have of this course
total contact hours is 32 total number
of lectures is 60 because each lecture
is for half an hour and after each week
we have assignments after 5:00 and and
as you know in each week we have 5
lectures after which as I already
mentioned just few seconds back
we have assignments so you have already
completed 10 assignments so with the end
of the 11th week you will have the 11th
assignment and my good name is Ragunan
and Sengupta from miami department and
iit kanpur so if you remember we were
discussing the Gert and the Q Gert
process general evaluation review
technique and queueing concept being
utilized in huge dirt or Gert now in in
this Gert method and Q Gert method you
have different type of logic gateways to
be utilized and or nor and so on and so
forth and based on that you have
different ways of how you implement that
so we did discuss at the last few
minutes for the ending of the tenth week
and based on that we will proceed into
the further discussion so what we will
do is that in the and the Gert process
we have some simple steps how you
basically will implement it so we will
discuss that first and then go into the
analysis so you will convert a
qualitative description of the overall
system on the problem to a model in a
network frame network model where the
network models as I mentioned that
you consider the pot and CPM they would
not be any looping back but in Gert and
QE Gert you'll basically have the
looping back concept there so based on
this assumption you will basically what
are the informations are given like
process one precedes process two by two
minutes or say for example process two
can only be started when process three
has also started or say for example
after process for finished you will be
say recheck it and it comes back to
process two so all these quality
statements which I mentioned for the
problem would be incorporated in the
dirt or Gert concepts so you'll collect
the necessary data to describe the
branches of the network so if you
remember the branches of the network
would basically have a probability and
the time period so time period
probability would be such that it will
consider that if the process follows any
any path it will have a probability and
this probability would be given by a
certain distribution and the time would
also be considerated in order to find
out the cumulative time from the process
from the source to the sink considering
the source is on the left hand side and
singing on the right hand side you will
obtain also an equivalent one branch
function between the two nodes of the
network one branch function means these
and/or concepts you will combine them
probabilistic and deterministic and and
all this combination will be done based
on this you'll basically formulate the
network on the whole process a project
network into the Gert logical network
such that it will be easy for us to
solve in the four steps we will convert
the equivalent functional form into the
following to performance measures of the
the network so if you remember I did
mention time and again that probability
and tiny important and will basically
utilize this concept of probability and
time to be seen find out the overall
metric based on which you will analyze
the the overall network so one would
with the probability that the specific
node is realized and obviously you will
have the time period for the realization
and another would be the
moment-generating function on the time
associated with an equivalent network
says that based on the moment-generating
function you can find out the average
time the variances the expected value so
on and so forth such that it will give
us a whole lot of information about the
network from the time point of view but
this time taken for a for a for an
information flow order for project flow
from eyuth to the jet phone if you
moving from the left to the right and
obviously definitely took and me from
the right to the left also considering
there's a looping so this time period
would would help us in order to find out
the overall average time taken to finish
the project based on the concept of Gert
and Q dot being utilized
you will make inference concerning the
system under study obtain in for above
so for above was basically you will have
the moment generating function based on
the time fact and you basically have the
the probability that a specific network
or a path would be traverse so one is
probability its distribution function
one is the time which will be found out
using the moment generating function so
consider this a very simplistic version
of the Gert network so you basically
have the you you are doing some process
whatever the process is so consider I
start from the left most node and let me
mean given number one two three four
which for this diagram is 1/2 the foot
but consumer in the overall description
on the problem the unmentioned known
number twenty two twenty three twenty
four twenty five obviously they can be
other nodes also going out from twenty
two or coming to an end to similarly
going out from twenty three coming in to
twenty three and so on and so forth now
consider that i will only consider the
networks and the diagram in the arrows
in which the network is connected so you
will basically define the problem so the
definition of the problem what you want
to do say for example at twenty second
stage you obtain the raw materials and
based on the processing work which you
do on a lathe machine on a
machine on up shaper pain a CNC machine
whatever it is you reach the 23rd stage
which is state number two but it may be
possible that the overall working on the
raw material based on the fact that you
have to do some grinding drilling
whatever was not perfect so obviously
the metric machine the work can come
back to stage one to be tested to such
that you can find out the dimensions
based on the fact that whether
dimensions right again it need to be
partly processed or processed
now once the twenty third stage is
complete with looping being they are
considered so it may problem the overall
movement can be in this direction so
there's need no reprocessing it goes
back so there's a probability of how
many such sets of in a particular unit
of time what is the probability that it
will proceed straight forward without
any rupee and in the case if there is
grouping we will find out that what is
the probability based one is the loop is
formed and to what stage the loops will
continue
so once 23rd state stage is complete we
go in to say for example the next step
is basically the research activity
research activity gives an output on the
stage 24 so it will basically find out
research means that consider here build
up you have been in developed a molecule
for a drug or see for example found out
a good chemical composition of some
fertilizer which can be utilized in the
fields you will basically evaluate the
overall output which you have obtained
so here it will come to evaluate the
research research means the overall work
which you have done based on the output
which you will get you will basically
from reach the stage twenty five now it
may be possible that the overall total
quality measurements you are doing the
soil quality measurement you are doing
considering that you using a fertilizer
the productivity of the soil the
irrigation level which should be
utilized in this soil everything is
obtained at the end of the
twenty-fifths a stage so consider if it
is perfect so we will proceed further on
depending on whatever the next steps are
considering it is not perfect again you
will basically be go into the research
stage try to change few of the
parameters of the input maybe so I'll
nitrogen content maybe water moisture
which is there in the soil or the water
content which is there in the soil or
the compost quantities you want to add
into the soil so all these things would
be reevaluated and some things would be
changed and based on that again we'll
complete the research activity and again
evaluate the results so this loop will
continue till we are satisfied to what
level of of competence our overall
output we want from the research
research means the overall work which
you want to do so it is important to
note however some of the inherent
disadvantages which are there when you
use Gert process obviously it is an
advantage with respect to pod CBM but
how be and it has some disadvantage or
else if there was no design frontages
then Gert would definitely would have
been used in a very big way or QED Gert
or cue judge should have been utilized
in a very big way in case of pert and
CPM so what are the disadvantages and
what are the so-called bottlenecks first
Gert typically employs activity on arc
format which you have previously noted
are not at all common with modern
project management techniques because
the use of dummy variables then comes
into the play based on the fact that you
are using the activity and our concept
now you are trying to utilize the
activity in node concept then obviously
trying to draw the god network is not
possible so thus with so few exceptions
god is not supported okay and moreover
which I should have mentioned before the
beginning of the other Gert concept Gert
somehow has lack the overall input which
should have come from the industry based
on which different type of software's
and tools could have been developed but
unfortunately it didn't happen so one is
basically the activity and our concept
why it is used and
why generally activity on node would be
conceptually much better in order to
analyze pert and CPM and obviously
activity on node should also have been
utilized for good but as you know that
God is only based on activity and arc
point one point number two that the
considering is it's basically a long
drawn process where you find out the
time where you find other probability
and and the probability of taking any
one path so if and if they're very
complicated networks obviously you need
good software to basically solve that
those type of problems which vary use
the Gert and the Qbert processes but
unfortunately the development of Gert
software has not taken place as it
should have been to continue this
discussion further so GERD networks can
become extremely cumbersome complicated
depending upon the size of the project
then the number of activities expected
feedback loops how many feedback loops
are there the probabilistic condition
that mint must be modeled there's
probability value the time period value
all these things are brought into the
picture and within the whole set of
project mass nodes and activities so it
becomes very difficult to model them in
the true sense so thus all the
purporting to offer a more accessible
and visual treatment of project networks
God can actually quickly become unwieldy
and obfuscate the visualization of the
projects network activity so considering
that it has loops it's as probability so
annals forth so even if theoretically it
is best to have the god network to
basically give us the best possible
results for interpreted type of
complicated project but when you see it
in the practical sense it becomes too
cumbersome in order to that to be
implemented so consider a very simple
example so there is a space mission
which consists of two vehicles to launch
vehicles both vehicles must be
successfully launched such that the
space mission is successful so consider
there are two booster rockets and both
of them are needed because one booster
rocket would help the overall space
spacecraft to go to a certain distance
for example 100 kilometers and then in
the next 100 kilometers considering that
it has to traverse a total distance of
200 kilometers that second booster
rocket should be working so that means
both of them should be operational for
this space mission which we are going to
consider as in the example so consider
those those booster rockets as vehicle 1
and vehicle 2 so in his vehicle 1 and
vehicle 2 so we'll consider that the
working principles would be so let me
take out I think it would be best if I
copy so give me 2 one minute I'll copy
that logical networks I'll copy it here
I have this so it would be easier for
all of us to because I thought I will
just go at one place then I change the
the speed depending on that how the
problems caused so these are the levels
of information which I would have
and the one one of these so we will copy
these also so for better understanding
of the overall problem
I just tried that we are all in sync
okay so we are all set if you remember
so these are the simple examples first
exclusive I am just repeating what we
are done
exclusive are the concept logical
network or gateways inclusive are the
third one is and and the characteristics
are all given on the right then you have
the deterministic and the probabilistic
so and the combinations are given so let
us go one by one I'll come back to this
slides quite often as we proceed so if
you consider vehicle one and vehicle to
the symbols are diamond so what is the
diamond I know that all of you are aware
of what is the diamond is the
probabilistic one but still for sake of
all of us to understand this diamond
means the probabilistic network that
means exactly one branch emanating from
the node is taken if the node is
realized so it is only one either
success or not success it cannot do both
so when when the vehicle one is launched
it is successful that means it will take
the the spacecraft from zero to hundred
kilometers and if wait who is also
successful it will take it from 101 to
200 kilometers so
vehicle one successfully launched so
this is for one let me use the blue
colors and red colors
[Music]
so I'm rooting for vehicle one this is
for vehicle to now I use the red color
unsuccessful vehicle one vehicle too so
now let us see how it is
so vehicle one is successful and vehicle
deuce is successful the mission success
is there vehicle one unsuccessful
vehicle two is on successful the machine
is a failure because these are non
successful launches and consider that if
you have mentioned that any one of the
the booster rocket working it would be
it will make the project successful
which means that in this case we will
basically have the concept that so in in
these combinations we are not going to
take successes and unsuccess that means
we are not going to take the concept
that one is a success and two is an
unsuccessful one that is blue one and
red two would not mean that the whole
project is a success or if I consider
this concept of red one and blue two
also doesn't mean that it is a
successful launch
so only combination would be yes for
success for one yes for success for to
any other combinations are unsuccessful
so we will basically proceed on this
concept now consider the space mission
consists of two vehicles again but it
now mentions that at least one on the
vehicle must be successfully launched
that means one is success other may be
an unsuccessful one so they will
basically have blue one red two or
bluray em red 1 blue 2 these are
possible it is a success and also
already we have that blue one and blue
to would do a success so out of the four
combinations three are yes one is known
so let us see this common concept so it
will become a more more
so-called involved diagram so vehicle
one I'll use the I use the same concept
of the colors blue for success and red
for unsuccessful one and then highlight
it accordingly
so vacant one and launches success now
if it is a success we have a vehicle to
also success so both of them are success
so let me use blue color
so this is 1 this is 2 this is 1 this is
2 so let us go from the extreme where
both us unsex Amtrak or ting from the
extreme case as that is easier for me to
proceed and explain
so this is unsuccessful this is an SEC
soon so you an successful launch so it
would be a sickly beer man ability
failure is there and the whole mission
is a failure so if it is a mission isn't
failure so is an unsuccessful launch so
he'll basically go back to the stage
that they are not helping the overall
okay not only just this I this diagram
would have been much better referring
basically combine it accordingly so this
is success
so yes so this is also success this is
so consider that the corn come this
combinations was such success and
success a combined in this case both
vehicles are successfully launched if
that is the case then the they can be
two states of failure later on one is
man awaiting failure for vehicle one and
this man ability failure for vehicle to
so if I combine both of them so even if
they are success later on man ability
failure would lead to the mission
failure as marked here as I'm
highlighting and in case man ability
failure is yes for one no for two no for
one yes for two it will be a success for
the whole project and obviously if the
combinations are success man ability are
yes for both of them it is an yes
success for the overall project now what
one of the nine cities of the Gert is
its usefulness at many levels within a
problem area because it is able to
delineate all different combinations of
the output diagram which are possible so
you will basically had a receipt of this
consider that now you have the subset
subsystems so you for a successful one
on an unsuccessful one you will have
different stages if you combine them and
you are trying to basically formulate
for this vehicle launch it can be say
for example you are doing a big
manufacturing work for a bridge or for a
boiler or or for an aircraft order for a
car which is there are lot of stages and
you can basically consider that all of
the works at the sub stations are sub
assemblies so it considers them as
subsystems so based on the fact that all
the subsystems are worked on you will
basically have the overall combination
combined and which you would leave to
the assembly checkout so here you will
basically have in this case this is the
pre processing work which you have been
going on before the vehicle is launched
and so launching would basically be a
success
now if it is a success there would be
combinations based on this which you'd
formulate the problem so these are the
combinations so if I consider the
combinations would be your can basically
combined in the input of the output so
if it is an exclusive or and a
deterministic output you have this and
parachute shaped the lying down
parachute which is a triangular part in
a hemisphere so that will give you an
exclusive or plus deterministic one and
diamond one would be the exclusive or on
the probabilistic one with a base so 10
which is the exclusive or and the
inclusive or would definitely not not
have the base so in that case you will
have the inclusive or the trying and the
so-called triangle so this is the
triangle which you consider the VV shape
and again you have the deterministic in
the probabilistic so-called output and
for the and deterministic and the and
probabilistic one and you have the
circular part with the probabilistic and
the deterministic concept so in the
dynamic concept so no change in the and
the overall structure occurs and the
produced economy over triangular or we
turned anti-clockwise 90 degrees
see of the check rod of the system
assembly check out the terminal count on
the successful launch and that will
leave to the orbital correction achieved
or the successful launch has been done
such that you can see the overall
project has been successful depending on
the combinations which you have for a
yes yes the number of diagrams are much
low for a yes yes been successful and
yes no no yes
also being successful the total
combinations would definitely increase
further continuing the discussion so for
the and note it plays a predominant load
no role in activities up to and
including a terminal countdown so this
is due to the fact that all activities
must be performed before liftoff
obviously you have to finish the
processing the part the checking the
electronic system if any repair has to
be done everything has been checked that
whether the cryogenic fuel has been
filled properly with the booster occurs
our protein properly everything has to
be checked and thorough before the
launch
after the terminal ground on either
possibilities as possible that means it
is a success an unsuccessful one but the
success in in unsuccessful ones are
combined for each and every stage which
you have already considered successful
node is an exclusive or node due to the
reason that successful launch can happen
in two mutually exclusive ways which are
proper operation during boost phase as
that is a success later on and
unsuccessful orbit after boost phase
with orbit correction being achieved so
there has been some answers will later
stage at the boot state but you are able
to correct correct it so obviously
that's the the rocket or the satellite
whatever was being planned to put in the
orbit has been put put there as required
the dotted lines represent the process
where it is where it signifies
activities that do not contribute the
successful launch so if you consider the
dotted lines that is the flight phase
failure and the boost phase failure
technically would mean that any search
field
here would signify that they do not give
any impact on to have any impact on the
successful launch
so with this I'll end the first lecture
for the 11th week and continue to
discuss more about and got in the
subsequent lectures for the 11th week
have a nice day and thank you very much
[Music]
[Music]
[Music]
