YT YouTube2Text

YouTube2Text → Transcripts → Lecture 28 : Supply Chain Models and Modelling Systems

Lecture 28 : Supply Chain Models and Modelling Systems — Transcript

by NPTEL IIT Kharagpur · 8,300 words · 1,248 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:01[Music]
0:22[Music]
0:25hello dear friends welcome back to nptl
0:27online course on Logistics and supply
0:29chain management I'm Dr vikas takur
0:32assistant professor Department of
0:33humanities and social sciences IIT kakur
0:36so now today we'll talk about uh supply
0:40chain models and modeling system we'll
0:42see what are the different options we
0:44have when we we will Define the overall
0:46layout for the supply chain and then
0:48we'll you know model all those scenarios
0:51and we'll see what are the results of
0:53the simulation if we go with one
0:55particular type of model and then how we
0:57can adjust all those parameters so that
1:00we can come up with the desired output
1:02whatever is the requirement by the
1:04customer either customer is maybe more
1:07uh cost focused or maybe a customer
1:10wants responsive kind of delivery so
1:13those parameters we'll see right so this
1:16modeling system will help us to you know
1:19model all those requirements of the
1:20customers we'll also uh consider the
1:24constants from the supply chain side
1:26right what are the capabilities of our
1:28supply chain we'll also so consider that
1:30all the stakeholders will take into
1:32picture and then we'll see this is what
1:34maximum we can do using that particular
1:37approach right so uh This Is How we'll
1:40proceed for this we'll just introduce
1:41about the supply chain modeling what are
1:44different components we'll consider when
1:46we'll go for the modeling process and
1:48then we'll talk about some supply chain
1:50models and uh obviously uh we'll talk
1:53about uh the future in in this modeling
1:56system right and we'll end up with some
1:58cases right so supply chain modeling
2:01designing and analyzing a supply chain
2:03to identify potential improvements right
2:06so that means it's a kind of design of
2:09experiment approach where we will see
2:12you know uh we'll try different types of
2:15inputs right and in different quantities
2:18right this is how you uh initially how
2:20you are learning the cooking process
2:22same way right so sometimes you are
2:25adding some ingredient maybe in more
2:27quantity so then you will suddenly after
2:29analyzing the output tasting that food
2:32you'll realize okay I need to reduce the
2:34quantity of that particular ingredient
2:36or maybe I need to increase that so in
2:38that way we can simulate different
2:40scenarios just to we'll find out what
2:43are the outputs whether the outputs are
2:45efficient the way we want that uh system
2:48should behave right so we can compare
2:51with the whatever uh you know
2:53pre-assumptions we have about that model
2:56so either we can talk about the offline
2:58model static or or we can go for online
3:01model this is known as offline model
3:04this is known as online model offline
3:07means we'll take care of the present
3:10scenario and we'll model the factors
3:12likewise right we'll consider all the
3:14situations uh whatever is right now
3:16we'll record the data on that and on
3:18that basis we'll run our model and we'll
3:20say this is going to be the output if
3:22we'll go with this approach right
3:24Dynamic model is a kind of model we
3:27which continuously feeds data inside the
3:29system and then is there any changes in
3:32the environment that also we can uh
3:34reflect on that so it's a kind of we
3:36know we call it online system as well
3:39and then we can go for the Monte Carlo
3:41simulation models where we can talk
3:43about different scenarios and those
3:45scenarios if will go by one way what
3:47will be the output and if will go by
3:50other way what will be the output and
3:52chances of happening that particular
3:54thing we can uh figure out all those
3:56things right so talking about either
3:59local supply chain or Global Supply
4:01Chain right so we have you know so many
4:04different stakeholders we have so many
4:06different processes different activities
4:09so many different products we are
4:10handling through the same supply chain
4:12right if raw material is there that is
4:14different product finished good is
4:16different product after returning back
4:18that packaging is another different
4:20product right or maybe after the end of
4:22life we are handling another kind of
4:24product which requires allog together
4:26another set of activities to perform
4:28that right and in in that way so many
4:30different organizations are involved so
4:33like we have raw material suppliers we
4:35have manufacturers we have Co
4:37manufacturers we have Distributors we
4:39have warehousing partners and then end
4:42mile delivery partners are also there so
4:45all those things we can model in this
4:48system right so any miscalculation will
4:51you know you will land up with the uh in
4:55the scenario where you will be facing
4:57more risk uncertainties will be there
5:00because you could not figure out the you
5:02know dynamic environment in that way and
5:04then interruptions to your smooth
5:06supplies and then obviously when you are
5:08you know reacting here and there you are
5:11trying to somehow manage the things your
5:14cost will be higher on that side because
5:16you are not you know standardizing the
5:19process so that will lead to extra cost
5:22so then after uh you know uh replicating
5:26all the scenarios within your model or
5:29different models you can pick the most
5:31suitable supply chain model that depends
5:34upon whether that most suitable supply
5:36chain model for one maybe it is cost
5:40efficient and for other maybe it is
5:44responsive kind of model which is you
5:47know responding so quickly to customer
5:49requirements that you are changing those
5:52things so depending upon the different
5:54choices we can then come up with the the
5:57final model either your supply chain now
5:59is cost efficient or you are lying
6:01toward the responsive side right so
6:03these two ends are there either you can
6:06be cost efficient or you can be
6:08responsive but yes we as supply chain
6:11you know managers or whatever
6:13organizations are working we are not
6:15purely saying that we are only cost
6:17efficient we will not take care of any
6:19kind of input which is coming maybe in
6:21terms of quality maybe in terms of
6:23quantity maybe in terms of variety you
6:25are putting to us so that will not say
6:28right so yes some somewhere we'll find
6:30out the mixture okay this particular
6:32product requires most uh maybe cost
6:35efficient supply chain we'll somewhere
6:37set the parameter here maybe this
6:40requires a kind of Emergency Services
6:43Healthcare Services we can set up our
6:45parameters somewhere here right so
6:48obviously the ultimate purpose is to
6:49reduce the cost increase the efficience
6:52and customer satisfaction within that
6:55particular uh Target area where we are
6:58saying that either we are responsive or
7:00we are cost
7:01efficient so when you need to model
7:04these uh different supply chain
7:07scenarios obviously you need to Define
7:09what are the components right how those
7:11components will change here and there so
7:13that the final outcome will change so
7:16let's talk about from the very beginning
7:17Point like procurement so now
7:20procurement again different decision
7:23different scenarios are there we have
7:25multiple vendors or we have single
7:27vendors right and then how many
7:30components we are Outsourcing right so
7:34that means
7:35procurement then how many Lots in one
7:39month you are uh procuring then what is
7:42the lot size right this lot size will
7:45help you to negotiate
7:47on your price as well right so these all
7:52parameters are there when you will
7:54consider the model you will uh consider
7:57all these inputs because if you'll
7:59increase is the lot size this is what
8:01Walmart is doing always feeding very
8:03large orders to their suppliers so that
8:06they can negotiate on the qu uh your
8:08price right so and then they can
8:10maintain the that minimum quality level
8:12as well because this procurement now
8:15your raw material suppliers they need
8:17not to go so quickly for change over
8:20because if you are buying product in
8:22small small Lots maybe then from one lot
8:25they are preparing for 50,000 items then
8:27they are changing over to next lot maybe
8:29for 30,000 they will prepare then 35,000
8:32but instead of that maybe if you are
8:35uniformally you are doing that you are
8:36producing one item 75k units and then
8:41you are changing over for the next model
8:43right so then in that way your supply
8:46suppliers will also be in that way
8:48efficient and you can further reduce the
8:51pricing so this is again one important
8:53component which you can consider second
8:55manufacturing where the manufacturing is
8:58happening right what are are the
8:59different locations are are you going to
9:02provide the uh different are you going
9:04to serve the different Market from
9:06different locations or maybe one
9:08centralized facility manufacturing Hub
9:10will be there and then you will be
9:12circulating throughout the Indian market
9:14that can be the another scenario that
9:16depends upon the initial investment cost
9:18of setting up any manufacturing unit
9:21right so then how much of one particular
9:25model you will if you are having four
9:27five different production lines so uh
9:31this is producing A and B model here we
9:33are producing c d and e model so when
9:37you will go for change over right so
9:39that optimal level you can decide that
9:41if I'll run this particular production
9:44line for 10 hours continuously that much
9:47is going to be the productivity because
9:49whenever we are going for change over in
9:51the process initial time we are taking
9:53to set up you know to standardize that
9:56and then we can produce some rejection
9:58as well so those factors we'll consider
10:01third is warehouse now we have two
10:04different scenario one is we are putting
10:06this Warehouse very near to the
10:09customer or maybe
10:12markets these are n markets are there
10:15and warehouse is very near so quickly
10:17you can respond to the customer right so
10:20that means you can reduce this distance
10:22outbound Logistics distance to the
10:24customer to the market and then quickly
10:27you can respond if this Warehouse is a
10:29kind of centralized facility which is
10:31far away from your main markets maybe
10:34the distance is higher in that case so
10:38then obviously if you are serving to
10:40Market 1 or Market 4 so then that much
10:43distance you are covering right because
10:46why you are doing that because you don't
10:48want to set up so many different
10:51warehouses at different locations If
10:54instead of you know going for 1 2 3
10:58different W houses to serve these 10
11:01markets you are maybe keeping inventory
11:04at three different locations but when
11:06I'm having centralized Warehouse I'm
11:08keeping inventory here only and serving
11:11all 10 markets from the centralized
11:13location in that way we can minimize the
11:16inventory cost but you need to take care
11:18about the transportation cost because
11:20then from Central facility every time
11:23the order is coming from the different
11:25Market you are moving to that particular
11:28location so in that way maybe economies
11:30of scale you also cannot utilize fully
11:33uh you cannot fully utilize the uh
11:36container vehicle uh capacity as well
11:39right Transportation as this is related
11:41to this structure only I told you we we
11:45talk about compare the inbound Logistics
11:47cost with the outbound Logistics cost so
11:51your outbound logistic cost is always
11:53higher that means you want your
11:56Warehouse should be very very near to
11:59your end Market because this distance if
12:01you are covering maybe 100 kilomet and
12:04if this is 5 kilomet so outbound
12:06logistic cost is higher we discussed in
12:08another session why this outbound
12:10logistic cost is higher right and if you
12:13are following this particular model for
12:16every Market you are maybe uh covering a
12:19distance of 35 km 40 km 30 km 25 km so
12:23that means outbound Logistics distance
12:26is now increasing right so in that way
12:28so this is if you will take example if
12:31you're buying furniture from Amazon flip
12:33cart so they are disassembling that
12:36furniture and that it is reaching to
12:38your doorstep and again you are
12:40assembling that so this is one way
12:42because when you will you know buy the
12:45raw material or you will disassemble the
12:47product it is easy to carry that product
12:50from one end to other end the other
12:52example maybe if you are carrying maybe
12:55of these um cars in one container maybe
12:59eight or 10 maximum car you can ship if
13:01that size is little small of the car
13:04right so but if you are carrying the raw
13:06material component you just imagine
13:09within 10 containers how much raw
13:12material you can transport so that means
13:14if your factory is near to the End
13:16Market then that outbound Logistics cost
13:19can be reduced but yes then again other
13:22challenges are also there when you are
13:24moving near to Market so that means you
13:27cannot operate with one facility for
13:30many markets you need to come up with
13:32many manufacturing units right so that
13:35way it will increase that cost so
13:37somewhere in your modeling system you
13:39need to put all these parameters and for
13:41a particular product you're depending
13:43upon the nature of the product size of
13:45the product how frequently you are
13:48shipping demand from the market how
13:50quickly you are changing that product
13:52all those parameters will consider in
13:54that model and will come with the best
13:56fit model which will tell you about all
13:59these parameters eight components how
14:01you can set where you should develop
14:03your warehouses where you should you
14:06know set your manufacturing unit which
14:08kind of Transportation mode you should
14:10use now this is the next parameter
14:12fourth one is transportation we talk
14:14about different modes of transportation
14:17either through Road through rail through
14:19air through ship right but yes all those
14:22parameters are there that if you'll go
14:25by ship the uh least cost method maybe
14:28and and sustainability also will take
14:30care if we'll go by Road it is taking
14:33more time if we'll go by air it is very
14:37you know Speedy Delivery but yes in that
14:40way your cost is increasing right so all
14:43those factors also you will consider
14:45availability of the infrastructure you
14:46can use ship only when the
14:48infrastructure is there right you can
14:50use air facility only when that
14:53infrastructure is there but for specific
14:55products maybe you can go for that kind
14:57of infrastructure right so fifth one is
15:00inventory already we discussed we are
15:02maintaining three different types of
15:04inventory either raw material or Shamy
15:06finished good or finished goods and some
15:09rejected also we are maintaining the
15:11inventory in the hope that we will
15:13repair these products and again we will
15:15resell in the secondary Market or in the
15:17same Market or somewhere recycling
15:20remanufacturing can be ensured right so
15:23now if you are again we'll take that
15:25scenario only when we are keeping our
15:27warehouse very near to the uh your
15:30markets so that means now you are going
15:33with multiple warehouses for one for uh
15:37maybe uh this Market these 10 markets
15:40you are serving by Warehouse 1 we have
15:43Warehouse 2 which is again serving maybe
15:46another 20 markets we have Warehouse 3
15:50which is again serving maybe another
15:522025 markets right so that means if we
15:55are moving very near to customer we have
15:59multiple warehouses so that means that
16:01will increase the cost of keeping the
16:03inventory because every Warehouse you
16:05will keep the inventory that's why
16:07online platform when we are going so now
16:10because they are opting for zero
16:13inventory model it's not inventory LE
16:15model where earlier they were taking the
16:18inventory from the you know
16:19manufacturers from the sellers and they
16:21were storing in their warehouses and
16:23whenever Amazon is receiving the order
16:26they were processing that order but now
16:28what is happening they are not uh
16:31keeping inventory in that way minimizing
16:33the inventory cost and directly
16:35channelizing the product from the seller
16:37to the customer right so then you can
16:42save the inventory cost so these factors
16:45we will consider because obviously when
16:48you will model these factors you will
16:51ensure that you are not stock out and
16:55also you are not keeping excess
16:57inventory with you that these two points
17:00you need to keep in mind right then
17:02Information Technology this is a kind of
17:05your component which will affect your
17:08all the players all the activities in
17:10that supply chain Network so you need
17:13quick relevant information transparency
17:15should be there where is the material
17:18right now right how much quantity coming
17:21how much Quant maybe units damaged
17:23during Transportation during handling
17:26how many are finally consumed in the
17:28production of floor how many finally
17:30consumed in the ne End Market consumed
17:33by the customer so then this information
17:36technology what type of system you are
17:38implementing to ensure the quick
17:41information sharing system then again
17:43customer service is another Point what
17:45type of expectations from the customer
17:48coming in terms of the quality level
17:51they are expecting in terms of the
17:53quantity in terms of the frequency how
17:56frequently they are buying in terms of
17:58the variety they are why uh they are
18:00looking for in terms of the innovation
18:02in the product they are looking for in
18:04terms of the delivery they are looking
18:06for right either they are looking for
18:08Doorstep Delivery or they are looking
18:10for some pickup locations or they want
18:13to F the product and then they want to
18:15purchase that product right so and this
18:18is not only the end then customer
18:20expectation in terms of returns how you
18:23will manage those returns right or in
18:25terms of if end of life of that product
18:30how you are recycling reusing
18:32remanufacturing that product right the
18:34final component is a distribution how
18:36you are ensuring the distribution
18:38Network right either the last mile
18:40delivery then last mile delivery then
18:42how you will combine consolidate
18:44multiple orders from the same location
18:47or maybe sometimes what is happening
18:49when you are not so efficient with your
18:51information then the same Amazon if you
18:54are placed two three different orders so
18:56maybe the delivery person different
18:58delivery person they are making multiple
19:00visits to your doorstep instead of
19:03consolidating into one order and then
19:06that cost can be reduced but you you
19:08have experienced that as a customer that
19:11from Amazon morning to evening you are
19:13receiving sometimes multiple deliveries
19:15so why can't they consolidate all the
19:17orders and then deliver at one time
19:20right but there may be then you know uh
19:23challenges with that maybe customer want
19:25some products uh before 10:00 in the
19:28morning related to maybe breakfast or
19:30kind of things right grocery products or
19:32some products maybe anywhere during that
19:35day you can transport or deliver right
19:38so then you need to manage means you
19:39need to model those factors in
19:42your uh this modeling uh analysis right
19:45so already I talked about all these
19:47components in detail so significance of
19:50supply chain modeling first is you need
19:52to this is the ultimate aim we need to
19:55minimize the cost but maximizing the
19:57services right so how efficiently we can
20:00ensure the flow of those three basic
20:03components material is one information
20:05is another one how efficiently we can
20:08ensure the flow of Finance right
20:10planning and controlling of the flow of
20:13these resources right we need to plan
20:16supply chain modeling now once you uh
20:19worked with all those eight components
20:22now you have developed the Matrix for
20:24all those eight uh components now you'll
20:27go for modeling that that you will put
20:29all those constraints these are the
20:31capacity limitations these are the
20:34requirements from the customer this is
20:36the capacity of our capability of our
20:39supply chain then you will go for that
20:41uh modeling and then again I was talking
20:44about experimentation and scenario
20:46analysis where you will come up with
20:49different simulation models different
20:51scenarios you will simulate and then
20:54depending upon different scenarios what
20:56is the output either you are cost cost
20:58efficient or responsive how you are you
21:01know uh making the tradeoff between cost
21:04efficient and responsiveness and types
21:07of model you can pick linear programming
21:09where we are defining the linear
21:10relationship that is the basic
21:12Assumption of linear programming we are
21:14defining the linear relationship between
21:16the uh all those independent variables
21:18right so whatever we are uh you know
21:21modeling in that particular model and we
21:23go for simulation where we are depicting
21:26different scenarios queing model what is
21:29the you know this also we are
21:31considering how many orders are in Q
21:33what is the Q length idle waiting time
21:35so right processing time all that will
21:38consider in queing model statistical
21:40model different statistical model either
21:42you talk about you know stochastic model
21:44you talk about nonprobabilistic model
21:46deterministic models right so those
21:49models we can work with in operation
21:51research we have seen all those models
21:54like when we talked about transportation
21:56management this we will discuss in our
21:58this lecture all always right so that is
22:01part of your Supply Chain management
22:02Transportation then assignment models we
22:04talk about Simplex method we talk about
22:07uh the project management using you know
22:09different uh Pro listic Models like
22:11critical path method and program
22:12evaluation and review techniques so
22:14those all are uh different types of
22:17modeling techniques kind of methods we
22:19which we can use to optimize our
22:22processes all right so selecting the
22:24right model depending upon the uh your
22:27requirements you will pick Pi the right
22:29model in the end so these are different
22:31types of model we'll just quickly go
22:33through what are these different types
22:35of model and in that way which model is
22:39uh you know best fit for a particular
22:42kind of industry if you talk about
22:44continuous type of model this is a kind
22:46of model where continuously your product
22:48is flowing through the supply chain not
22:50only product product information and
22:53your uh your money right your Finance
22:56these three are continuously flowing how
22:58when you can ensure the continuous flow
23:00of these resources obviously when you
23:02are having fixed kind of demand and
23:05fixed Supply right you are pretty sure
23:08about your forecasting process your
23:10demand function and your supply function
23:13so in that way uh it should be a
23:15continuous kind of model right so
23:17material should flow continuously from
23:19the supplier to uh customer but what
23:23should be the objective okay first we'll
23:25take the examples uh for this continuous
23:28kind of model already we talked about so
23:31many all the industries related to you
23:33know commodity kind of product I you
23:36talk about grocery item usually
23:39continuous kind of model we can uh go
23:42for right or maybe uh fmcg products many
23:46products we can use a continuous kind of
23:48model steel industry continuously we are
23:50producing sugar industry salt industry
23:53these are the best example because as
23:55for the population we can fix this is is
23:58going to be the demand and we can go for
24:01that right so this is uh uh these are
24:04some of the examples now when you are
24:06saying you are sure with the demand and
24:08Supply so obviously the ultimate
24:10objective should be you should be very
24:12very efficient right so because
24:16otherwise what is there to compete the
24:19first thing is how
24:21efficiently how with the minimum cost
24:24you are providing that product or
24:25Services right so that should be the UL
24:28imate objective of your continuous kind
24:30of production system other
24:32characteristics just in time we can
24:33Implement because we are sure about the
24:35demand and Supply so we can just get the
24:37supplies continuously uh feed those
24:41supplies to the production lines and
24:43right we in that way we can reduce the
24:45inventory and inventory cost as well and
24:48waste also can be eliminated because now
24:51that waste which was we were generating
24:54earlier from the inventory while
24:55handling managing updating the inventory
24:57we were damaging the products in the
25:00inventory only sometime product were
25:02going obsolete so that waste we can
25:04reduce right so it is kind of smooth
25:06flow one piece flow kind of production
25:09system it is coming from one end leaving
25:11the other end so that is there
25:14advantages all those advantages we can
25:16minimize the cost we can be more
25:18efficient in that way productiv
25:20productivity will be improved right
25:23disadvantages basic disadvantage of
25:26continuous system is less flexible
25:29because we are not considering the
25:30dynamism of the environment because it
25:33is a continuous model and any disruption
25:36major disruption will interrupt the
25:38suppli sometimes will interrupt the
25:40demand sometimes right so that uh is the
25:43basic disadvantage if we'll talk about
25:45the Amazon they usually focus on this
25:48where type of model where they are
25:50continuously you know sourcing the
25:52product from the sellers and delivering
25:55the products at the minimum cost right
25:58so you can see uh this model is suitable
26:01for those business already I talked
26:03about the examples where you have stable
26:05demand and Supply you can see this is
26:08the standard shipping where you were
26:10having you were placing the order but
26:12your fulfillment center is far away then
26:15they are using some USP services to
26:17transport that product and finally here
26:20you are where you were getting the
26:22product right now the concept is same
26:24day delivery when you are very your
26:26fulfillment center is very near to your
26:29customer that is what I discussed in the
26:31initial slides right but then if you are
26:33near to the customer you are adding more
26:35responsiveness right but in that way
26:37cost efficiency will not be there
26:39because so many locations will be there
26:42you need to serve all those locations
26:44you are maintaining inventory at all
26:46those locations so inventory cost will
26:48get high but outboard Logistics you will
26:50you will reduce right and in that way
26:53somewhere you can manage in that model
26:55which you want type of product and then
26:57you your customer is ready to take same
27:00day delivery then yeah you know uh we
27:03cannot separate these two things at two
27:05different centers why to make multiple
27:07stoppage right so then it is ultimately
27:10adding to the cost so why don't we do
27:12this in one station and then crowd
27:14sourcing means local career Partners we
27:17can pick who are frequently deling here
27:19and there and we can collaborate with
27:21them and finally within same day
27:24delivery or 1 hour delivery you can go
27:26for second is fast mod model obviously
27:29this is obviously the speed is the main
27:32parameter of this and where you need
27:34quick delivery and timely response and
27:37Zara is the best example where they are
27:40not only quickly delivering the product
27:42but they are replenishing the new
27:44inventories with the old inventory right
27:47so we have discussed one case where the
27:51you know the the new trends whatever new
27:54fashion they are coming with the average
27:56of that industry is six to seven weeks
27:59but Zara is almost every other week they
28:03are changing the Trends they are coming
28:06up with the new products right so this
28:08fast model is best fit for those product
28:11which is having very limited Cy uh life
28:14cycle right if we talking about
28:15entertainment industry if today any
28:18movie or web series is launched it
28:20should be distributed fully wherever you
28:22want to distribute right it should not
28:24happen that now today you are launching
28:26it somewhere in the North then you're
28:28going to South Eastern part West this
28:30cannot happen right so this these kind
28:33of products which uh got very limited
28:35shelf life you need to distribute using
28:38your fast model right so and seasonal
28:41products also because those products
28:44maybe for four or 5 months season so
28:46within that time you need to ensure the
28:49raw material uh acquisition production
28:53and final consumption as well within
28:55those four five months right then we
28:57talk about efficient supply chain model
29:00so this is again we can uh extend the
29:03first model which was your continuous
29:05flow model uh we can relate it with that
29:08because here also we we uh Target a type
29:11of supply chain model which is you know
29:14which can minimize the uh cost right
29:17this is the only thing we are adding
29:19here efficiency right so when you are
29:21sure about the demand you are sure about
29:23the supply only thing you the fight is
29:26how efficiently you can deliver the
29:28product to the End customer from the
29:31suppliers right so that is that should
29:33be the priority so you can see uh here
29:36we keep priority on efficient efficient
29:39Inventory management and getting the
29:41most Auto production uh workers and
29:44Equipment right so that the productivity
29:47can be improved with minimum outputs you
29:50can get the maximum out input uh output
29:53right with minimum inputs you will get
29:55the maximum output this is how you can
29:57increase the the productivity so this is
29:59efficient model obviously the challenges
30:02are there to the efficient model because
30:04you cannot change the cost of the raw
30:05material and commodities whatever you
30:07are using and then capacity constants
30:10are also there maybe 24 hours into 7 you
30:13are running your plant even then you
30:15cannot manufacture more than 50 lakh
30:17units so then how you can you know deal
30:21with that capacity
30:23constant uh shall we come up with the
30:25new machine or new plant then again can
30:28be utilized that full capacity of the
30:30new machine right and some other
30:32challenges label shortage delays are
30:35there right so highly competitive
30:38sectors are the reasons why we need
30:40where you know almost products are
30:43similar kind of thing features you are
30:45providing so it becomes very very
30:48important that you provide efficiently
30:50with minimum cost that can become your
30:53one USP with the minimum cost you are
30:56providing the products to the customer
30:58so Zara we can go this is the best
31:01example is suiting for uh different
31:05cases you see when we have very short
31:08lead time this is the only lead time
31:11right if you'll see lead time is on y
31:13AIS so we should Source from the local
31:17markets like Spain Portugal or Morocco
31:20right so so that it can be shipped
31:23quickly right so items with high fashion
31:27components means which you are changing
31:29very frequently you need to ship
31:31immediately you have limited lead time
31:33items with reasonable lead time you can
31:36go for offshoring but nearby offshoring
31:39because that much time is not there
31:41maybe compared to this it is little
31:43higher but still we need to go for
31:47minimum maybe distance whatever we can
31:50uh you know manage up to that point and
31:52we can be those location can be
31:54somewhere cost efficient location or
31:56maybe some extra Advantage we are
31:58getting out of the that location in
32:00terms of technology in terms of skills
32:02we are getting in terms of the cost
32:05those can be there right if we are
32:07having that much lead time little higher
32:09lead time so then we can go for lowcost
32:12country like China India where we can
32:15get the cheap labor where we can get the
32:18cheap uh you know resources right all
32:20those resources and we are producing
32:22some you know raw material silk also
32:24there and then we have when more time
32:28all the time and then you can go for
32:30highly standardized kind of you know
32:33items and uh you can go anywhere in the
32:36world wherever you are finding the best
32:39locations in terms of quality cost
32:41availability of raw material
32:43availability of manow skills technology
32:45whatever is required so then you can
32:47permanently you can Source those product
32:49to uh those countries right so this is
32:52the best example you can understand
32:54agile model next is where you are are
32:58responsive to you know quickly adapt the
33:01changes in the market so Nike is doing
33:03it beautifully any changes are coming or
33:06recommendation are coming from the
33:07customer they are quickly amending the
33:09product right so aile model works on how
33:13quickly you are responding to the cust
33:16any changes in the scenario how that can
33:18happen when you your communication and
33:21collaboration network is very strong
33:23that any change is here you are uh you
33:26know uh you are sharing immediately with
33:29your supply chain stakeholders right so
33:32you can
33:34see uh agile kind of supply chain is
33:37required when you are dealing with those
33:39kind of products which where you are not
33:41sure about the demand where everyday
33:43renovation Innovation is happening with
33:45the product or process right so in that
33:47way the four characteristics are you
33:50need to align your process virtual
33:52integration with all the stakeholders
33:54and all the activities throughout the
33:55network a network was should be there
33:58and Market sensitivity you need to sense
34:01the market trends and then you can come
34:03up with that plan right so this can be
34:05only I told you agile model is only
34:08possible when you are jointly managing
34:11with your suppliers so how supplier can
34:14manage the inventory at some point of
34:16time right it's not only the
34:18manufacturer responsibility and then if
34:20you are changing the product design
34:22process design accordingly your
34:24suppliers need to supply the material
34:27custom configured model where we are
34:30providing the customizing Services right
34:32supply chain according to the specific
34:34customer needs so this is the best
34:36example always we talk about Dell where
34:39you go to Dell showroom you tell your
34:42specifications in terms of processor you
34:44want in terms of ram ram capacity
34:46storage speed processing speed you want
34:50which generation so uh you know
34:52processor you want all those
34:54requirements you can give and then
34:56through there you know customer uh
34:58customer uh configured model they will
35:01provide you that product right and how
35:04they are doing that it combines aile and
35:07continuous flow methodology right so
35:10they are buying all the raw material
35:13components from uh the various part of
35:15the world usually from the Asian
35:17countries and then they are as per the
35:20configuration you are giving they are
35:22assembling the product and providing you
35:24so this is direct kind of supply chain
35:27flexible model again kind of aile model
35:30flexibility where you are dealing with
35:32the you know weing customer demands
35:35anytime you need to change you that
35:37should be incorporated throughout your
35:39supply chain so part segmentation
35:42precise talking algorithms and flexible
35:44planning are some of the main components
35:47when you are talking about the flexible
35:49model so you cannot go for you know huge
35:52pile of inventory and then you are
35:54consuming slowly in the market maybe
35:56tomorrow scenario will be different and
35:58then you require another set of raw
36:01material another set of inventory so
36:03then how you will ensure the consumption
36:06of the inventory already you acquired so
36:08that flexible model doesn't work for
36:10those kind of items right and uh it's
36:13kind of you know you need to turn it on
36:15turn it off kind of pro thing and H will
36:19packet uh is the best example for this
36:22kind of flexible model where they are
36:24you know sensing the market conditions
36:27and recording the customer preferences
36:30so frequently and providing the flexible
36:33Supply Chain Solutions right another is
36:36your virtual supply chain model right
36:38which is you can consider the future
36:40supply chain model as well where we are
36:42taking the help of cloud computing and
36:44data analysis so uh AI iot system will
36:48be part of this the only purpose here is
36:51we have the greater visibility and
36:53communication the things I was talking
36:55about to be more efficient C is flexible
36:58what you need you need to share
37:00information quickly and Proctor and
37:02Gamble they are doing this they are
37:04sharing the information quickly and
37:06managing the minimum inventories right
37:08in that way because only when product is
37:10required they are shipping that product
37:13right so best practices for supply chain
37:16modeling you need to understand the
37:18business needs already we talked about
37:20different products required different
37:22types of supply chain if we are talking
37:23about steel industry Continuous Flow
37:26efficient supply chain if we talking
37:27about Healthcare emergency supply chain
37:30you need to go with an asile and
37:32flexible supply chain disaster
37:34management That's Entertainment kind of
37:36supply chain you need that uh you know C
37:39supply chain Speedy supply chain right
37:41choose appropriate approach so you need
37:45to pick the uh modeling method approach
37:48how you'll model those factors whatever
37:50you have figured out right and validate
37:53with the accurate data so it's not only
37:55once you picked the model it will be
37:56forever
37:57you need to validate the established
38:00model again and again right and you need
38:02to test all the scenarios what if that
38:05happens right what if it will go the
38:07other way we thought maybe in this way
38:09and it the some other factors came into
38:12picture we did not model those factors
38:14and now you know playing very important
38:16part right so then we can consider
38:19different models obviously we can
38:21involve the
38:22stakeholders when whatever model if you
38:24are going for aile kind of model you
38:26need to uh you know take the views from
38:29your stakeholders whether they are ready
38:31for that kind of flexible aile quick
38:33model or not regulate regularly you need
38:36to update and refine that model right
38:38already I talked about document
38:40assumptions see whenever you are
38:42developing any model there are
38:44limitations like we talked about linear
38:46programming and then we said that the
38:49independent variables are having linear
38:51relationship if you're talking about
38:53Simplex Mo method right where we are or
38:55Transportation method where we are
38:57talking talking about the linear you
38:59know relationship right so if those
39:02assumptions you are forgetting when you
39:04are implementing then your model will be
39:07then it will be a disaster right so you
39:09need to train and educate the people how
39:12they can use the best model and those
39:14best model can be tracked against the
39:17set of performance measures you have
39:18already defined obviously we have some
39:21challenges because complex
39:23relationships different stakeholders
39:25different activities it is very
39:28difficult to configure all those
39:29activities in one model but still we are
39:32trying to cope up with that and Define
39:34those eight components so that we can
39:37include parameters from all those things
39:39because these kind of models work only
39:42with the data you need data to validate
39:44right you need data to develop those
39:46models right so then how accurately you
39:49are getting timely information you are
39:51getting so that also Big Challenge
39:53multiple objectives because multiple
39:55organization so one organization may be
39:57competing on responsiveness other
39:59organization is competing on cost
40:01efficiency so then multiple organization
40:03multiple objectives so that can be
40:06another challenge Dynamic environment
40:08this because Dynamic modeling is
40:10required where everyday situation is
40:12changing you need to change your model
40:15as well integration issues are there
40:18because uh multi Supply chains are there
40:22and various stakeholders are there so
40:24then sometimes those conflicts can be
40:27there whether we want cost efficient we
40:29want responsiveness we want to manage
40:31inventory where the customer we should
40:34keep inventory centralized so those
40:36issues can be uh you know sometimes
40:39making decisions can be little difficult
40:41resource constants always we are working
40:44with the limited resources in terms of
40:46all the resources time money machine
40:48method material whatever we are using
40:50scenario complexity because these days
40:53we have multiclone networks and Omni
40:57Chan distribution multiclone Network
40:59where we are providing the distribution
41:01multiple distribution points from the
41:03distributor and or supplier end to the
41:06customer om Channel distribution where
41:08the customer can come from the multiple
41:10channels right like you can book air
41:13ticket through goo through make my trip
41:16through exigo through your IRCTC through
41:19the Indigo website or B whatever
41:22Airlines you are picking right so that
41:24scenario complexity is there risk
41:26management
41:28obviously because you are working in the
41:29dynamic environment and that any change
41:32either in all those pestal Dimension
41:34political ethical legal technological
41:37social all those environmental any
41:39change in that environment will disrupt
41:41your supply chain and how quickly you
41:44will manage that CH change and how
41:46efficiently you will manage that will
41:48help you you know change management this
41:51is the future of modeling supply chain
41:53control Towers now this is the
41:55combination of machine learning learning
41:57artificial intelligence iot because now
42:01we were talking about eight components
42:03and we have seen under manufacturing so
42:06many components can be there so many
42:08model we are manufacturing within those
42:10production lines on one assembly we are
42:13making multiple models then we are
42:15keeping the inventory for each model
42:17each component packaging Distributing so
42:20many different parameters how you will
42:22keep the track of all those parameters
42:24then we need these cloudbased Solutions
42:27supply chain control Towers which will
42:29take care of every parameter you need
42:32information related to warehousing
42:34whatever parameters you'll say we'll see
42:36in the next session what are the
42:38different you know parameters when we'll
42:40talk about you know Transportation
42:43facility inventory so all those
42:46parameters you just say and supply chain
42:48control tower will track those
42:50parameters will give you the data now
42:52you feed that data in your model and
42:54you'll get the output right so if you'll
42:57travel from here to the other point so
43:01how much cost will be there how much
43:02lead time is required when the end time
43:05delivery will happen and all those
43:07things you can figure out so this is a
43:10small video on uh this uh the smart
43:13solution we'll just go through this
43:15video so you can just see that these
43:17supply chain control Towers when because
43:20when we are talking about different
43:21types of mathematical modeling we'll go
43:23for we need data for all those
43:26parameters whatever you have set for
43:28transportation right per unit uh
43:30shipment cost right per unit
43:33manufacturing cost Warehouse cap
43:35capacity utilization different
43:37parameters we are recording so this
43:39supply chain control tower will help you
43:41to record all those parameters and then
43:44you can uh you know develop your best
43:46fit model which model is best for your
43:49type of product right so quickly we'll
43:52go through the case studies first is
43:54case study is about Intel who is the Lar
43:57largest manufacturer of computer chips
43:59the problem here I'm highlighting how
44:01they they became more efficient right
44:04earlier they were producing uh chips and
44:07that chip were costing around $100 right
44:10so they were spending almost
44:12$5.5 per chip for transportation supply
44:15chain cost right but now because they
44:19are coming up with the new chips and it
44:21is a low cost atom chip so which cost
44:25around $20 so if it is costing $20 and
44:29the shipment supply chain cost is still
44:325.5 that is big question so then what
44:35can be done so they analyzed all the
44:37scenario and they came up with the
44:40discussion this that inventory is
44:42something which we can you know finally
44:44reduce the cost but why before inventory
44:47they worked with you know uh uh
44:50packaging they scale down the level of
44:52the packaging required for that so that
44:55high value to weight r ratio can be
44:57maintained packaging material can be
44:59reduced already they did that right but
45:02still the cost was around $5.5 per chip
45:05right they already worked with the
45:07distribution cost and they they know
45:09know that now further it cannot be
45:12reduced so only thing is we can reduce
45:15the inventory and there we can save that
45:18cost right so but to reduce the
45:21inventory the basic problem was 9 week
45:24order cycle if today you will place the
45:27order after 9 weeks you will get the
45:29order so that means minimum 9 to 10
45:31weeks inventory you will keep that so
45:33first point while in reducing an
45:35inventory cost how we can bring this
45:37cycle time down maybe somewhere two to
45:40three weeks so then we'll maintain
45:42inventory only for two to three weeks so
45:44then they thought that okay we'll go for
45:46make to order once we'll get the order
45:48only then we'll make that right this is
45:50one approach they opted for and then
45:53they instead of you know cutting the
45:55chip assembly test window from 5day
45:58schedule to Y weekly 2 Day schedule
46:01right so 3 days they cut there and then
46:04they analyzed very frequently and
46:07minutely their sales and operation
46:10planning right so how they you know
46:12sales is happening that uh throughput
46:15rate how the end sale is happening in
46:17the market right accordingly they
46:20planned the production and then they
46:21asked their vendors you keep the
46:24inventory we don't want to keep the
46:25inventory so this is how how they worked
46:28on these two three parameters and then
46:30you can see they reduced
46:33$4 per unit for that atom chip which was
46:37costing them 20 and you can just imagine
46:41the the saving they they have done just
46:43working on the inventory but minimizing
46:46the
46:47inventory again have some challenges
46:50right so if you are minimizing the
46:51inventory that means production
46:53schedules are very very tight schedules
46:55means small orders you producing so we
46:58already know that if you are producing
47:00in small orders production cost will be
47:03higher because economy is of scale then
47:05Transportation if you are shipping small
47:08orders full capacity may not be utilized
47:10then again economies of scale right will
47:13be in question but yes then that's why
47:15we have this m mathematical modeling
47:17system we'll we will keep our data feed
47:20our data and we'll get the best result
47:22because you cannot count on fingers this
47:25is best in this case this is better in
47:27this case this is so that we cannot do
47:29that so let us feed all the data in that
47:31system and we can go for that another
47:34quickly case of pharmaceutical industry
47:37and you know that pharmaceutical
47:39industry you know everyday scenario
47:41where Acquisitions are happening and
47:43divestment studies is are happening
47:45every day so this is one organization
47:48which is expanding by you know baning
47:50with other organization and uh keeping
47:53out their known uh you know core
47:55businesses and that non Core Business is
47:57one of is your distribution Network and
48:00then they suddenly realized okay this is
48:02going very inefficient right and
48:04fragmented way we need to organize this
48:07how we can reduce cost enhance customer
48:09services and establish a reusable
48:12modeling tool which can automatically
48:15you know model the scenario and will
48:17give you the best option so we are doing
48:19fairly well with the manufacturing as
48:21that is our Core Business activity but
48:24we need to work on the distribution
48:26Network right so they used mixed integer
48:29linear programming problem where they've
48:32model all those spectors which were
48:33require required like manufacturing
48:36sites how many sites where the location
48:38they will decide distribution centers
48:40again location how many required demand
48:43from which reason demand is coming what
48:45are the demand pattern aggregated
48:47product groups right so they in included
48:50all those products and then they
48:51included the global as well as your uh
48:56local business business levels right so
48:58then they model that and you can see the
49:02result indicated significant Improvement
49:05and 10% reduction in the overall cost
49:08and next day delivery they ensured in
49:1195% of the products right just working
49:15on their distribution Network how they
49:17can model using mixed integer uh your
49:20linear programming model right where
49:22they fed all this data and then they
49:25reduced the overall cost cost so despite
49:28this because obviously when we are
49:30talking about you know linear
49:32programming and all that getting data
49:34from different sources different
49:36stakeholders and then you know uh then
49:39you are so frequently recommending the
49:41changes and we are always we are as a
49:44human being we are you know first we are
49:46resisting to the change so that is one
49:49Challenge and other challenges in terms
49:51of how you will take care of the data
49:53coming from different stakeholders
49:55different activities right so then you
49:57have to make the tradeoff right so
49:59decision makers to understand tradeoffs
50:01and make informed decisions you can make
50:04informed decision when you have that
50:06information which is you can easily
50:08record through your supply chain Towers
50:11right so uh we can here we can conclude
50:15that modeling system will help you to
50:19simulate all the different scenarios
50:21what can be different outputs what are
50:23the possibility of getting the success
50:26or
50:27and how the environment will behave all
50:29the environmental factors you can model
50:31into that right so this is particular
50:33thing but then you need to feed regular
50:36data it's not onetime fixed model
50:39regularly you need to feed the data from
50:42where you can get that data already we
50:44talked about your supply chain Tower
50:46solution control tower where you know
50:48the virtual word is there for all the
50:50physical items you have right so
50:52anywhere you are depleting inventory it
50:54can record quickly right anywhere you
50:57are damaging the product it can be
50:59quickly replenished re uh it can be
51:01shared throughout the network any
51:04customer complaint is coming because of
51:06Any defect can be quickly shared with
51:08the manufacturing process right so with
51:11this uh we end this that this modeling
51:14system will help you to improve the
51:16efficiency as well as the responsiveness
51:18of the overall supply chain so these are
51:21the references for further reading so
51:24that's all thank you very much

About this transcript

This page contains the full transcript of Lecture 28 : Supply Chain Models and Modelling Systems by NPTEL IIT Kharagpur, generated from the public captions YouTube serves with the video. The transcript has 8,300 words across 1,248 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.