沃顿商学院全套笔记-八-

沃顿商学院全套笔记（八）

沃顿商学院《商务基础》｜Business Foundations Specialization｜（中英字幕） - P122：6_如何以微观角度看待工作.zh_en - GPT中英字幕课程资源 - BV1R34y1c74c

In this module， we will look at the operations of a process at the very micro level。

We get our stopwatch and look second by second on how people work。

The case study I have picked for this module is Subway Restaurants。

I somewhat assume that all of you have eaten in a Subway restaurant or similar venues。

so it should make for a good case study。 I also assume that every one of us has made a sandwich for ourselves。

Now unless you are a professional cook or have a big family。

your experience is all about making one sandwich at a time。 How long does it take you？

What is your processing time？ How long does it take you to slice the bread？

Here's some official Subway information of how long it takes to make a sandwich。

This will vary a bit by sandwich and employee， but， let's just take these numbers as given for now。

We see that we have a grand total of 120 seconds of work per sandwich。 That is two minutes。

In other words， one employee could make a sandwich every two minutes。

which is equal to 30 sandwiches an hour。 That's the employee's capacity。

But what would Subway do when demand is higher than this？ Maybe 80 or even 100 sandwiches per hour。

which are common numbers during busy lunch times。 Well。

they certainly need more employees to do this， but， how should we organize the workflow？

Save me higher two more employees， so they now have three。

One thing that they could do is that each of the three employees serves。

the customer from beginning to end， from grading the customer all the way to checkout。

Each employee would then take 120 seconds to serve a customer。 Interestingly。

that is not the process that they use。 And we later on would understand the logic behind this。

For now， here's what they're actually doing。 Take a look。 What Subway does is here。

rather than having one person， doing the whole thing， making the whole sandwich。

there are three people involved in the job。 They're really setting up a mini assembly line of sandwiches consisting of employee one。

two， and three。 Each of them has their hands in our sandwich， yummy。

So the first employee is greeting the customer， and then employee one does all。

the things after putting the cheese on the sandwich。 Then the second employee takes over， and。

that person starts with putting on the onions and goes all the way up to wrapping。

up and begging the sandwich。 And then the third employee is gonna offer the fresh value meal and。

then rings out the cash register。 That completes one customer。 So for now。

let's see the numbers here， these processing times I exact。

I know this is a very mechanical view to the world， but as you will see， it is a good start。

It takes exactly these 12 seconds to put on the meat。

It takes 37 seconds per sandwich per unit per guest at employee number one。

The processing time is 46 seconds per unit for the employee number two and。

37 seconds per unit at employee number three。 So these are the processing times。

Here's what I want you to do。 I want you to draw me a process flow diagram of this subreassembly line for。

sandwiches。 And then I want you to think about the capacity of this process。 As usual。

put me on hold， pause me， and then see how far you can get this done by， yourself。 All right。

here we go。 Let's look at this together。 Let's start by drawing the process flow diagram。

For the process flow diagram， we start out with a triangle of waiting customers。

So that's a triangle。 Then comes the first box， the first resource。

That's going to be employee number one。 There could be a buffer then。

We have employee number two afterwards。 So that's the second resource， the second box。

And after that， there's another buffer。 And that is the other employee number three。

And that completes the process flow diagram。 So we have three resources。 Next up。

I want you to think about the capacity of this process。 For the capacity of the process。

I have to find the bottleneck。 There's going to be one of these three workers。

Each of them has a capacity。 We're going to find it by taking one over the processing time。

So one over 37 is a capacity of the first worker。 And that is in sandwiches per second。

One over 46 sandwiches per second is a capacity of the second yield。 And one over 37 is the third。

So we see here that the lowest number is one over 46。 And so that is by definition our bottleneck。

Rather than messing around with mental maps。

let's switch over into Excel and do this on a spreadsheet。 We have the three workers。

That's three resources here。 And each of them has a processing time。 And remember those were 37， 46。

37 seconds per unit。 Let's make sure we keep track of the units here。

These are all in seconds per unit。 Now the capacity we set is simply one over the processing time。

So one over 37。 And that is 0。02727。 That's a really hard to interpret number。

This is expressed now in units per second。 And to make it easier to interpret， I would suggest。

we just convert it to units per hour。 We do this by simply multiplying this number by 3600， right？

3600 seconds in an hour。 And 97 sandwiches per hour is a number that， is much easier to interpret。

So remember then the process capacity reset， the process capacity is simply the minimum。

of these individual capacity levels。 That was the whole idea of the bottleneck。 The bottleneck。

not surprisingly， is at station number two。 And the process capacity is 78 units per hour。

From there， the next thing that we， want to think about is flow rate。 Flow rate， we said。

was the minimum of demand and capacity。 So we need to make an assumption about the demand rate。

Let's see for now， the demand is 100 sandwiches per hour。

And so that means our flow rate is the law of these two， numbers。

But the minimum between demand and capacity and the flow rate。

now is this going to be 78 units per hour。 We also talked about the idea of utilization。

So for utilization， I'm taking my flow rate。 So I'm fixing the 78， and I fix them in Excel。

by pressing the dollar sign。 And I'm dividing it by the capacity。

So that means my first resource has a utilization of about 80%。

And then the other resources have a utilization of 100%， and 80% respectively。

Now just for intuition building， imagine my demand rate， would go down。

Say we have a demand of 50 customers per hour。 Then my utilization， even at the bottleneck。

would decrease to a value that is less than 1。 In fact， with the demand of 50， it。

would do down as you see here to about 63%。 But for now， let's not worry about that。

Let's go back to the base case of demand being 100。

So we have a utilization of 100% at the bottleneck， and about 80% at the other two resources。

Next up， I want to define a new measure。 I will define the new measure as the cycle time。

Let me define cycle time as 1 over the flow rate。 Let's be very careful with the units here。

I've been a little sloppy in the last row， skipping track of the units。

Flow rate is measured in units per hour。 And so my cycle time since it's 1 over the flow rate。

is now hours per unit。 Now 0。012778 hours go by between two sandwiches。

It's a little hard to interpret。 So I suggest we just convert this back into seconds。

and we divide the 3，600 seconds of the hour by 78 sandwiches。

Now we see something that I hope is intuitive， as we are now back to our seconds per unit。

our cycle time is 46 seconds per unit。 And that means we're making a sandwich every 46 seconds。

So work repeats itself。 We are cycling through every 46 seconds， to make a new sandwich。

Now that we have the cycle time， I want， to ask you another question。

Imagine we have a school bus with 20 hungry kids， arrive at a subway restaurant。

Every kid wants one sandwich， and right now the restaurant， is still really empty。

How long will it take to please the crowd？ How long will it take to make 20 sandwiches starting。

with an empty process？ This is harder than it sounds。

So let's think about this systematically。 How long will it take to feed the first kid？ Well。

the first kid， the first customer in order to be served。

we need to basically go through all three steps--， 37 seconds at station one， 46 at station two。

and 37 at station three。 Now how much later will we please the second customer？ Think about that。

The second customer will come out 46 seconds later。

And another 46 seconds later comes out the third customer。 So from then onwards。

we are completing a customer order， every 46 seconds。 That is exactly the idea of cycle time。

So to please the crowd of 20 customers， to feed 20 customers， we're going。

to have to feed the first one。 And then 19 times more， we're going。

to go through the cycle time to have completed， the total order of 20 units。

That will be after 994 seconds。 Before advancing further， I would。

like to review the calculations that we did in Excel。

a moment ago。 Remember， we have these three resources， with processing times 37， 46。

and 37 seconds per unit。 I try to visualize this here in this graph with 37， 46， and 37 SEs bars。

We can imagine already that station two， is going to be the bottleneck， but that。

would be jumping ahead。 The bottleneck is defined based on capacity， not based on processing time。

So to find the capacity of a resource， we look at the number of parallel servers at that resource。

M divided by the processing time。 Now here， each resource， each station。

is really stuffed by one employee。 And so my capacity level is 1/37。

and that is expressed in sandwiches per second， 1/46， and 1/37。

That makes this step here the lowest number， and so employee 2 is going to be the bottleneck。

So my process capacity is 10， 1/46， and it's 10， which is 10， which is 10， which is 10。 For now。

I want to leave it in seconds， as opposed to converting it to hours。

as I did in the excess spreadsheet， and you will see why in just a moment。

That gives me my process capacity， and we assume there's plenty of demand。 So my flow rate is。

as usual， minimum between demand， and process capacity。

And this is going to be driven by the capacity， of the bottleneck。 Next up。

I'm going to look at the utilization， and the utilization is the flow rate， which。

is going to be 1/46， divided by the capacity levels。 And so for the first resource， that's。

going to be 1/46， divided by 1/37， and that means 37 divided by 46。 For the second resource。

it is 1/46， again， flow rate， divided by capacity， which is also 1/46。

So the utilization there is 1。 And then for the third employee， it's again 1/46， divided by 1/37。

and that gives me again， 37 divided by 46， which we said is around 80%。 Next。

let's look at the cycle time， which is what we said， was 1 over the flow rate。

And here's why I didn't convert it into hours。 If you write it this way， it's quite simple。

The cycle time is simply 1 over the flow rate， and the flow rate was 1 over 46。

So all that combined gives us the cycle time of 46 seconds， per sandwich。 Every 46 seconds。

life repeats itself。 And then finally， to make 20 units starting， with an empty process。

we have to first produce the first， sandwich。 And for the first sandwich。

it's going to take us 37 plus 46， plus 37， which together gets us 120 seconds。 So after 120 seconds。

we have one sandwich。 So now we need 19 times an additional sandwich。 And guess what？

We're making those on a 46 second cycle。 So we have 120 plus 19 times 46， which。

gets us to the 994 seconds。

So far， with the sub-week， it's remaining repeated， our capacity calculation。

I also introduced the concept of cycle time。 Please be careful with the word cycle time。

I've seen it used very differently across the industry。 Most importantly。

I've come across some companies， that use the word cycle time for what we call flow time。 Now。

not surprisingly， I think my definition is the right one。 The word cycle time， in my view。

captures the idea of a cycle。 Every 46 seconds， a process in our sub-week case， repeats itself。

Every 46 seconds， a new cycle begins。 The cycle time is particularly useful for processes。

with discrete flow units。 I wouldn't use a concept in an oil refinery or brewery。

It just works better when you're dealing with customers， or widgets or sandwiches。

The cycle time helps us find how long it would take， to produce a certain number of units starting。

with an empty system。 As we will see， the cycle time also。

will be at the heart of some further productivity calculations。 Speaking of further calculations。

in the next video， we'll talk more about the importance of a labor productivity。

I will introduce two new measures， and then we will link those to financial performance。

See you then。

[BLANK_AUDIO]。

沃顿商学院《商务基础》｜Business Foundations Specialization｜（中英字幕） - P123：7_劳动生产率.zh_en - GPT中英字幕课程资源 - BV1R34y1c74c

Today I want to talk about labor productivity。

I will introduce two measures of labor productivity， the average labor utilization and the cost。

of direct labor。 I then want to spend some time with you reflecting upon the financial importance of labor productivity。

Let's go back to the Excel model that we had started in the last video。

So we have the three resources here。 Remember we had the processing times， the capacity。

all the way down to the cycle time。 Now imagine you and I would run the separate restaurant。

We are the owners。 We pay the workers。 What would make us really angry？ Exactly。

We have a worker that sits or stands there and is not working。 We say the person is idle。

How much idle time do we observe in this picture here right now？ Again。

it's for the moment assume that we have unlimited demand or a really big number。

of customers asking for our sandwiches。 Where are you going to go see the idle time？

Who is sent around doing nothing？ Well worker number two， the bottleneck is never idle。 In contrast。

the other two workers have some slack relative to the longer processing time， of worker number two。

And that number is simply the cycle time minus the processing time of that station。

So the first person has an idle time of nine seconds。

Again the bottom of the neck will not have any idle time and the third one has again。

nine seconds of an idle time。 The total idle time we get by summing up all the idle times for the individual stations。

and voila that gives us a grand total of 18 seconds。

Now 18 seconds are large number or is it a small number？ That's really hard to interpret。

In absolute terms it's really hard to judge。 So we need to take this number relative to something else。

Let me introduce another measure。 I will refer to this number as the labor content。

The labor content is simply the sum of all the processing times that involve labor。

In our case here we get a labor content of 120 seconds per sandwich。

So these 18 seconds of idle time should be compared to the 120 seconds of making a， sandwich。

In contrast in an autoplanet 18 seconds of total idle time relative to maybe， 10。

000 seconds of labor content for a car would be far less significant。

This is the intuition behind of what I now will really find as the labor utilization。

Labor utilization looks at the labor content。 For every sandwich I'm going to get 120 seconds of labor but I need to pay the workers for。

the 120 seconds at their work plus the 18 seconds they are idle。

It ratios called again the labor utilization。 Don't worry for now about the details we're going to go and review all the definitions。

later on in this video。 So the labor utilization looks at the labor content relative to the idle time。

I can write this slightly differently。 Again， I will summarize this on a PowerPoint slide later on。

For now I look at the labor content and rather than dividing it by the labor content by the。

idle time I can divide it by three times the cycle time。 It shouldn't surprise us。

This gets us the same utilization from a moment ago。

Now let me emphasize what labor utilization measures。 Labor utilization measures balance。

Balance is good but it's not the only thing that matters。 Just as a thought experiment。

imagine I have three very lazy but really smart employees。 What are they going to do？

So they tell each other， "Look， Christian is a kind of stupid manager。

He's going to look at the labor utilization to judge our performance。"。

So why don't we all slow down and everybody has a processing time of 60 seconds？

While what happens to the labor utilization with this evil plan， if all the workers have。

a processing time of 60 seconds there will be no idle time。

The process is perfectly balanced and the labor utilization goes to 100 percent。

Has this process become more efficient？ Clearly not。

So we need another measure here for labor productivity。

Let me propose another measure which I'll refer to as the costs of direct labor。

What's the cost of direct labor？ What we're doing with this measure is we look at the wages per unit of time。

It's assumed for now we have three people here and we pay each of them 12 bucks per。

hour and over that hour we've created ourselves a flow rate of what we said was 78 sandwiches。

So the cost per sandwich in terms of direct labor is 46 cents。

Think about this the next time you're going to a subway restaurant。

You typically tend to pay around 5 to 6 bucks for a footlong sandwich and in there there。

are about 46 cents of labor。 Just as before we have the three resources。

The processing time is about 37， 46 and 37。 So those are the workers， one， two and three。

We found that the cycle time is driven by the second resource 46 seconds per sandwich。

That means the other two resources one and three have some slack relative to the cycle。

time and that slack is going to be their idle time。

So they have 46 minus 37 which is 9 seconds of idle time each and we're going to get a。

grand total of 18 seconds of idle time。 But how do we interpret that number？ Is it big？ Is it small？

We have to take it relative to something and that something we said is going to be the。

labor content。 Labor content， this is some of the processing times， 37 plus 46 plus 37。

the 120 seconds。 So when we're going to judge whether the 18 seconds are larger or small。

we're going to， take them relative to those 120 seconds。 That gives us the labor utilization。

And so one way of computing labor utilization is taking the labor content of 120 seconds。

It is really the green here in my three bars， 120 seconds。

And taking it relative to the total which is a green plus a red。

That measures balance and we found the number to be 86。9%。

So that's one way of computing that number。 There's another way you can do this。

For reasons I don't want to tell you right now， I would prefer you use this other way。

And that is by involving the cycle time。 You see the definition here to the left。

With this definition you're going to go and look at again the labor content， so it's 120， seconds。

And dividing it now by three times the cycle time。

As you see just as before we get 120 divided by 138 and surprise the price， we get the same， 86。9%。

So that's the average labor utilization as a measure of balance。

And finally we talked about the cost of direct labor。

For that you just look at the wages per unit of time， three workers for $12 per hour divided。

by the flow rate which we found to be 78。something which is expressed in sandwiches per hour。

So the per hour cancels out and we're left with the 46 cents per unit。

That gives us the cost of direct labor。 Before we talk more about labor productivity。

let me address a question that I oftentimes， get asked。 Does labor productivity really matter？

So here's some cost data from the order industry compiled by a gentleman， Dan Woodney， at MIT。

and what Dan looked at is the total cost for an OEM in the order industry。

In this case the data was from Daimler。 The total cost you can basically find in the profit and loss statement in the accounting。

of Daimler and from the total cost for Daimler you notice that the vast majority of costs。

sit in purchasing。 About 70% of the total cost is going to be in purchasing and only a relatively small piece。

of the total cost is going to be labor costs。 Most of the costs are coming through our purchasing department。

Daimler pays relatively little money in labor。 So why care about labor productivity？

I would agree that this view is a little bit misleading because keep in mind that the。

70% of costs that Daimler spend from those 70%， those are the total costs that go to their。

suppliers and these suppliers in turn also incur labor costs。

If you factor in the labor that the suppliers like Bosch or Continental spend， if you factor。

those in as well， the total percentage of costs making a vehicle is getting much larger。

They're not on Daimler's book but they're real costs。 Now the suppliers again have their suppliers。

we call those as second tier suppliers， and， those have suppliers of their own。

If we roll this up off of the value chain， you do see that they are a fair bit of costs。

sitting around in labor。 The raw material costs are much smaller than you think。

That has a couple of important implications。 First and foremost pennies matter in a product but oftentimes these pennies will not be on。

your book。 You really have to go out and work through the supply chain and find these costs somewhere。

upstream。 By the way， the same is true for improving your carbon footprint。

With globalization and global trade， suppliers carry more and more responsibility and so tapping。

into their cost base is central。 Even if this is not visible in the P&L statement of the manufacturing company。

those costs matter。 Let's go back to our sub-way example one last time。

This time was an iron profits。 Profit is simply revenue minus cost。

So let's start understanding what the revenues here look like。

Revenue is simply the number of sandwiches that we were making per hour， flow rate， times。

the average price per sandwich。 Let's just assume we're selling these sandwiches for $6 per unit。

Say it takes us $2 in food costs to make a sandwich。

So the first item here that we have on the cost is we're spending the same 78 units per。

hour times $2 per item in material costs。 We also have a labor cost。 We agreed on the labor costs。

We assume that we're paying these workers $12 per hour for all three workers。

We also have some other forms of fixed costs。 And fixes are a whole topic in itself。

We could talk about cost allocation and what drives these costs for long sessions by itself。

whether it's rent， the insurance， the marketing， the overhead and what have you。

Let's just say for now with some plausible cost allocation， we have $250 of fixed costs， per hour。

Remember， most of the business will be over the lunch hour and that is why I allocate so。

much of the fixed costs to that one hour。 This allows us now to compute the profit。

So we have the revenue minus the food cost minus the labor cost minus the fixed cost。

which gets us a profit of about $27 per hour。 So we're making about $27 per hour。

I don't want to comment whether there is a lot or a little。 For now， let's do something else。

Let's think about some improvement strategies， which is something that we'll talk about in。

the next video。 And let's just for the sake of argument imagine we could shave off some seconds of the bottleneck。

This could either be done by automating or by having a more skilled worker or by taking。

some of the 46 seconds and move them to the left and to the right。

Again more on how to find these improvements in the next videos。 For now。

assume this number would go down to 43 seconds。 What's going to happen to profit？

Clearly profit is going to go up。 The question is by how much？ Let's take a look。

We have $27 in the baseline scenario。 Three seconds faster at the bottleneck corresponds to an improvement of about 7% of the processing。

time at the bottleneck and about 2% to 3% in labor content。

Three seconds faster at the bottleneck is translating into a profit increase from $27。

per hour to $48 per hour。 We almost doubled our profitability。 Wow。

Almost doubling profits just because of shaving of a couple of three seconds at the bottleneck。

That is the key insight here to the value of improving productivity。

It's not just about cutting the cost of labor。 Poor productivity creates a real opportunity cost of not selling more sandwiches。

More productivity helps us to increase the top line and serve more customers。

In today's video， we saw that the labor productivity is important for two reasons。

First little things add up。 The example from the auto industry has reminded us is not just about the labor that happens。

in our process。 Our suppliers use labor to generate our inputs。

Our supplier has its own suppliers who also use labor。

When you add up the labor costs to the value chain， the cost numbers become quite sizable。

And second， we learned how labor can act as a constraint on the process that gets us back。

to the bottleneck。 The reason why we want to shave off a second or two at the bottleneck is not to save ourselves。

the associated labor costs。 Rather， we want to do this with the aim of generating some additional capacity。

If we are capacity constrained， more bottleneck capacity is worth a ton of money as it adds。

to top line revenues。 And what is not to like about that？ In the coming video。

we will talk more about process improvement strategies。 Until then。

grab a sandwich and look for the bottleneck。 [BLANK_AUDIO]。

沃顿商学院《商务基础》｜Business Foundations Specialization｜（中英字幕） - P124：8_改进流程的工具.zh_en - GPT中英字幕课程资源 - BV1R34y1c74c

The reason why we analyze processes， the reason why we compute things such as idle time and。

labor utilization levels or the cost of direct labor is of course to improve the process。

We want to make things better。 In today's video I will introduce three tools that will help you improve your process。

I will refer to them as offloading the bottleneck， balancing the line using tech time， and we'll。

talk about specialization。 Let's start with the idea of offloading the bottleneck。 In 1984。

Ellie Goldret wrote a highly successful book entitled "The Goal。"。

Hard to believe that the book is about operations management and its operations management textbook。

written as a novel。 Unlike my books， it has sold millions of copies。

The hero of the book is a fictitious plan manager by the name of Alex， who discovers the。

principles of process analysis。 Alex has his epiphany when he takes a group of boy scouts for hiking the forest。

The group takes a single file path and sets us close to each other。 But the longer they march。

the more the group spreads out。 The boys in the front tend to be the fast ones。

so they walk at their own rate leaving， the others behind。

One shall be scoured by the name of Herbie， is holding the group back。

He has hundreds of meters of empty trail in front of him， and a group of impatient boy， scouts that。

due to the narrowness of the trail， cannot overtake him behind him。

Alex realizes that this hike has a lot in common with his production process。

He associates the speed with which the entire group is moving with the flow rate in the plant。

Just as the plant cannot produce faster than the bottleneck， the group of boy scouts cannot。

walk faster than Herbie。 So any process improvement starts by looking at the bottleneck。

In the case of Herbie， Alex soon found out that this ability to walk as fast as the other。

kids was not just constrained by his theological limitations， but also by Herbie carrying a。

rather large and heavy backpack with plenty of snacks inside。

To improve the hiking speed of the group， Alex then takes Herbie's backpack and puts the。

content up among the other faster scouts。 Yes， those scouts are now slowed down a bit。

but Herbie is able to walk faster， and that， is all that counts。 In general。

we refer to the improvement strategy of moving work away from the bottleneck as。

"offloading the bottleneck"。 Offloading the bottleneck can take mind to perform。 First。

we can think about reassigning activities to other resources with more capacity， an improvement。

strategy that we will refer to in a moment as "line balancing"。 Second。

we can think about automating some of the activities， consuming time at the bottleneck。

by using technology。 For example， we might be able to automate the dispensing of a wrapping paper。

which has， the potential of reducing the processing time for employee number two。 As a side note。

McDonald's key innovations on their way to becoming the fast food empire。

was the invention of a condiment dispenser， which would provide an exact supply of ketchup。

or mayonnaise by the pressing of a button。 Third， we can think about outsourcing some of the activities consuming time at the bottleneck。

Imagine it would be possible to put the condiments in the back at the beginning of the shift。

or to even purchase back that are already loaded with condiments。

This would reduce the time for employee number two， which， given that employee number two。

is the bottleneck， would increase the capacity of the process。 Next。

let's turn to line balancing and tack time。 In our prior calculations。

we found the capacity of the three-person line to be about 78 sandwiches， per hour。

What are you doing when demand is higher？ The busy stores at lunchtime have demands all the way up to 160 sandwiches an hour。

And for that， well， we clearly need more labor。 So I want you to wrestle with two questions while you put me on pause。

The first one is about the cycle time。 What cycle time do you need to hit to serve 160 customers per hour？

And the second one is on staffing。 How many workers will you need？ Okay， put me on pause now。

All right， here we go again。 Let's get this done together。

So what cycle time do you need to meet the amount of 160 sandwiches per hour？ Well。

there are 3600 seconds in an hour， and I want to make myself 160 sandwiches during， that hour。

So the hours cancel out， and I'm going to get 22。5 seconds per sandwich。

It is the cycle time I need to achieve to meet demand。 In a moment。

we're going to refer to this number as the tech time。 That's going to be the tech of demand。

Tech really tells us how fast we have to run， and in what cadence we have to operate to。

be in sync with demand。 That's the whole idea of tech。

Tech also helps us find out how many people we need to hire at the very minimum。

That number is not going to be an exact number， but it gives us a lower bound。

The logic behind this is as follows。 We're going to look here at the 120 seconds of labor that I know are in a sandwich。

That is our labor content， 120 seconds per unit。 And we're going to divide it by the tech time。

the 22。5 seconds per unit。 And so we have seconds per unit in the nominator。

and also seconds per unit in the denominator， so that cancels out。

And we're going to go and get a staffing level of a minimum of 5。33。

If we round this up to the next integer， I see that at the minimum， I need six workers。

It's not going to be exact， but it's the target。 We're going to see in the next slide。

that they actually would need a little bit more， labor than that。

But this number really provides a first-grid shot。 So let me formalize this。

We formally now define tech time as a ratio between the available time and the demand rate。

So available time was in our example， 3600 seconds per hour， and we want to have 160 units per。

hour of demand。 So the tech is 22。5 seconds per unit。

The word "tact" comes from the German word "tacht"， which is used in the musical world。

We want to make sure that not every musician is playing a taser or her own pace， but everybody。

stays in sync。 So I think it's very appropriate to use the word "tact" in the case of process analysis。

Let me emphasize that a "tact" time is an ambition。 It is a target。

It does not depend on the capacity or the capability of the process。

It's just turning demand into a target cycle time for the process。

Once we have done the "tact" time， we can compute the target manpower。 The target manpower。

as we saw， is the ratio between the labor content and the tactime。 In our case， it was 5。33 person。

which was 120 seconds of labor content， divided by 22。5， seconds of the tact per unit。 So ideally。

we would like to hire 5。33 persons。 Since you cannot hire a third of a worker。

this is clearly not feasible， and we were rounding， this up to 6。 As we were seeing in a moment。

even 6 might not be enough。 The reason for that is that work cannot easily be divided into smaller chunks。

So for example， the cash register operation here， that takes 20 seconds。

It is a block or a chunk that we cannot move easily， force and pack。

And so you notice when you're allocating the activities across the resources here， the， employees。

even the bottleneck has some slack， has some idle time relative to the tactime。 Again。

line balancing means that you take the tactime 22。5 seconds in our example， and。

then you spread out the activities from greeting the customer all the way to the cash register。

and you take to the constraint that no individual should have more work and their processing。

time than the tactime。 If you notice you're going above the tactime， well。

then you just move the work to the next， resource and continue until you found out that in this case。

we'll actually need 7 employees。 Now let me point out some subtlety that I find somewhat counterintuitive。

It's counterintuitive， at least for me， not sure about you。 So here because of the demand rate。

we said that even the bottleneck would have some idle， time。 Let's ignore that。

Let's make us capacity constraint and assume that we are running at the rate of the bottleneck。

which is producing a unit every 21 seconds。 We can then compute how much idle time each of the workers has。

It's going to be 5 seconds for the first worker， 0 for the second worker， 2 seconds here for。

the next one， 7 seconds， 5 seconds， 7 seconds and 1 second。 Now how much idle time do we now have？

So we have 5 plus 0 plus 2 plus 7 plus 5 plus 7 plus 1。 We have 27 seconds per unit of idle time。

Do you remember how much idle time we had with 3 workers？ It was 18。 Well， less than the 27。

So as we specialize， that idle time is going to go up。 The more workers take on the work。

the harder will it be to find a perfect balance。 As I said， I find this counter-intuitive。

Today we talked about improvement strategies。 We talked about offloading the bottleneck。

We talked about balancing the line using the concept of tech time。

We have not yet talked about specialization。 So here we go。 Adam Smith。

the famous philosopher and economist， wrote about the benefits of specialization already。

in the 18th century。 First looked at a pin factory。

You see a picture of a pin factory here and a quote from Adam Smith that nicely captures。

the concept of specialization。 Now it is a stretch to move from a pin factory to a subway restaurant。

but let me try。 What benefits might we get in our subway restaurant as a result of specialization？

I propose that we would benefit in at least four ways。 First。

think of a reduction in processing times due to elimination of setups。

The times of the various activities might be reduced and thus the capacity levels increased。

as workers specialized in fewer activities。 For example。

an employee who is doing all the activities might have to take off his gloves。

when he switches between touching a sandwich and operating the cash register， which would。

increase his or her processing time。 A person who does nothing but operate the cash register would not face such an inconvenience。

Oftentimes such disruptions are referred to as setup times。 Second。

I can imagine a reduction in processing times due to learning。

If you cut a thousand tomatoes per day， chances are you'll get pretty good at that。 Simply put。

practice makes perfect and this perfection also manifests itself in shorter， processing times。

Third， we can potentially get away with lower skilled labor。

More specialized labor tends to require a shorter training period and oftentimes receives。

lower wages。 It takes less time to teach an employee how to assemble a review mirror compared to teaching。

her how to assemble an entire car。 So even if specialization does not decrease the processing times。

it can still save us， money。 Finally， I think about the need for equipment replication。

Another benefit of the specialized process flow relates to the amount of equipment that。

is needed to operate the process。 Six parallelization in which each employee carries out all the activities would require。

that each employee would get her own equipment， including a cash register， which is rather。

expensive。 Again all of these benefits have to be weighed against the cost of specialization。

As we specialize， the idle time is going up， which at least to me was somewhat counterintuitive。

Moreover the coordination needs and the interdependence among the workers are increasing。

If I have only one person who can slice tomatoes and that person does not show up for the job。

I have a problem。 In contrast， if each employee is capable of making a whole sandwich from beginning to。

end， my process is much more robust。 This might not be a big deal in the production of a sandwich。

but it is a super hot topic， in producing cars and phones in today's global and interdependence supply chains。

Okay， enough about the 18th century。 In the next video we will talk about flying to Mars。

It should be a nice contrast to the regression in history。 See you then。 [BLANK_AUDIO]。

沃顿商学院《商务基础》｜Business Foundations Specialization｜（中英字幕） - P125：9_学习回顾和实践问题.zh_en - GPT中英字幕课程资源 - BV1R34y1c74c

In the last video， I promised you will be flying to Mars today。

I will keep my word， but first， I want to annoy you with some rather earthly calculations。

I want to first repeat the definitions that we have covered until now in this module。

In the set of videos we talked about the labor content， the cycle time， the idle time。

the labor utilization， and the cost of direct labor。

We also talked about the idea of line balancing。 Now that we have arrived at the end of this module。

I want you to be comfortable computing the cycle time。

finding the labor content and the labor utilization， and computing the cost of direct labor。

You should also be able to determine how long it takes to produce X units starting with an empty system。

And finally， you should now be able to find tech time and the target manpower。

as well to think about some process improvement strategies。

Down here you see the key formulas that you will need to remember from this module。

The problem I have chosen for this module is Mr。 Kears， here's someone。 Mr。

K is a problem that's somewhat similar to the sub-work case in the sense that there are multiple activities that are done by a set of workers。

Let me emphasize that there are five activities。 There are five activities， not five workers。

There are three workers and five activities， and the five activities are spread out。

as you can see in the question。 As usual， I want you to put me on hold while you're tackling the problem by yourself first。

So pause now。 Alright， here we go again。 We start by looking at the bottleneck。 And for that。

we draw a simple process flow diagram。 So we have a box for the first worker or a triangle。

a box for the second worker， another triangle， and a box for the third worker。

So then it's our process flow diagram。 And then for the bottleneck。

we need to find the capacity level， and that requires computing the processing times first。

The processing times we are at 10 minutes per guess at the first， 20 at the second。

and 30 minutes per guess at the third resource。 And to get to the capacity level。

what we do is we're going to take M divided by the processing time。

And since it's a one worker per station， it's simply one divided by the processing time。

So it is one divided by 10 as a capacity for the first resource， one by 20， one by 30。 Okay。

And this is now expressed in guess per minute。 Alright。

that makes it clear that the third resource is going to be the bottleneck。

and that answers our question。 Next， let's look at average labor utilization。

I mentioned in the module two ways of finding the average labor utilization。

And so let me illustrate how both of them would work。

The labor utilization we said is labor content， which is a sum of the processing time。

divided by three times the cycle time。 Okay， so what does this mean？

The labor content is 10 plus 20 plus 30， the sum of the processing time。

and that's 60 minutes per unit。 And we're going to be on a 30 minute cycle。

so three times 30 is 90 minutes， which gets us a 66% labor utilization。

Now the other way that I mentioned in the video， we could compute the same thing by looking at the labor content。

and then divide it by the labor content plus the idle time。 Now again。

we're on a 30 minute cycle here， and so that means that my first resource has a 20 minutes of an idle time。

processing time is 10 minutes， and the cycle time is 30， so 20 minutes of an idle time。

The second one has 10 minutes of idle time， and the third one is zero。

So that means we have a total idle time of 30 minutes。

So 60 divided by 60 plus 30 is going to be my labor utilization。 Same result。

And then finally at the wage rate of $20 per hour， what is the cost of direct labor per customer？

For that， we have to look at three people， each make 20 bucks per hour， so it's $60 per hour。

and we're going to divide it by the floor rate。 We're serving two guests per hour。

that's our floor rate。 And so the per hour is going to cancel out。

and we're going to get a cost of direct labor that is $30 per guest。

Here's the next batch of questions。 How long would it take the process to serve 10 customers starting out empty？

Well， let's think about that。 How long would it take to serve one customer？ Well。

that's going to be 10 minutes at the first resource， the first employee。

20 minutes at the second employee， 30 minutes at the third employee。

so after 60 minutes we're going to have our first customer out。

From then onwards we need to serve nine more， and those nine are going to come out at the rate of the cycle time。

which is 30。 So 60 plus nine times 30， and that gives us 330 minutes。 All right。

how could this process be improved by offloading the bottling？ Well。

that question kind of gives the way the answer， so we have to move work away from the bottleneck。

And we saw the bottleneck was station number three， right？

And so when you look at the activities here， station number three is doing the hair。

but also spending five minutes here on the checkout。

And if you could move these five minutes to somebody who has less work to do。

and the key suspect I guess here is going to be station number one。

you would offload the bottleneck。 So in short， what you would be doing here is you would move five minutes of work from station number three。

the third employee， which is currently the bottleneck， and this is going to be super busy。

You're going to take these five minutes to station number one who really has nothing to do。

That doesn't always work because it might be precedence relationship。

It's a good move you should consider。 All right， what's the tack time that we need to achieve if we want to have 40 customers per day served？

Remember， when we look at tack time， we look at the available times of 10 hours or 600 minutes per day。

and we're going to divide it by the demand rate， which we said is a target demand。

and we said that this year 40 customers per day。 So the per day cancels out。

and we have 600 by 40 that is 15 minutes between customers。 That's going to be our tack。

We're going to be serving a customer every 15 minutes。

And the tack time intern allows us to compute the target manpower。 For the target manpower。

remember we have to look at the labor content， and the labor content divided by the tactics simply going to give us a target manpower。

So again， labor content， we already computed there was 60 minutes per customer。

and we're going to divide it by the 15 minutes per customer that we just computed as a tack。

So that means at the very minimum， if everything goes super smoothly and there's no idle time。

we're going to need a total of four employees。 My philosophy， the observation of today。

relates to the difference between project management and process management。

Many of you likely have experience with managing a project。 You define activities， set an end goal。

and you look for something known as a critical path。 Recently。

I heard an interesting quote from Elon Musk。 Think of Elon Musk whatever you like。

but Tesla and SpaceX are arguably remarkable businesses。

Elon was talking a bit about his favorite topic， which is flying to Mars。

He said flying to Mars is more of a D-day operation than it is an Apollo mission。

Let me repeat this。 Flying to Mars is more of a D-day operation than it is an Apollo mission。

What did he mean？ The Apollo was an amazing project with a single mission and a historical success。

On July 20， 1969， NASA landed the Apollo 11 and its astronauts on the moon。

Personally， I don't remember it。 I was exactly 100 days old when that happened。

Land a man on the moon。 Though ultimately， multiple people walked on the surface of the moon。

the entire management focus of the Apollo mission was on landing one man on the moon。

The mission of D-day and the invasion of Normandy by the Western Allied forces was different。

It was not about landing one boat in Normandy。 On June 6， 1944， the Americans。

British and the French began sending over 150，000 soldiers to Normandy coastline。

including the famous Utah Beach， Omaha Beach， Juneo Beach， and Swart Beach。

I like this picture here of American airplane production before the invasion。

The focus of management was not to build a boat or build a plane。 It was building many of them。

Process management ultimately only makes sense if you are repeating something。

Elon Musk doesn't want to build one rocket to fly to Mars。

He intends and in fact has even started to build many starships。

His dream is to establish a colony on Mars。 For management problems， so you only do something once。

you should take a course on project management。 If however you want to do something multiple times。

you have to think about process flows。 For most of us， I propose。

planning a wedding is a project you plan for one marriage。

The wedding planner in contrast thinks that weddings is a process。 Other examples。

The more units flow through the process， the easier would it be to use process analysis as a tool。

When in the late 1800s， Gottlieb Diamond built the first car。

he at first did think about bottlenecks， utilization or inventory。

His focus was on managing one project which was building a beautiful car。

When three decades later Henry Ford mass produced cars。

he would make a vehicle roughly every minute。

For Ford， it was all about the process。 Similarly in healthcare。

big emergency rooms see between 100 and 200 patients per day。

Even something medically challenging as a bypass surgery is not thought of as a process。

In contrast when Dr。 Daniel Williams preferred to perform the first heart surgery。

he most likely thought of it as a project。 So you see that when work gets repeated。

when we move from the first units to the end unit。

we need to change our focus on project management to process management。

And that is really what this course is all about。 [BLANK_AUDIO]。

沃顿商学院《商务基础》｜Business Foundations Specialization｜（中英字幕） - P126：10_浪费的DNA.zh_en - GPT中英字幕课程资源 - BV1R34y1c74c

Welcome to the module on lean operations。 The goal of this module is to help you spot。

waste in the processes that you analyze。 Though each and every process is somewhat unique。

we will see that there is a common pattern when we look for inefficiencies。 What I want。

you to do in this module is really learn how to figure out the DNA of waste。 The module。

is relatively light on math， but please don't hold this against the idea of lean。 And we。

also emphasize that the idea of lean operation is really present throughout all modules in。

this course rather than being squeezed into this single module。 As often， let me kick。

off this module with a personal story。 Now a little while ago， I needed to get an MRI。

with a contrast agent， which is a relatively simple radiology procedure。 And so I got。

an appointment here at Penn Medicine's radiology department for 10 o'clock， 10 a。m。 She had an MRI。

I arrived at the hospital at 9 45。 I arrived at the radiology， I gave。

them my name and they conferred my appointment time。 And then to my pleasant surprise， similar。

to what you see at some restaurants where you are checking in to wait for a table， the。

hospital gave me a little buzzer。 I said to myself， "That's kind of cool。" They really。

thought about the waiting time experience here。 So I took my buzzer and I sat down。 To。

my even bigger delight， about 15 minutes later， at 10 or 5， the buzzer started buzzing。 That。

was fast， I thought。 Unfortunately， it was not my turn to be examined as scheduled with。

my radiology appointment。 Instead， I was now allowed to really check in。 So I checked。

in at the different check-in booths at 10 10， go back and wait it。 Now at 10 20， also， not too bad。

eventually the nurse called me。 I was directed to the changing room over。

here in this part of the hospital。 I changed and I waited。 I was stuck in a room with a。

television that couldn't be turned off。 Finally， at 11 o'clock， I was called by the。

nurse and she directed me to the imaging room。 I waited in the imaging room for about another。

20 minutes until the radiologist showed up。 For a few minutes， he explained me what he。

was about to do with me。 We had the procedure which took about 10 minutes。 I was directed。

back to the changing room and instructed to wait to make sure that the images turned out， alright。

Then I could ultimately change back into a normal close and about two hours later， at 12 o'clock。

I was out of there。 Let me ask you this。 Where did you see inefficiencies。

during my morning as I journeyed through the hospital？ We'll get back to some inefficiencies。

of this story later on in this module。 For now， I wanted to talk about two grandmasters。

of operations management。 Let me start with Fredrik Wences of Taylor。 Taylor wrote a famous。

book "The Principles of Scientific Management"。 He wrote this book well over 100 years ago。

In his studies of workers， Taylor believed in careful and systematic observation。 Oftentimes。

he was able to obtain dramatic improvements in productivity， typically it's a waste reduction。

picking the right man and tools for the job and setting the right incentives。 For the。

better or for the worse today the word "tailarism" is unfortunately used more as a curse word。

One of the reasons for Taylor's decline in popularity is that Taylor had a rather difficult。

relationship when it came to human beings。 He was interested in their muscles， not in， their brains。

But please allow me to nevertheless share two quotes with you from the principles。

of scientific management。 So first Taylor wrote， "We can see and feel the waste of material。

and things。 Awkward， inefficient or ill-directed movements of man however， give nothing visible。

or tangible behind。" And then the other quote I wanted to share is kind of weird but I really。

like this one。 Employers derive their knowledge of how much of a given class of work can be。

done in a day from either their own experience which has frequently grown hazy with age， from。

casual and unsystematic observation of their man or at best from records。 I really like。

these quotes。 I think they are as important now as they were 100 years ago。 At the heart。

of this modulus the idea that we need to measure the amount of work that is done by。

our resources and that measurement is hard。 It is really hard。 The data is not easily available。

but the fact that it is hard shouldn't keep us from doing it。 The second gentleman I would。

like to quote is Thijeone。 Thijeone was really seen as a father of the Toyota production system。

So two owner quotes。 First owner wrote， "Moving is not working。 The Toyota style is not to。

create results by working hard。" And then the second quote reads， "All we are doing。

is we are looking at the timeline from the moment the customer gives us an order to the。

point we collect the cash。" As I think back to my radiology experience。

then begin with the second owner quote。 Here is my timeline from the moment I entered the。

hospital to the moment I left the hospital。 You can see that only a small fraction of。

that time had really to do anything with medical imaging。 I needed that day about 20 minutes。

of care but I spent two hours and 15 minutes in the hospital。 So that was really a waste。

of my time。 Now maybe the hospital design or process was the objective of maximizing hospital。

efficiency rather than optimizing for the patient experience after all hospitals are under significant。

amount of cost pressure。 So maybe they did a good job on efficiency and not on my willingness。

to pay。 So let's think about the hospital efficiency。 How about the person who gave me， the buzzer？

Was that person adding value？ A massive waiting room in the lobby？ Was that。

needed for improving my patient experience？ How about the person who checked me in and。

confirmed my paperwork？ Was he or she adding value？ So you see where I'm going with this。

As an example of hospital resources， let's look at space。 Space is really expensive for。

hospital operations。 Hospitals are really space constrained and have to pay for expensive。

real estate。 Since I had consulted for this hospital at prior times， I asked the management。

to share a floor plan of the radiology unit with me。 Now let me ask you this， which space。

on this floor plan adds value。 Right。 At the end of the day， the value is generated in the。

imaging room。 Given that I was waiting for 20 minutes in the imaging room， that room and。

that expensive equipment in there were clearly not running at a high level of utilization。

Now given that patients don't spend 20 minutes in the department， but over two hours， we have。

to put these patients somewhere。 And what are patients sitting around？ Exactly。 They're， inventory。

And an inventory of patients requires space。 And what do you need when you have elderly。

man like me sit around waiting for an hour， especially when you think of a water fountain， nearby？

Let's not go there。 As you can see from this example， hospitals， the hospitals。

that are really short in space， but they really primarily consist out of lobbies， waiting， rooms。

restrooms， utility rooms and so on。 All of this space exists for a good reason。 Don't， get me wrong。

But most of this space doesn't add value to the patient。 This gets me back。

to my process analysis module。 Remember how I introduced two perspectives to the process。

in an operation？ We can take the floor units perspective in the case that would be the。

patient and look for where the patient wastes the time。 This was really the idea of owner。

talking about the timeline。 We can also take the perspective of the resource。 It is the。

employee's real estate， the imaging machine。 That is a much more tailor-risk perspective。

The waste of my time or the waste of the time of the resources here， note that none of this。

really has left a record behind。 This is why I like the tailor quotes so much。 Awkward。

inefficient or indirect movements of man leave nothing visible or tangible behind。 So one。

of the key things we need to do in this module is to learn how to carefully observe how work。

is done。 See you in the next video。 [BLANK_AUDIO]。

沃顿商学院《商务基础》｜Business Foundations Specialization｜（中英字幕） - P127：11_浪费资源的时间.zh_en - GPT中英字幕课程资源 - BV1R34y1c74c

As I mentioned in the opening video， we need to be mindful of the time that is wasted。

at resources， as well as the time it is wasted by our customers， by our flow unit。

Today we will start looking at the time wasted at the resource。 Again。

I would like to start this video with a medical story。

A number of years ago my colleague Evan Feudson and I did a study in the ICU at the Children's。

Hospital of Philadelphia。 My research question was simple。

We wanted to know what fraction of time an ICU bed was used to provide the care it can。

and should provide and how much of the time was wasted。

As we were launching our study I felt reminded of Taylor's words， "waste of time， leave。

nothing tangible behind。"， How true？ Where would we go and find the data of wasted time？

Like in any modern operation there was no shortage of data。 Medical records， billing records。

hospital IT systems， you name it。 But where would we find how much time was wasted？

In our study of the ICU capacity at shop， here's what Evan and I did。

We found a motivated student and that student would go to the ICU and go around the beds。

every hour， over and over again。 It's a really academic approach and I feel somewhat guilty of asking our students to。

spend so much time。 But I still think that this was the only way to get the job done。 Anyway。

this approach allowed us to collect thousands of hours of our observations on。

ICU beds and how they were used。 We then classified these hours as follows。

Here you see the hours that we have data for， the hours the students observed。

We then looked at the hours that from a medical perspective were not really needed。

That included the hours the bed was staffed， meaning it had a nurse assigned to it， but the。

bed was empty。 Hours the kid in the bed was waiting for a transfer and the hours that the bed was waiting。

for cleaning or an incoming kid。 So in this case， we see that an ICU bed at shop at that point was basically used about。

82% of the time for the kid that it was designed to deliver。 That by the way is a really。

really large number。 I show you the state of the two reasons。 First。

to introduce a new tool and then to discuss the limitations of the tool。

Now this chart illustrates a method of really rigorously collecting data and analyzing。

value at time。 You map out the total time that you have available at the resource and then you subtract the time。

that you're comfortable defining as waste。 We talk more about what exactly constitutes waste in the next video。

For now， waste is the time that your feed is not adding value。

It's not helping to get the kid healthy。 We then can define the overall equipment effectiveness or OEE for short。

offer resource as a percentage， of time in which the available time of the resource is used productively。

either resource， is not waste in capacity。 We can do this for any type of a resource。

not just for equipment。 You could do this for a machine， a factory worker， a lawyer or an airplane。

Second， let me also be upfront about the limitations of the tool。 Again。

let's go back to my ICU example。 When we collected the data。

we told the student that a kid in the bed that is on a ventilator， always should be in the ICU。

So any kid was observed in our study that was unevent was coded as value at time。

I think this was a plausible approach， but again， not without limitations。 Imagine， for example。

a kid being on the vent， but the vent really should have been disconnected， a couple of hours ago。

But the care team was busy taking care of another kid and so the vent never got disconnected。

or got disconnected six hours later。 And so you really have to ask yourself。

are these six hours value at time or are they， not？ That gets me to the owner quote。

"moving is not working。"， I think this one is really deep。 Moving is not working。

That means that even the dumbest consultant can look at a worker who is standing around。

idle and realize that that worker is not adding value。 To look at a busy worker， however。

somebody who is working， somebody who is moving and。

determine how much of that work is really adding value， that is a lot harder。

Now according to owner， we can break up the time of workers that the workers are at work。

into two buckets， work and waste。 Owner defines waste as needless waste of time and worker movements that should be eliminated。

immediately。 Again， more on this will come in the next video but for now， it's wasteful。

Worker time minus waste gives us work。 But really not all workers adding value。

Owner defines non-value at work as a work that was needed but it's not directly adding。

value to the customer。 Keep in mind this framework originates from a production setting but I think it's really。

quite universal。 Every process has a custom so we can say that non-value at work in a processes work that。

doesn't directly contribute to the willingness to pay from our customer。

Such work might be needed but not adding value to the customer is oftentimes referred to。

as incidental work。 For example， the person that was checking me in during my radiology visit and confirming。

that I would be indeed the person who was on the prescription， well that was not outright。

wasteful but we might imagine other ways to confirm my identity。

And clearly from my willingness to pay perspective， that check-in didn't improve my health， my。

happiness or my health outcome。 All right， back to OEE。

An OEE of 50% means that you can do twice as much work as you do right now with the same。

amount of resources。 Or you can cut your resources in half and still serve the same number of customers。

Now if your capacity constrained， these productivity improvements will really add to top line。

revenues。 And if your demand constrained， they really guide you to the way to save an operating。

cost。 To give you a non-health care example of this， consider the OEE of an airplane。

Airplanes are really expensive pieces of capital equipment as you want to get as much。

airtime out of an airplane as possible。 Here's an OEE chart for American Airlines Delos Fort Worth。

Every good consulting presentation that I've seen in operation says some version of this。

slide somewhere in the deck。 So if you look at the 24 hours that are in a day。

the plane is about one-third of the， time in the year。 It's pretty impressive by the way。

and it's certainly much higher than what it was 20， years ago in the airline industry。

Not to mention the situation and the summer of 2020 with the collapse of air travel that。

we've seen during COVID。 Now why is it？ What happens during the other time？ So boarding。

de-plaining， maintenance， all of this stuff is clearly important， but it's。

not directly adding value。 It is the idea of incidental activities。

Personally when I ran these numbers I was more surprised by the taxi time， but really。

more than an hour and a half per day that $100 million airplane is basically acting as。

a bus。 All of us are busy， which means we typically call for more resources。 We want more staff。

we want more equipment， we want more office space， more and more and， more。

Now the OEE framework reminds us that our first job in process improvement is to get。

more out of our existing resources。 And OEE of 50% means that we can double our capacity without adding more resources。

The OEE framework is primarily there to help you organize data and the data collection。

from the front line。 It tells you what to look for as you're looking for process improvement opportunities。

Like any framework the framework is not right or wrong， it is up to you and your data collection。

to decide what you make with it。 But it supports the spirit of Genshekin-Buzu。

Go and see for yourself。 In the next video we'll talk about what to look for when you are at the front line looking。

for ways。 See you then。 [BLANK_AUDIO]。

沃顿商学院《商务基础》｜Business Foundations Specialization｜（中英字幕） - P128：12_7大浪费.zh_en - GPT中英字幕课程资源 - BV1R34y1c74c

The OE framework has helped us to look at a resource and separate a waste from value at。

time。 In this session， it's talking a bit more about what waste looks like。 According to Atei Giono。

there exists seven sources of waste。 Already introduced the key idea in the last session。

moving is not working。 So as we look for waste， it is not just the idle resource that provides the opportunity。

Beeziness per se is not a virtue。 A piece of equipment might be highly utilized or an employee might always be overworked。

but that is no guarantee that the process or the employee is adding value。

Let's start by going back to the Taylor quote。 Remember we talked about awkward， inefficient or。

indirect movements of man？ While your awkward， inefficient and。

indirect operations professor will not define seven sources of waste。

The first way I want to define is wasted motion。 It's unnecessary movement of people during their work。

I just changed the tire of my bike this morning。 It took me probably twice as long as it takes the skilled bike mechanic down the street from。

here。 And so that is a form of wasted motion。 The same is really true for a chef。

I love watching chefs in the kitchen。 Every motion adds value， there's no fumbling。

there's no pausing。 It's just perfection。 That's just not me。 The second way relates to defects。

Defects are obviously not adding value。 They're really annoying for the customer if they remain undetected。

And if you detect the defect within the boundaries of your process， they require you fixing them。

something that is known as rework。 So defects are a second source of waste。

Over processing is a third source of waste， which relates to producing a higher quality。

than what is needed or demanded by the customer。 It's like serving a fancy meal to your relative who only eats cheeseburgers or maybe with。

my German family only loves part worst。 So now we are up to three。 Motion。

over processing and defense。 All of those three can be found by looking at one resource in isolation。

To see the other four sources of waste， we really have to move beyond one worker performing。

one task and look at the process and many workers in a line。 Just think back to the sub-break case。

the three persons making a sandwich every 46 seconds。 As you switch to looking at the process now。

we typically encounter idle time。 Some resources by definition have more capacity than the bottleneck。

and so they will incur， idle time。 Now， all we refer to idle time is waiting as the workers really waiting for the floor。

unit， the part。 Personally， I like idle time better because I don't want to confuse it with customers waiting。

for a resource。 But again， all we call it waiting and to remind to criticize the God of operations。

Next there's inventory。 If the floor is not well coordinated。

some of the resources will produce faster than the， others。 In particular。

it's going to be the bottleneck， it's going to be the slowest step， and we're。

going to end up with inventory in the process and typically in front of the bottleneck。

That is our fifth source of waste。 Inventory also happens when we produce more than what is demanded by the customer。

We go back to our discussion of tech time and its role of keeping us in sync with demand。

Uno named this overproduction and he made it the sixth source of waste。 And finally。

as we have multiple resources in the process， we have to move the floor。

unit from one resource to the other， which is known as transportation。 So here they are。

seven sources of waste。 Again， more to come in future videos。

but I wanted to introduce a framework first。 Let me make two observations。 First。

how do you remember these seven sources of waste？ They are really multiple memory tricks out there。

but the one I like the best is Tim Wood。 Tim Wood simply combines transport， inventory， motion。

waiting， overprocessing， overproduction， and defects。 I know it's somewhat stupid。

but at least for me， it has always worked。 And then second。

the operations community now increasingly refers to an eight source， of waste。

and that's a waste of intellect， which means not using the ideas of those working， in the process。

That is a wasted opportunity and thus an eight source of waste。

We talk more about involving the workers and future discussions on quality。 At this point。

think back to the comment I made on tailorism at the beginning of this， module。

In the 20 years that I've been doing consulting work and research collaborations with companies。

I've always been amazed how universally applicable the seven sources of waste end up being。 Yes。

waste really might look different in each industry。 Yes。

this framework has its root in manufacturing。 It also works well in other settings。

including healthcare。 I show you some examples here on this slide。

I find it almost shocking how directly the seven sources of waste apply to healthcare。

and similar settings。 Rework in healthcare is called readmissions。

Hospitals have a whole department called patient transports， rooms， equipment and providers。

are at times idle as we discussed at the beginning of the module。

And don't get me started on overproduction， the provision of unnecessary procedures by。

the providers。 Anyway， I'm sure you have seen these things in one form or the other in your own experience。

Now spotting waste really requires careful observation。 In the past。

this has mostly been done by stop watching people。 For example。

I've done some work myself using a method called time and motion studies。

What I did is I video recorded primary care providers and then I would dissect the encounters。

that they had with their patients， second by second with the help of some medical experts。

and other doctors to look for waste。 This would give us a sense on how doctors are spending their time with patients。

It was fascinating to me to see how little time was spent on actual examinations。

Such data collection is obviously very tedious。 However。

when you show up with either a video camera or a stopwatch to collect data， chances。

are that you're not really getting representative data from the operations。

Let me make two observations on how such time and motion studies are changing as we're moving。

along in history。 First， most service operations these days are really digitized。

And so oftentimes you can get timestamp data from the operations， particularly call centers。

claim processing， customer support or reading x-rays。

And all these processes are something that I refer to as digital exhaust。

Such processes store when an employee touches a record and how long they work on it。

That allows you to compute the processing times even years after the work has happened。

Taylor's claim of movement leaving nothing tangible behind is simply no longer true in。

our modern digital world。 Second， we now really have technology to use image processing and AI to do this observation。

for us。 To help illustrate， I've been working with a startup run by two pen grads。

And these guys have developed machine vision technology that helps manufacturing firms。

to measure processing times， spot waste and detect quality problems via video cameras。

that are installed at the assembly line。 What you see here is the assembly line of pride mobility。

Pride is making high-end wheelchair。 A business also run by a warden grad。

And the video footage here on this slide shows you line segment 170， which consists out of。

some manual assembly operation。 And you notice that the image recognition of the software is tracking the flow unit。

which is the wheelchair。 And it's also tracking the resource。

which is the gentleman here in the white t-shirt。 And then the software is smart enough to spot when the worker is idle or the worker is spending。

more time on the flow unit as usual。 All this data is then fed into a reporting system that spits out the report that looked。

like this。 Now， I'm sure you've seen this in your own life。

This is really more and more 1984 George Orwell than is emerging in our society。

But it does matter if you drive a modern car， serve the internet or go for a walk with your， phone。

And it's also said we now do， can be tracked and leaves a data trail behind。

This is certainly worthy of a longer discussion about privacy and individual freedoms。

If you're interested in this topic， I make sure to provide a link to some work I've。

been doing in this space。 I do however think that these various tracking technologies will and should play an increasing。

role in operations management。 In the previous video。

we defined the OEE as a percentage of time that the resource is， doing what it is designed to do。

And now， waste was an abstract enemy that was somehow reducing our OEE。

The seven sources of waste framework helps us to see how exactly we waste capacity。 Tim Wood。

in other words， is the main enemy to a lean operations。 In the next video。

I want to switch the perspective and talk about the view of the flow unit， be。

it a customer service process or widget in a production process。 See you then。 [BLANK_AUDIO]。

沃顿商学院《商务基础》｜Business Foundations Specialization｜（中英字幕） - P129：13_流失的流动时间.zh_en - GPT中英字幕课程资源 - BV1R34y1c74c

Unfortunately， many experts and consultants focus entirely on the OEE metric we discussed。

in the last session。 To them， lean operations is all about making the resources more protective。

And how do you make a resource like a doctor more productive？ Well。

since either time is one of the seven sources of waste， having a waiting room packed。

with enough patients to keep our dark busy for an entire day seems like a good idea。

We talked about customer waiting times in another module of the scores。 For now。

let me just point out this is not lean at all。 Remember that inventory is also a form of waste and wasting customers are inventory。

As I mentioned in the context of my radiology experience at the beginning of the module。

lean is equally about the customer。 Customers waiting and customers running around are also forms of waste。

And so in this session we introduce some tools and measures to capture this。

We're not changing the perspective from the resource to the perspective of the flow unit。

When owner looked at manufacturing flows， he said he postulated products have to flow， like water。

Products have to flow like water。 Well， why wouldn't they？

Water keeps products from smoothly flowing through the production process。

Take a look at the three-step production process here。 Parts here are taken out of inventory。

then they are produced in a casting operation。 Once all the parts are done with casting。

a transporter moves them over to the machining， operations。

They have to wait it again until the machine is completed。 The previous production cycle， again。

it works and then comes and moves the output over to， the assembly station。 Of course。

the other parts are not ready yet in the batch and so they have to wait。

And so the assembly only starts when all the parts are there。

And so our part spends even more time waiting。 Finally。

the parts once they are assembled to get shipped。 Now we can define the flow time efficiency。

also known as the percentage value at time， as the ratio between the value at time and the flow time。

So in the three-step process you'll notice that only a small fraction of the flow time。

is really adding value。 Most of the flow time， the flow unit sits around waiting to move forward in its journey。

In my radiology example， at the beginning of the module， the ratio was 20 minutes of care。

spread out over more than two hours of flow time。 But here's another observation about customer experience。

Personally， I would argue that it's not just about the time investment of the customer。

but it's also about customer friction， customer effort and pain points。 Personally。

I'd much rather spend 10 hours at home waiting for an email that takes my。

doctor five minutes to write rather than spending one hour in the waiting room for a 10 minute。

discussion。 Now you see that the flow time efficiency is superior in the second scenario。

but I prefer， the first。 There's this observation behind a tool that is known as a service blueprint。

But the service blueprint does， it takes all the activities related to one or multiple customer。

episodes。 And then the service blueprint reorganizes them along two dimensions。

From the left to the right we have time。 We start with the early activities on the left and work our way to the later activities。

on the right， just like with the previous production timeline。

But unlike the production timeline I showed with the casting operation， we now also have。

a vertical dimension。 The vertical dimension distinguishes where the customer service is delivered。

On top we have activities that happen at the home of the customer。

Really from a customer perspective， this type of work is preferable。 From my customer's perspective。

it's generally preferable to deal with work from home compared。

to work that happened on what's called on stage。 On stage stands for work that happens at the retail branch in the hospital or inside the。

factory。 But also from the firm's perspective， caring out work on stage tends to be expensive。

You visually incur real estate costs and the firm also has to have the resources ready。

exactly when the customer walks in。 So wherever we can， it's nice to move work away from the stage。

You want to move work off stage。 It's just like checking into your fly from home as opposed to doing this at the gate。

Just think about my radiology example one more time。 Verify my insurance status。

This is work that could have been done before I showed up in the hospital。

More convenient for me and cheaper for the hospital。 Beyond having the work happen at home。

we can also think about another way of moving it， off stage。

Imagine we as a firm could do the work for the customer。 I mean。

you think about buying stuff on Amazon。 You won't enter any billing or shipping information anymore when you buy something on Amazon。

You know that Amazon has stored everything and remembers it over time so that reduces your。

work the next time you're going to buy something。 Why can't my hospital do this？

Why do I have to fill out a four-page survey in the waiting room？

Every dime I see my doctor only then to be asked exactly the same questions on the survey。

when I'm talking to the doctor in the room。 To be fair。

this has gotten a lot better over the last years， but you get my point。

So another way of improving the customer experience is through automation。

We move effort away from the customer by automating some of the activities。 In my latest book。

"Connected Strategy，" they are wrote with my colleague， Nikolai Sigokow。

We discuss how modern technology can create seamless and magical user experiences that。

do not always require the customer to take action。 Flow units are impatient。

and so we should get them through our process as fast as we， can。

That is due for a crankshaft and an autoplant that should not sit in inventory for a week。

but rather should be turned over many times during the week or even multiple times during。

the day in order to reduce the inventory cost。 But it's also true for a service organization。

where the inventory consists of waiting customers。 And clearly， customers want to move as a whole。

so they will be sitting around and waiting。 In either setting。

we can compute the percentage of flowtime that is used for value-adding activities。

as a good diagnostic tool to see where flowtime is wasted in the process。 That's it for today。

I see you soon。 [BLANK_AUDIO]。

沃顿商学院《商务基础》｜Business Foundations Specialization｜（中英字幕） - P13：12_什么是客户本位.zh_en - GPT中英字幕课程资源 - BV1R34y1c74c

[MUSIC]。

Welcome back to Module 2。 Just to review in Module 1。

we looked at traditional ways of doing business， particularly for a strategy associated with performance superiority or。

operational excellence。 And we looked at the different characteristics of businesses that do。

that kind of thing， which of course I called product centricity。

We discussed why product centricity isn't quite as great as it used to be。

and started touching on businesses that turn in a different direction， customer centricity。

But we haven't defined it yet。 We've given a number of examples of companies that we'd say are。

or at least have been highly customer centric， IBM， Harris， Tesco。

as well as a number of companies that are terrific companies， companies like Walmart， Apple。

Starbucks， Nordstrom， but don't qualify as necessarily being over the hurdle。

but are making great strides towards customer centricity。 So what about your business？

What about these businesses around me here on， South Street？

How do we determine whether business really is or isn't， customer centric？ In other words。

what is the definition of customer， centricity？ So what I like to ask my students to do is to write that down。

So how do you define customer centricity based on what we've discussed so far？

Based on the examples that we've looked at。 So in fact， I'd like you to take a minute。

and just jot down whether it's a full sentence or even just a few words。

that you would associate with customer centricity。 Take a minute and do that。

and then I'll give you my perspective， my definition on what customer centricity is。 Okay。

so you gave me your definition of customer centricity。 I'm going to show you mine。

But before I put the words on the screen， because a lot of it's just going to be words on a screen。

I want you to look at these words and think about them a little bit differently。

I want you to think about how this definition of customer centricity and what it implies。

just how radically different it is from conventional product-centric business practices。 In fact。

I want you to look at these words and tell me， if you were to start doing exactly these kinds of tactics。

if your company was to start having these kinds of perspectives， why you'd be fired。 In other words。

there are fireable offenses in this definition over here。 And what are they？ Okay。

if you look at it， there's a lot of things that might make sense。 It all fits together， sounds nice。

Hopefully it's well-lined with your own definition of customer centricity。

But I really do want to emphasize just how different it is。 So for instance。

what are the fireable offenses here？ One of them would be this idea of select set of customers。

In the product-centric world， you can't have a select set of customers。

In the product-centric world， we're so dependent on generating as much volume as possible。

on selling as much stuff as we can， that we can't really afford to be selective。

It's going to be hard to keep our costs down if we're selective。

So the whole idea of having and emphasizing a select set of customers very much runs against the grain of many businesses。

Another would be the bottom line on this definition。

The idea of really focusing on maximizing the long-term financial value of certain kinds of customers。

In most situations， it's hard for a company to do that。

Given the pressures of Wall Street and just the conventional ways we look at business。

we're so short-term oriented。 We've got to hit the quarterly numbers。

Whereas in the customer centric world， and going back to many of the examples that I mentioned before。

we want to invest in the right customers。 We are willing to recommend products and services that we're not going to make any money off of。

For instance， going back to the IBM example， there was a case where a company was willing to recommend other products and services。

So IBM was actually losing opportunities。 But locking in customers for the long run。

being seen as a trusted advisor， in some cases can be worth it。

That the long run profits that we can get from customers can be greater than just trying to get them to buy another thing right now。

So again， that's a radically different way of doing business。 Another part。

higher up in this definition， is the idea of aligning our research and development activities around our customers。

The way it usually works is we go to the R&D people and we say， "Hey R&D guys， gals。

come up with the next blockbuster for us。"， You've been so good at coming up with these terrific products and services。

What's the next big thing that you have for us？ But here we're talking about something different。

Here we're saying， "Hey R&D guys and gals， here are really valuable customers over here。

Let's come up with something for them。 Something that's going to make them even more locked in。

Something that's going to create greater long run value for them。

And something that's going to help us recruit even more customers like them。 R&D folks。

come up with something for them。 It's a totally different way of doing business。 Now。

if you think about that last point， it's not quite as radical as it might sound。 Because after all。

what made our valuable customers so valuable？ The fact is。

they liked the products and services that we develop。 And so if we leave it up to the R&D people。

whatever they come up with next， our customers will probably love it anyway。 But it's the mindset。

It's the idea of going R&D and putting these valuable customers front and center。

It's the way it just changes the conversation and perhaps the design within the organization。

That's what starts making it customer centric。 So that's my definition of customer centricity。

But again， those are just words on a page。 What we really want to think about is what this means。

See， there's a lot of companies that might adopt that definition or something else like it。

And then put a big banner on the lunchroom wall for all the employees saying。

"We are now customer centric。"， Well， it's not that easy。

There's a lot of challenges in actually bringing this definition and this mindset to life。

And so I want to think now about some of those challenges as well as some of those opportunities。

So we can see in the rest of the slide over here about what customer centricity really implies。

And I want to give you a few examples about that。 So again。

thinking about the fact that customer centricity requires us to be forward looking。

We're looking at not which customers have been valuable， but which customers will be valuable。

using the data， the models， the technology that we have available to us。 So what does that mean？

So here's a very specific example。 So so many companies have some kind of sales person of the month incentive。

And they tend to be backwards looking。 They're going to reward salespeople based on how much stuff they sold last month or quarter or year。

I want it to be forward looking。 Think about it this way。

If you have that kind of backwards looking program， you're encouraging。

you're in centering your salespeople， to try to close sales that were going to happen anyway。 Like。

you know， "Hey， I got to get this one done before the month end so I can get my bonus。"。

That's not necessarily helping the company in the long run。 In order to have real long run benefits。

you have to be future looking。 So here's the way a sales person of the month incentive should work。

I want a company to calculate the lifetime value of each and every customer。

And let's do that at the beginning of the month or the quarter or whatever。

and then do it at the end of the month of the quarter。

And let's ask ourselves not just how much stuff we sold to the customer。

but how much did we elevate their lifetime value。 So instead of us going to customers who are going to buy things anyway and just watch them buy things they were going to buy。

let's try to build relationships with customers。 Maybe they weren't inclined to buy。

And you know what？ Maybe they didn't by the end of the month。 But we're closer to making the sale。

We've improved the relationship。 We've lengthened and maybe deepened the relationship。

That we think that in the long run， we will create much more value that wouldn't have been there。

That's how I want to reward the salespeople。 On future value that they're sowing the seeds to create。

Now that's risky。 It requires some faith。 It requires some data。 It requires some models。

But if you can do it， and I'm aware of a number of firms that have in a variety of different businesses。

then you're actually much better off。 Think about it from the sales person's perspective。

Instead of just rewarding them based on what they've done。

you want to encourage them to build relationships。 You don't want to just close sales。

You want to build long lasting relationships。 You want them to invest in the customers。

even if they're not getting anything out of it right away。 And then after all。

that's what salespeople want to do。 They want to build and strengthen relationships。

They don't want to just close sales and move on。 So if you have this kind of forward-looking perspective。

not only might it be better for your shareholders in the long run。

because of the profits you'll create， but it's even better for the salespeople。

because that lets them do what they're really good at。 And again。

I can point to examples of companies， I'm thinking of a particular pharmaceutical company。

that changed its salesperson incentive program to be forward-looking instead of backward-looking。

and wonderful things happened。 The salespeople were happier。 The company made more money。

And the salespeople actually looked to the marketing people， say， "Hey。

can you help me identify other good prospects that I should be going after？"。

So instead of just trying to， you know， shake down customers to just make sales right away。

that kind of relationship building is good for absolutely everybody。

These kinds of forward-looking incentives work in other ways as well。 Think about airlines。

Think about MBA students。 Spend a lot of time thinking about MBA students。

What happens to our warden students when they come to school？

So they were working in industry before， spending a lot of time flying。

Now what happens for the two or so years that they're at warden？

Their status with the airline drops， and then when they start on a new job after graduation。

they have to start all over again。 If the airlines were really forward-looking。

they would recognize that some of these students， are going to take a temporary hit on their travel。

but after they graduate， they're going to be traveling even more。

far more than they ever did before。 So if the airlines were smart。

they would go to our students the day they're admitted， and say， "You know what？

We're going to put you in the President's gold medal。

Chairman's red carpet club for the next five years。"， Because we recognize。

based on what we know about you， that you're going to be a really good customer in the future。

And even if you're not going to be a great customer tomorrow， it's worth the investment for us。

That's what I'm talking about， and that's what we don't see a lot of。

Customer centricity requires us to look ahead， figure out who the valuable customers will be。

and do things for them to help them recognize， that we have their best interests in mind。

That's the kind of investment that I'm looking for。

Those are the kinds of incentive structures that I want。

and some of the organizational designs associated with it。

That's what customer centricity should be all about。 [Music]。

[BLANK_AUDIO]。

沃顿商学院《商务基础》｜Business Foundations Specialization｜（中英字幕） - P130：14_实现流程的顺畅流动.zh_en - GPT中英字幕课程资源 - BV1R34y1c74c

One of the privileges of being a Warren professor is that I get to work with some really smart。

doctoral students。 My students are my collaborators in research and more often than not they're also my teaching。

assistants in my class。 At this point I should confess that it's really my teaching assistants who grade the。

final exam。 The exam consists of a dozen or so questions， each of which has two， three。

four sub questions。 So let's look at how the TAs organize the grading process。

Here's a picture of two of my favorite doctoral students， Leslie Meng， now on the faculty， at Yale。

and Dawson Coward， now on the faculty of Georgetown。

Leslie and Dawson were grading exams on this day， about 120 of them。

Not surprisingly they had to split up the work between the two of them and you see how each。

of them is grading about 60 exams。 What do you notice when looking at this picture？

First we see that Dawson has 60 exams in front of him。 All of them are turned on a particular page。

Leslie in contrast has only one exam in front of her and she has one stack of exams to。

her left and one stack to her right。 Can you imagine what they do differently？

Well let's start with Leslie。 Leslie picks up an exam， goes to the exam from beginning to end。

and then enters the， exam score into a table。 One exam after the other until all of them are fully graded。

Make sense？ Doesn't it？ Look at Dawson。 Dawson starts with question one。

say question one has three sub questions， and then Dawson， grades question one for all the exams。

Then he moves to question two and repeats the whole thing。 Ask yourself for a moment。

What are the pros and cons of the two strategies？ Now since few of you will be entering the exam grading business。

I want to generalize， this example and talk about a flow units journey through a process。

What we've done so far in this course， we've simply assumed that the flow unit makes its。

way through the process whenever there's resources downstream that has capacity。

But how that flow unit is managed and what triggers a flow unit to advance is actually。

pretty important。 In this video I want to talk about three ways we can manage flow。

We first talk about push versus pull， including just in time deliveries。

We then talk about mix model production。 And finally we will talk about single unit flows as opposed to transport batches。

Imagine you have a three step process with the capacity levels that you see here。

You have enough input for that process and you have hard working and motivated employees。

What's going to happen？ Well it's clear that station two here is the bottleneck。

Since station one is motivated and works really hard， he or she is going to push units into。

the process。 All of them are going to pile up in front of station two。

This really doesn't help the flow rate because the capacity constraint is at station two。

but it will increase the inventory。 In a setting like this。

what we really want is station one to slow down。 We prefer station one to be idle for half of the time as opposed to just flooding the。

process by pushing in more inventory。 You might say that we could send the work out station home after we completed a pile of。

inventory at station one in the first half of the day。 There's certainly some truth to that。

but the problem is it's certainly not lean。 Because now we have a pile of inventory between station one and two。

Instead what we should do is we should pay the process based on the speed of the bottleneck。

Even better， we should think about how we can pace this process based on the speed of demand。

Remember the concept of tag time。 We should operate at the tag of demand。 Now ideally。

we would also balance the line so that big differences in capacity levels。

would not occur in the first place， but you will just never get a perfect line balance。

So what can we do to avoid undesirable inventory buildup？

You have to really pull the work through the process from the downstream as opposed to pushing。

the work in from the upstream。 Now one way of achieving such a pull system is known as a conmances。

Imagine you're sending these lovely black boxes here to the market。 For sake of argument。

say the downstream here is the last resource in the process。

Once you've depleted a certain amount of inventory of the boxes at the last resource， typically。

one container， you need to inform the upstream that you need more。

Connected to that container is a conman cart。 Once a container is empty。

that conman cart is going to be handed over to the station。

the resource upstream that is feeding you， your internal supplier。

Now that person or that resource feeding you should sit there idle and do nothing until。

they have received the conman cart。 That's why a conman cart is oftentimes referred to as a work authorization form。

Now again， it sounds really stupid because you're forcing people to be idle。 But in the lean world。

it's better to be idle rather than having to run ahead and push， inventory into the system。

So through the conman carts， you really control how much inventory is in the system。

You authorize the amount of inventory。 Now the extreme version of that approach to pulling work is you have a conman cart for。

every single flow unit。 That means a container size is really equal to one。

And every unit is now replenished one by one。 Consider for example the replenishment of complex modules such as seeds or dashboards or。

exhaust systems in an automotive plant。 It's fascinating to see how when the vehicle enters a certain point in the production。

say the painter， there's an electronic signal that is sent to the supplier requesting the。

modules for the final assembly。 And then the supplier has something like three or four hours to basically go through their。

operation of making the module， putting it on a truck and delivering it to the final assembly， line。

And then magically just when your vehicle that is in need of the seat is at station 255 seat。

assembly that pops up just in time。 So this is real just in time often referred to as synchronised production。

Next let's talk about the idea of mixed motor production。 Remember Dawson the TA。

Why did he like to grade all the exams on question one before grading all the exam on， question two？

The reason for that is what in operations we call a setup time or a switching time。

If you've watched my module on product variety in multiple flow units you might remember。

the idea of setups and that they require capacity。

Imagine an automotive plant that is producing two types of cars， a wagon and a sedan。

Since the body panels of a wagon are different from the body panels of a sedan， some of the。

stamping machines need to be retooled as you switch over。

The operator is having to do something that is called a changeover。

Think about making cookies for exams， you're baking cookies and you're switching over between。

making cookies in the shape of a star and sometimes you make cookies in the shape of， a heart。

The problem of course is that changing over a stamping machine that is as big as my house。

and a stamping dies which weigh multiple tones of steel。

Well that is much harder than switching over from a star cookie cutter to a heart cookie， cutter。

So setups and changeovers create a form of scale economies which is really a form of specialization。

This is by no means unique to an auto plant as we saw again with my TA example。

You might also think about a surgeon who might find it easier to do six miniscule surgeries。

in a row as opposed to doing a knee surgery， then a hip surgery and then doing an anchor， surgery。

People like specialization， people like scale economies。 Anyway， back to my auto example。

Even that retooling is hard and will really eat up my production capacity。

The plant manager doesn't want to switch forth and back between sedan and wagons all too， often。

So he or she does what Dawson did。 He or she runs a long production run of sedans。

Now suppose the amount for sedans and the amount for wagons is about split evenly 50-50。

What happens？ While we're producing sedans， we're not producing wagons。

So any sedans from wagons has to come from inventory。

So wagon inventory is coming down and in contents since we're allocating all our capacity to。

sedans， the sedan inventory is going up。 Then eventually we'll switch from producing sedans to producing wagons and the opposite。

occurs。 Now the sedan inventory is coming down， the wagon inventory is going up。

You see the resulting oscillation here， the up and down of the inventory cycles。 And on average。

the inventory is exactly going to be in the middle。

Now contrast that with the plant that is able to change over production with less effort。

or even without any effort at all。 With smaller ultimately no change over time， what would you do？

Exactly。 We will produce at the rate of demand。 And since we assume that demand for wagon's incidents would be evenly split。

we produce， in much smaller batches and ultimately if there are no change over time。

we will produce， in batch sizes of one。 The technical term for this is a mixed model production。

also known as Hejungka。 Hejungka helps us keep inventory low。 It also has another nice property。

Imagine that you're rolling the processes to install the trunk opening mechanism which。

is specific and unique only for the wagon。 That means that when the plant is producing wagons。

you're running all out。 But then once the plant switches over to sedans。

you really have plenty of idle time。 You alternate between all outsprens and leisurely walks。

In contrast with mixed model production， you'll work on this level， right？

That employee putting in the trunk opener simply works on every other car。

So the moment you do Hejungka， work starts to be much more level。

I know that by now you're probably sick and tired of my Hejungka quotes， but here's one， last one。

Run like the turtles， not the hair。 This advice is really based on the ancient fable describing the race between the hair。

and the turtles。 The hair really cools the slow moving turtles which challenges the hair to a race。

And in the race， the hair quickly gains a significant lead and hence the hair is comfortable taking。

a little nap even in the middle of the race。 The turtles though， tired as it is。

keeps on growing at a steady and even pace。 When the hair eventually wakes up。

the hair is so far behind that all the hair does is， see how the turtles springs into victory。

Unless you control your processes， processes have the tendency to operate like the hair。

Your resources are motivated and so whatever they do， everything is pushed into the process。

as quickly as possible。 This is especially true in service processes where it's the customers who push themselves。

forward。 But you have to run an operation at an even and steady pace。 Unlike the turtles。

not the hair。 That gets me to my third concept I wanted to talk about in this video here and that is。

the idea of single unit flows。 Because of these change over times。

we have a reason to create a production batch or production。

run consisting out of multiple flow units。 Now similarly。

if we have a significant transportation time between stations， we have an incentive。

to wait for units to pile up before we do the next transportation trip。

Just think about how you're buying stuff like toilet paper。

Even short of the COVID-19 toilet paper craziness， chances are that you don't buy toilet paper。

every day。 Despite the fact that you're probably using toilet paper every day。

You might purchase 10 rolls at your local supermarket。 You might purchase 100 rolls at Costco。

Now buying 100 rolls of Costco toilet paper is likely to be cheaper and more convenient。

But if you do this， on average， you now have 50 rolls of toilet paper in the house。

The Toyota mindset is that if I consume one roll of toilet paper in my house every day。

I should be purchasing one roll of toilet paper every day。

If everybody would have followed this strategy， the March 2020 craziness around supermarket。

shares being emptied of toilet paper would certainly not have happened。

The idea of changing over from one model to the other every minute or two and the idea。

of replenishing your toilet paper run roll at a time sounds appealing from an inventory。

perspective but one that translated into lots of wasted production capacity and many。

trips to the supermarket。 Yes， that's a problem。 This gets me to my last insight in this video。

The key to implement a lean operation is not just to change your production plans or your。

planations and cycles。 Instead， you have to build processes that are flexible enough to support this。

You have to invest in flexible production equipment， you need to train your workforce。

and you need to involve your suppliers potentially asking them to build a plan next to yours。

This is why in the introduction module， I made inflexibility one of my three system inhibitors。

All this requires a much bigger transformation than you alone deciding that you， from tomorrow。

onwards， are going to run a lean operation。 In the next video。

we'll talk about how such a bigger transformation looks like。 I see you then。 [BLANK_AUDIO]。