Studying The Goal and TOC [AM]

Project Summary

I’ll be studying The Goal by Eli Goldratt. I’ll try learn and master the concepts deeply.

I would gladly read non-fiction TOC that explains the same concepts. The Goal can partially act like a starting point to guide what to learn.

I’ve already started reading and taking notes three days ago and I liked it a lot. I liked it more than I thought I would. I’m at chapter 9.

I think there’s a big risk I won’t finish this so I have a plan-B which is just to read the book and take notes. Not going deep and making sub-projects for each major concept.

Goal

What’s your goal? Why do you have that goal? How will you judge success and failure? What bigger picture goals or values are you pursuing? How is this relevant to CF?

I want to learn concepts from TOC deeply. TOC is useful in itself but is also an advanced prerequisite of CF.

I want to use TOC to think better and become more productive. I don’t expect TOC to be a full solution to productivity, just help with some parts.

CF, philosophy and being productive are bigger picture goals.

Plan

What’s your plan? How big is the project? What resources do you expect it to require and what have you allocated for it? How confident are you about succeeding? What sort of errors or error rate do you expect and how will you deal with that? Got any error correction mechanisms? What are the risks of not finishing the project or failing and do you have any plan to address those risks?

The plan is to read the book and take notes on the plot relevant to the ideas.

I won’t just do a mega project and evaluate it at the end. When some idea is fleshed out enough in the book I’ll make a sub-project dedicated to learning that specific idea. In the sub-project I’ll look for non-fiction about that idea too. I want to master those ideas, so I’ll practice them and practice things like noticing when they’re useful. Success for this project means going through the entire book and having success in each sub-project.

I don’t know how long it’ll take because I don’t know how many ideas there are that I’ll make sub-projects for. I’m fine with it taking like 6 months even. Most likely I’ll cycle between this project and prerequisites, probably just grammar. I could do both at the same time but given my recent project activity I don’t think I will.

I’ll try to do 3 hours a day for 2 weeks, and then my semester starts so I’ll do at least 1 hour a day then. I’ll try to capitalize on my enthusiasm right now and do even more than 3 hours a day for these 2 weeks.

It’s a long project so there’s lots of risk in that. Also I haven’t practiced and mastered any ideas like I’m planning to. There’s risk that I’ll lose the enthusiasm I currently have about the project. I think there is a big risk I won’t be able to do the project as laid out. If I see that I’ll have to learn to master ideas with easier things before I can master TOC ideas then I’ll change this plan to a much easier one. I’ll just read the book and take notes like I already have. I won’t make sub projects and I won’t have to dive as deep into the ideas. I’m confident I can get through the book while taking notes. Length of that plan will depend on how long conversations on things I’m confused about will take, if people will discuss with me.

Other People

What help are you asking from others? What value are you offering to others? Will you complete the project independently if no one else participates? Why are you sharing this with others? What sort of criticism do you want?

I can complete the project independently. At least the plan-B project. The plan A project is more likely to require help, but I’ll try my best on my own.

I’ll be happy if someone will help me on stuff I’m confused about, I already have one thing I’m confused about.

Criticism about my ideas and my project are good. I haven’t done anything to warrant demanding unbounded criticism yet so I can’t ask for that. I trust people to give any criticism they think would be helpful to me.

Context

What’s the context? What’s your relevant background and track record? Why are you prioritizing this over alternative projects? Why are you doing it right now? What have you already done?

I recently discussed TOC ideas in Introduction to Theory of Constraints - Elliot Temple - Critical Falliblism and Optimize Limiting Factors [CF Article] - Elliot Temple - Critical Falliblism.

I haven’t been doing projects lately which I’m disappointed in myself for. I’ve done 4 grammar projects three of which were successes and Another successful mini-project.

I also wanted to study AS with Elliot’s detailed analyses, but I figured The Goal and TOC ideas are easier than AS and Oism. I also hope to get productivity boost from learning TOC ideas.

I’ve been wanting study The Goal for and also just study a philosophical book in depth for a long time.

I’m copying some of the style from Notes on "The Goal" by Goldratt. I haven’t read the topic in detail, I’ve just done a really fast skim and read some meta stuff between Alan and Elliot. I don’t remember things he wrote. I remember he noted that Peach parked in Alex’s spot which I thought wasn’t worthy of noting down.

I spent under an hour writing this down, which is faster than usual. I’m more excited to do this project than others and therefore more happy to fill out this template. It felt like I knew more what to write and that I had useful things to say. Sometimes I felt like I was just filling things out in the other templates.

Introduction

Science is simply the method we use to try and postulate a minimum set of assumptions that can explain, through a straightforward logical derivation, the existence of many phenomena of nature.

Goldratt is in favor of explanation in science, not just using science as a tool for predicting.

The Law of Conservation of Energy of physics is not truth. It is just an assumption that is valid in explaining a tremendous amount of natural phenomena. Such an assumption can never be proven since even an infinite number of phenomena that can be explained by it does not prove its universal application. On the other hand, it can be disproved by just a single phenomenon that cannot be explained by the assumption.

He recognizes the asymmetry between criticism and positive supporting arguments. I think he rejects justificationism then.

This disproving does not detract from the validity of the assumption. It just highlights the need or even the existence of another assumption that is more valid.

He’s in favor of conjectural knowledge or contextual knowledge. He says we can get closer to the truth.

Einstein’s assumption is not true to the same extent that the previous one was not “true”.

I think this is fallibilist.

That all seems very Popperian. Did Goldratt read Popper?

I searched “did goldratt read popper”, the AI said there was no direct evidence but linked to Curiosity – Eliyahu Goldratt Discussion:

Goldratt says TOC is application of scientific method. Second paragraph of the Forward of The Choice:

Early on I was aware that Eli is actually on a quest to demonstrate that the approach and methods of the hard sciences can and should be applied to the social sciences. He initially targeted management science, claiming that since in that branch of the social sciences results are measurable, people find it harder to dispute the superiority of using the hard science techniques. It was fascinating to see how gradually the business world accepted Eli’s work in spite of the fact that so much of it is a drastic departure from tradition. His Theory of Constraints (TOC) is now taught at almost every business school and MBA program and has been used by thousands of companies and government agencies worldwide. TOC has been successfully applied in almost every area of human endeavor, from industry to health care to education.
“I” = Goldratt’s daughter. The book is a collaboration between him and his daughter. It’s written from her perspective.

Cool.

Somehow we have restricted the connotation of science to a very selective, limited assemblage of natural phenomena. We refer to science when we deal with physics, chemistry or biology. We should also realize that there are many more phenomena of nature that do not fall into these categories, for instance those phenomena we see in organizations, particularly those in industrial organizations. If these phenomena are not phenomena of nature, what are they? Do we want to place what we see in organizations to the arena of fiction rather than into reality?

I think knowledge about organizations are more like philosophy since it involves people. But that’s using a strict definition of science. Applying some methods and attitudes that made science great, i.e. rationality, to organizations is great. That’s treating organizations as a part of reality and not fiction.

But production in an industrial plant feels like science. It feels like computer science. idk really why.

It’s not that important to determine whether it’s “science”. More important to figure out whether the ideas are good and work in reality.

Introduction to the First Edition

Understanding of our world is not something to be pursued for its own sake, however. Knowledge should be pursued, I believe, to make our world better—to make life more fulfilling.

I believe there is value in knowing things to satisfy your curiosity, but I also believe it’s best to pursue knowledge to improve your life. I like Frisco’s motto “What for?”.

Family Plot

ch 2

Alex had planned to go on a date with his wife but has to cancel because of the emergency. Julie is lonely and hurt that Alex doesn’t make time for her. She seems ok or even excited that the plant might close such that they can move to a different town.

ch 6

Alex is late contemplating how to make money. He calls Julie and says “Guess who had a rotten day.” which is a mistake since Julie has also had a rotten day. Alex had forgotten the postponed date was that day.

ch 7

Alex’s daughter is up late waiting to show her getting all A’s to Alex.

ch 8

Alex was busy the whole day at work and remembered Jonah at the end of it so he cancels dinner with the family.

Factory Plot

ch 1

work is usually hectic. seems like there are lots of things going. Alex wants to get work done “before the fires break out.”

A big order is so late (7 weeks) the people at the plant weren’t even aware of it. they have the parts ready but can’t assemble because subassembly is missing. the subassembly parts are there but are not being worked on, instead some other urgent parts that took an hour and a half to setup.

So the parts are all there but they’re not working on it and they’re about to wast an hour and a half of setup. Super uncoordinated! now they have to spend time deciding what to do as well when they’re already in a pinch.

the machines aren’t ready on demand. when they find out they need to do something urgently they don’t have excess capacity to work on this thing. it seems like they have spare parts but they can’t work on them because everything is occupied. if some bad luck and urgent orders come up they have to disrupt other production to a large degree.

that’s like not having open slots on your schedule when things come up or take longer than expected.

Peach tells Alex his plant is lots of losing money. Alex partly blames it on not having enough employees. Alex has three months to turn things around.

The machine they needed for the late order is now broken. It’s the only one they have of that type.

ch 2

the machine is working again and almost everybody is working on the late order. parts per worker is low but at least they’re getting the really important order shipped that day.

Alex says they can’t always do things like these because it’s really inefficient.

Goldratt says that there’s something good in what you did to solve an emergency. In this emergency they had to really focus on a really important order and give up general “efficiency”. You wouldn’t want almost the whole plant working on a single order, but the lesson is to focus on throughput of the important things rather than general efficiency.

I don’t think I noticed this on my first read. I think I wouldn’t have noticed this if I hadn’t read the emergencies section in Introduction to Theory of Constraints.

Alex tries to figure out why things aren’t working. They’ve got good people, machines and materials. He blames the competition from Japan whom I think did some of the things that Goldratt says are good, e.g. Toyota.

Alex has already costs a bunch and his plant is efficient. He contemplates making it even more efficient, but he figures it’s already so efficient there’s not much more to squeeze out.

They have lots of inventory, but the material does end up as finished products on time.

Late orders are late in America. Stuff has to be expedited to get done on time.

Alex says gremlins are causing trouble, jokingly. But actually they are. The gremlins represent variance and Alex’s balanced plant is not robust against variance so everything turns to chaos.

ch 3

Peach calls in leaders of the plants to tell them the whole division is in danger. Alex could be totally without a job in three months.

ch 4

Alex meets Jonah.

Alex is about to go give a talk about how robots will revolutionize productivity in American manufacturing. Jonah’s isn’t impressed.

Alex says the robots increased productivity but Jonah asked whether they increased profits, which they haven’t.

They didn’t lay people off, inventory has not gone down and sales hasn’t increased so Jonah concludes the robots can’t have increased productivity.

Alex argues his efficiencies are up and costs are down. Alex says this is important. The robots are running constantly. If they didn’t they would lose savings on cost per part.

Jonah already knows that their inventories have gone way up and that they’re late on everything.

If they’re running everything all the time at max capacity then some stations will produce more than the next station can handle. Therefore inventory piles up some places.

  • why is everything late? figure it out yourself

Jonah says the plant is inefficient according to the right measurements. Alex is using wrong measurements because of poor thinking. He says the world hasn’t been critical of production/organization theory and are almost not thinking at all.

Jonah asks Alex “what is productivity?” he answers “value added per employee equals…”

Jonah manages to get Alex conclude that productivity is accomplishing something in terms of your goal. Alex says it’s obvious but then says that one goal of his company is to increase efficiencies. Jonah says every company has one goal. Alex suggests power and market share. Jonah leaves Alex to figure it out for himself.

The goal is to make money.

The meeting with Jonah was 2 weeks ago when Alex didn’t know the whole division was in danger of being shut down.

ch 5

Alex remembers the talk with Jonah and starts doubting that anyone at the division knows what they’re doing.

Alex leaves the meeting to ponder what the goal of every business is.

Alex figures out cost-effectiveness is not the goal. Because they have lots of inventory bought at great prices, but what for?

They’ve laid off people so it can’t be to supply jobs.

Alex sees no reason why it couldn’t be producing products but Jonah said it wasn’t that.

It couldn’t be quality since UniCo has improved its quality to great and are still losing money. He thinks quality is important but not enough. Alex thinks the problem is quality can be costly. He concludes quality and efficiency are the goals. But he’s not satisfied. He can think of efficiently produced quality products that went bust.

He thinks of technology, but he remembers R&D isn’t usually emphasized in organizations and also thinks the just having the newest best machines wouldn’t make his plan productive either.

Alex realizes the goal is not to fill warehouses with products. They want to sell the products. So the goal could be sales. But Jonah said it wasn’t market share and what’s the point of selling a lot if you can’t sell at a profit.

He realizes the big thing is money. So he has to stop the company from losing money otherwise it’ll have to shut down. Then he realizes there’s no point in running the company if it just breaks out even. The goal is to make money.

So being productive is making money. The other things he thought of are just means of achieving this goal.

I remember patio11 telling about programmers not being fully aware of this. He said they shouldn’t sell themself purely on technical explanations but telling their clients how they can make them money or solve a problem.

ch 6

Alex walks by three idle workers and gets upset at their supervisor. Their quotas were finished and now they’re just doing some busywork because Alex pushed the supervisor. Alex realizes what they’re doing might not be productive at all, it might not earn them any money.

Alex contemplates all the variables and complexity of the plant wonders how can he control it and make sure they’re making money. The measurements they make are supposed to help with that, but he’s not convinced.

Alex asks Lou the minimum necessary measurements in order to know whether they’re making money.

Absolute money made is a start but we have to consider it relative to how much was invested. ROI. But lots of profitable businesses with good ROI have gone bankrupt because they ran out of money. So cash flow is necessary too. Net profit, ROI and cash flow are what matters.

You need enough cash to survive, if you go below the required amount the company is finished. So nothing else matters when your cash flow is bad. They realize their cash is running out.

Lou blames the poor performance on lots of common things. The union, work ethic, low cost foreign labor, not caring about quality, etc.

Alex wonders whether he can prioritize one of the three measurements at the expense of the other two. He realizes you have to increase all three simultaneously. His goal is clear.

He has to figure out how to make what happens in the plant connect with increasing those three measurements. He’s stuck.

I think the goal in the plant is throughput. Finished products that they can sell.

Alex talks to a supervisor at the plant what impact they’ve had on the three measurements. The supervisor doesn’t think about those three, he has lots of other measurements he worries about.

I think Alex realizes that he has to think about something to connect the two. He can’t expect the plant workers to think about net profit, ROI and cash flow, those are for the HQ to think about. They need some measurements in the plant to be a proxy for making money.

ch 8

Alex is on a call with Jonah.

Jonah says the goal can be expressed in different ways. You need to express the goal which can guide daily operations at a plant which net profit, ROI and cash flow can’t. Jonah’s measurements are throughput, inventory and operational expenses.

I remembered throughput but that couldn’t be it alone because you at least need to spend less than you sell for. Inventory is less obvious and seems less direct, I do remember it was important though.

Jonah:

“Throughput,” he says, “is the rate at which the system generates money through sales.”

Alex asks if he can substitute sales with production. If you produce but don’t sell you’re producing something no one wants. You’re filling up warehouses, not making money.

Jonah:

“The next measurement is inventory,” he says. “Inventory is all the money that the system has invested in purchasing things which it intends to sell.”

“Operational expense,” he says. “Operational expense is all the money the system spends in order to turn inventory into throughput.”

These are carefully designed definitions. Changing one of them likely causes change in another.

“But what about the labor invested in inventory? You make it sound as though labor is operational expense?”

“Judge it according to the definitions,” he says.

“But the value added to the product by direct labor has to be a part of inventory, doesn’t it?”

“It might be, but it doesn’t have to be,” he says.

“Why do you say that?”

“Very simply, I decided to define it this way because I believe it’s better not to take the value added into account,” he says. “It eliminates the confusion over whether a dollar spent is an investment or an expense. That’s why I defined inventory and operational expense the way I just gave you.”

I’m confused here.

Labor would be operational expense since it is money spent turning inventory into a product which gets sold. But that inventory will have value added from the labor. So if inventory was the value of things that is intended to sell then then the labor would be counted as part of both inventory and operational expenses. But Jonah defined inventory as investments in things purchased which is intended to sell. But he also says the value added by direct labor could be inventory. But he also says it eliminates confusion over whether something is an investment or an expense. I don’t get it.

I interpret “direct” as labor done by their workers, as opposed to indirect labor which would be them buying some material from a different company which spent labor adding value to that material.

Jonah will explain local and global optimums later.

This is being discussed in chapter 10. I’ll read that before asking for further help. Probably the characters will figure it out.

I can first try to do a sub-project to learn to practice and master philosophical ideas before I do the easier The Goal project.

Factory Plot

ch 9

The CEO will come to the plant for publicity in a month.

Alex matches the three measurements with the questions Jonah asked in their first meeting:

  • Throughput - did your sales go up?
  • inventory - did your inventory go down?
  • operational expenses - did you lay off anybody?

So the way to express the goal is this?
Increase throughput while simultaneously reducing both inventory and operating expense.

Did the robots do that? Inventory and operational expenses went up and he doesn’t know whether throughput went up or not. He knows efficiencies have gone up so cost-per-part has gone down.

But did the cost really come down? How could the cost-per-part go down if operational expense went up?

The answer would be that they’re producing more things (without necessarily selling more). In addition to labor costs that takes it probably also costs to run (oil, electricity) and maintain the machines. If they’re building up more inventory that means each part could be made cheaper but they spent more in total. More inventory means they’re putting more money in the system but not necessarily producing more finished products.

Alex and Lou find out sales have not changed after the robots were installed. Instead overdue shipments have increased.

Alex want to know what happened to inventories. He asks for stats on work-in-process on parts produced by the robots. This seems like measurement they’re not currently keeping track of.

They ask Stacey who manages inventory control. She says inventories have gone up on the robot parts.

She says the robots were only at 30% efficiency when they came in. So she released more material to the robots such that they would be working at as high efficiency as possible. Since then those parts end up as surpluses and inventory has been growing. Those parts aren’t being consumed because they’re parts no one asked for, they’re just producing extra to keep the efficiencies up. And somehow they’re not getting the parts they need in order to finish their products on time.

I think I can answer:

The machines are not ready when needed. They’re not producing things just-in-time. They’re not producing things that are needed when they are needed. Like the example late order in the beginning, the machines are working on other things when an urgent order needs to use that machine. They may in general have lots of inventory but they could be missing buffers for certain things. They’re not focusing on the urgent orders.

Bob says they can get the parts that are needed but too late. They’ll have a lot of some of the components but not enough of other components and which they have lots of and which are in shortage changes.

Alex questions whether cost of parts has gone down because of increasing operational costs of managing larger inventories.

ch 10

Alex tells them what he has learned from Jonah.

Bob asks what sales has to do with them, the manufacturers. He says sales is for marketing. Alex points out that they’re not in business to fill warehouses with finished goods.

“Interesting, isn’t it, that each one of those definitions contains the word money,” he says. “Throughput is the money coming in. Inventory is the money currently inside the system. And operational expense is the money we have to pay out to make throughput happen. One measurement for the incoming money, one for the money still stuck inside, and one for the money going out.”

“Maybe Jonah feels direct labor shouldn’t be a part of inventory

I think I misread what Jonah said.

“But the value added to the product by direct labor has to be a part of inventory, doesn’t it?”

“It might be, but it doesn’t have to be,” he says.

“Why do you say that?”

“Very simply, I decided to define it this way because I believe it’s better not to take the value added into account,”

Alex said value added has to be a part of inventory. Jonah said it doesn’t have to be. Jonah said that value added by direct labor can be accounted for using different definitions. In his definition value added is not taken into account in inventory. I thought he said it could be dependent on the situation. But now I think Jonah is consistent with direct labor not being a part of inventory. His definition eliminates the confusion of whether value added by direct labor is an expense or an investment. He says it’s an expense.

I think Jonah wants to take what they sell minus total cost that went into the product in order to figure out profits. Total cost is materials and tools they bought and labor paid for to produce or maintain the materials and tools. Every cost is covered by those things.

That would be contrary to my guess that TOC doesn’t track every little cost:

“All this is, if I understand it correctly, is a different way of doing the accounting. All employee time—whether it’s direct or indirect, idle time or operating time, or whatever—is operational expense, according to Jonah. You’re still accounting for it. It’s just that his way is simpler, and you don’t have to play as many games."

I’m guessing the other way of accounting is cost accounting? And they differentiate direct or indirect, idle or operating and more.

(It seems like idle time is measured and generally minimized in cost accounting:
Idle Time: Causes and Accounting Treatment | Employees

The reasons for the idle time are to be analyzed and the management needs to know the reasons for avoidable idle time so that correction can be formulated to reduce and minimize the idle time. An idle time report is prepared as shown in Table 6.1 for giving necessary information on idle time with an analysis of causes.
)

I would guess that TOC thinks those measurements are local optimums. Minimizing idle time is a local optimum. While operational expenses is a global measurement.

Jonah mentioned confusion between investment and expense; are we confused enough now to be doing something we shouldn’t?

I guess the confusion is in how they measure the value of workers time and the value of the inventory. Currently they’re counting any work done on materials to be valuable. The material was improved therefore they did valuable work. But they’re not considering whether it gets put into a finished product on time. They’re not considering whether it’s blocking something else more important to be worked on.

Then I think about the “soft” things in business, things like knowledge—knowledge from consultants, knowledge gained from our own research and development. I throw it out to them to see how they think those things should be classified.

Classified as either operational expense or inventory.

I would think of R&D as an investment. That’s also knowledge you can sell. You can sell a patent or a course teaching the knowledge.

Knowledge from consultants I would imagine is not unique to you and I don’t think you can sell it. It’s also more of a fix now solution rather than R&D which is setting up for future progress. So I think I would classify knowledge from consultants as operational expenses. It’s a one time operational expense to improve now.

If it’s knowledge, say, which gives us a new manufacturing process, something that helps turn inventory into throughput, then the knowledge is operational expense. If we intend to sell the knowledge, as in the case of a patent or a technology license, then it’s inventory.

I thought about the definition for inventory so I got patents being inventory. But I didn’t think about the definition for operational expenses so I was worse on that one.

Definitions for reference and refreshing:

“The next measurement is inventory,” he says. “Inventory is all the money that the system has invested in purchasing things which it intends to sell.”

“Operational expense,” he says. “Operational expense is all the money the system spends in order to turn inventory into throughput.”

The R&D and consultants divide isn’t fundamental. Though I got the tendencies right. I don’t think consultants would give you knowledge you would sell as a patent. They would tell you how to improve operations and increase throughput. R&D can also produce knowledge that is used to turn inventory into throughput.

But if the knowledge pertains to a product which UniCo itself will build, it’s like a machine—an investment to make money which will depreciate in value as time goes on. And, again, the investment that can be sold is inventory; the depreciation is operational expense.

Project Notes

I’m in the middle of chapter 10. I’m taking more time to think through things and come up with some guesses before the characters figure things out.

I spent 5 hours today.

Family Plot

ch 9

Julie is gone when Alex gets home the day after, he had slept at his mothers place where he made the call with Jonah.

ch 11

Julie is not happy that Alex is leaving for New York to meet Jonah on such short notice. Alex says he didn’t know and always tells her when he knows he’s leaving. Julie is frustrated with the unexpectedness and lack of reliable planning from Alex.

Alex thinks Julie doesn’t understand Alex’s problems and that she’s unfair to him. Julie thinks Alex isn’t being fair.

Alex tried to call Julie when he arrived at the hotel. Julie didn’t answer.

ch 12

Julie seems to have been crying when Alex comes home.

Alex wants to know where she had been when he was calling her. Julie says he doesn’t care that she always home with the kids but cares when she’s away just once and that he has been late and away many times.

Alex says that’s always business. He says he always tells where he is so asks again where she had been.

She had met with a female friend because she had to talk with someone.

Alex asks “How did this come over you all of a sudden?”. Mistake, she’s been very lonely because Alex hasn’t been there.

She says he’s been prioritizing work over everything else since he got promotions and got into management. She doesn’t say whether he should’ve turned down the promotions or not.

He promises more time with the family.

She says he’s not even present when he’s with the family. He’s always thinking about work.
He says it won’t be like that when he’s out of the crisis.

She says he always says the job is in danger then how come he gets promoted all the time? Good point. Seems like she doesn’t believe him that his job really is in danger this time.

They bargain about him spending more time home. They agree that he will take the paperwork home. He also promises to do something with Julie this weekend.

Project Notes

I started labeling sections where I made guesses and did more of my own thinking rather than repeating what was said in the book with “I say” and “My guess”. At the end I thought I would make them more visually distinctive by making them bold. I think I missed some small parts.

I spent almost 4h30m today.

ch 10

But if the knowledge pertains to a product which UniCo itself will build, it’s like a machine—an investment to make money which will depreciate in value as time goes on. And, again, the investment that can be sold is inventory; the depreciation is operational expense.

I’m confused here.

Why would it depreciate? Maybe because other businesses will copy their product design?

“the investment that can be sold” that’s selling the knowledge as a patent I suppose.

Why would depreciation be operational expense? Wouldn’t that just be getting less money for each product sold, not an expense?

Alex says it’s like a machine. The machine depreciates because it gets worn down. But knowledge doesn’t get worn down.

I made grammar trees to check if I read it correctly:
Simplified tree:


Complete tree:

Missing “again” which should be a child of “And”. “But” connects from the past sentence. “if” conjoins within the sentence.

That was how I read it. So I still don’t get it.

  • try to understand this quote later if no one helps

“Here’s one that doesn’t fit: Granby’s chauffeur.”

He’s an operational expense. He helps Granby be more productive and therefore helps the company make more money.

Alex says you don’t have to directly work on the product to help turn inventory into throughput. You can do more abstract work to help or manage the people who do directly work on the product. You can also make big level decisions which does impact the plants ability to turn inventory into throughput.

Alex says the robots have been counterproductive. Alex tells Bob and Stacy the plant may shut down in three months.

Stacy suggests not pushing through so much material in order to decrease inventories. Bob objects that efficiencies will go down.

ch 11

Alex tells Jonah about the plant being in danger. Jonah is busy. He can’t solve Alex’s problems for him, but Jonah will help him solve them himself.

Jonah says Alex has to focus on the fundamentals, not on the robots. Alex says the robots have to efficient in order to be productive in terms of the goal.

Jonah asks whether an idle employee is always bad. Alex says yes. He says it’s a waste of money. Jonah says a plant in which everyone is working all the time is inefficient.

My guess is that you need idle employees to be flexible. You need employees to be ready when bad luck happens or just when some important order comes in and you need to change priorities.

Jonah points out that only having excess manpower could their plant have made so many excess inventories. Alex thinks Jonah is suggesting to lay off people. Jonah says it’s that Alex isn’t managing the capacity of the plant according to the goal.

“A balanced plant is essentially what every manufacturing manager in the whole western world has struggled to achieve. It’s a plant where the capacity of each and every resource is balanced exactly with demand from the market. Do you know why managers try to do this?”

I say: They’re trying to produce the exact amount that is demanded while spending as little as possible. They’re trying to buy just enough capacity to produce the exact amount demanded.

Alex says having less capacity is missing out on profit while having more capacity is wasting money. Essentially the same I said but said better.

Jonah:

“Why do you think it is that nobody after all this time and effort has ever succeeded in running a balanced plant?”

I say because of variance. Some stations will underproduces which means the next station won’t have anything to do and will therefore have idle workers and machines. Other stations will overproduce causing large inventories to build up. I think balanced plant only refers to capacity but this would also unbalanced in the sense that you have an unbalanced amount of inventory at different places and compared against the demand.

Alex says it’s that conditions are always changing on them. Is Alex referring to variance in the plant or conditions placed on them by the market or something else? Alex blames different things with vendors, the work force and the market. Jonah says that’s not the number one cause.

Jonah says the problem is being closer to a balanced plant. Jonah says people are focused on trimming the capacity in terms of the goal. They lay people off and that reduces operational expenses, but also trims the capacity and that causes throughput to go down and inventories up. He says the goal is to improve all three measurements at the same time. Not to only improve one at the detriment of the others. That would be local optima.

He even says the operational expense doesn’t necessarily go down because the carrying cost of inventory goes up.

Jonah says that happens because of statistical fluctuations (variance) and dependent events.

I got variance but forgot it on really matters so much because of dependencies. Note: improve my understanding of dependent events.

Jonah says some information can be determined precisely while others are subject to variance. Jonah says the chef won’t know exactly how many eggs he’ll need. Alex says in a banquet he’ll know exactly. Jonah says the chef might drop a couple of eggs on the floor, so Alex agrees the chef needs some excess capacity.

Alex says that the fluctuations in the work of a worker will average out over time. Jonah says it’s not variance or dependent events alone that cause a balanced plant to not work, he says it’s both together.

ch 13

Alex is in charge of an overnight hike with Dave and the boy scouts.

I suppose I’m walking at about two miles per hour, which is about how fast the average person walks. At this rate, I think to myself, we should cover ten miles in about five hours. My watch tells me it’s almost 8:30 now. Allowing an hour and a half for breaks and for lunch, we should arrive at Devil’s Gulch by three o’clock, no sweat.

Alex notices the gaps between the boys are stretching. He decides to go in the back in order to leave behind anyone.

I say: What’ll happen then is that some other boys will start moving too far ahead of Alex’s group.

Alex ponders variation and dependent events. He connects it with their hike they’re currently doing. It’s dependent events since the boy ahead has to pass before the next boy can and so on. There is variation in the speed each boy is walking at. He doesn’t see the problem. He thinks the speeds will average out at about two miles per hour so it should be fine.

Herbie the fat kid was making Alex and his group hold up before a long steep hill. When Alex gets up he notices Ron, the one who’s in the lead, is nowhere to be seen. They hurry up and Alex has to tell Ron to stop. Alex tells Ron he was supposed to set a moderate pace, which Ron said he did.

Alex notices the gaps appearing again. Everyone seems to be walking at the same reasonable average pace, Herbie too. He wonders if it could be statistical fluctuations that makes the gaps grow ever wider. He thinks again it can’t because the fluctuations should average out.

I say: The flaw with the averaging out idea is that the boys can only catch up with those ahead, they can’t walk past each other. So if a kid ahead is walking slow it puts a max on how fast the kid behind can walk. And at the same time the kid two steps ahead is likely walking faster and increasing the gap. The kid behind will waste his fast pace period and fall behind later when he has a slow period and the kid ahead has a fast period. If the kid behind continues in an average speed instead of intentionally speeding up he will on average not make up the gap.

Dave stops to adjust his packstraps causing a 20 foot gap between him and Ron.

I say: If they both continue at a reasonable speed they won’t really be independent events anymore and therefore the speed differences will average out and they’ll keep the 20 foot gap.

Alex realizes he can’t shorten the gap himself because he can’t walk past the kid in front of him.

Alex realizes the fluctuations aren’t averaging out, but rather accumulating. Assuming everyone walks at the same reasonable speed all the time once a kid slows down and holds somebody up that means a permanent set-back (it’s possible for both to walk faster simultaneously and catch up, but on average that won’t happen) because the “dependency limits the opportunities for higher fluctuations”. They way to make up is to purposely make everyone who were held up walk faster at the same time. And so they would not all try to walk at the same speed all the time.

Alex realizes that farther back a kid is in the line the more he has to walk at a higher speed to catch up. The kid has to make up for all the excess gaps ahead of him.

Alex compares the situation with a manufacturing system. They’re producing a “walk trail”. A kid walking some distance is him processing some material. And he has to process it before the next kid can process it too. The product gets sold when the last person, Alex, finishes the trail. The throughput is measured by Alex finishing not by the rate at which Ron walks.

The gaps in between is inventory. It is material that the next kid has to process.

I guess operational expense is the energy they put into walking. When you have to walk at higher variance speeds your spending more in total energy. Walking at a steady pace is the most energy efficient way to walk. So Ron is using less operational expense than Alex is who has to stop and speed up a bunch.

Since throughput is the rate at which Alex walks the inventory ahead of Alex is the inventory for the entire system. The variance causes Alex’s inventory to grow over time. And for the plant operational expenses go up since the carrying costs go for the inventory goes up. I think the operational expenses are going up in the hike as well.

ch 14

They’re taking a lunch break. They’re behind on schedule. Probably because it was calculated assuming 2 miles per hour as average speed of the group as a whole.

Alex ponders how to overcome the effects of dependent events and variance in a manufacturing plant. He thinks: wouldn’t a balanced plant decrease inventory and remove parts shortages? He doubts the hiking model fits to a manufacturing plant.

He wonders how he could make the hike a balanced plant by trimming the capacity of the kids he way plant managers do to their plants. He realizes he can’t realistically achieve that. I say there will always be variance in the plant. He can’t make everything work at the same capacity all the time.

Alex creates a game to model a manufacturing plant to convince him of the effects of dependent events and variance.

The game is about processing matches. They’ll move matches through a succession of bowls to create finished matches coming out of the last bowl. A die determines the capacity of each bowl. The number you roll is how many matches you were able to process at your station.

The bowls setup a series of dependent events because the next bowl can only process as many matches as it has been given by the previous bowl. The die causes variation in output.

Throughput is matches coming out of the last bowl. Inventory is the total amount of matches in the system. Market demand is assumed to match the capacity (3.5 matches) of the bowls which is the same for each bowl.

Alex will record the deviations from the average. So -2.5, -1.5, -0.5, 0.5, 1.5 and 2.5 to match each possible roll. He will count cumulatively, so adding the deviation of that roll from the previous tally.

I’ll write a simulation of the match and dice game. It’s not necessary but I think it’ll be fun.

Factory Plot

ch 15

Alex realizes that the kids downstream it the balanced model didn’t have extra capacity to make up for bad luck. They have the same capacity all the time. Which means the following kids will be starving for parts and rolling above their inventory and therefore keep deviating negatively.

With the hike and the plant they have excess capacity. Alex tells the kids behind to hurry up when gaps (inventory) appears. Perhaps the machines are somewhat balanced in capacity but they can move workers around to different stations to increase the capacity of some stations creating an unbalanced plant.

Herbie thought he should be at the back because that way he won’t hold everyone up. The kids has changed positions such that the line is in order from fastest to slowest. This means nobody is held back, everyone works at their optimal speed, at optimal speed. This is like if in the plant everything was to work all the time. Like there was no idle time. Idle time in the hike is being held up by someone else. This causes the line to spread as fast as it could, i.e. inventory is increasing as fast as possible. Since for everyone the one ahead of them will run ahead of them since they’re faster.

Alex realizes that since he finishing the trail is throughput and that since he’s being limited to Herbie’s rate, it is actually Herbie who’s determining maximum throughput.

The kids who’re walking faster than Herbie aren’t helping the troop as a whole walk faster. The troop as the whole is dependent on the slowest walker finishing.

They’ve reached halfway of the trail and see they’ve been walking at a mile an hour as opposed to two an hour.

Does it matter where Herbie is positioned in the line? When he’s positioned at the end he’ll actually never be obstructed, so he can walk at full capacity all the time. However when he was in the middle when there was no clear bottleneck he would sometimes be obstructed. And since the bottleneck was switching around the current bottleneck could be held up by someone else. So when Herbie was in the middle they would walk at a slower pace than Herbie’s capacity. To keep max throughput, which is Herbie’s capacity, and minimize inventory they need to put Herbie in the front.

They divide up Herbie’s backpack to let him go faster. Make the bottleneck go as fast as it can. The others can reduce their capacity in exchange since it doesn’t help them be even faster so long as they’re faster than Herbie.

ch 17

Alex tries to explain to his team what he learned. They’re skeptical because they say robots don’t have statistical fluctuations. Alex says the non-robotic operations has variation.

Important job for Hilton Smyth:

  • Has to go through two more departments before done. First fabricating then they’re welded.
  • Has to be done at 5 o’clock. They’ll start at 12 and the welding robot will be setup at one o’clock.
  • They’re supposed to fabricate 25 parts an hour and the robot is capable that much an hour.

Alex explains how statistical fluctuations and dependent events apply to this situation. The fabrication will do 25 an hour and has variation. The robot should have no or minimal variation. The robot is dependent upon the parts being fabricated. They’ll make it an experiment in a balanced plant, so the robot will work at exactly 25 parts an hour. They’ll keep track of how many parts each department gets done every hour.

The fabrication got done making 100 parts by 4 pm. However they had a slow start only making 19 parts the first hour and 21 parts the second hour. They later hurried up to reach the 100 by 4 pm quota. Because the robot was set to only make 25 parts an hour that 10 part negative deviation couldn’t be made up for by the robot. The deviation the fabrication department passed on was the starting point for the robot to work on.

ch 18

Alex tells Jonah about what he learned. He said they shouldn’t try to optimize local optima but rather do what improves the system as a whole.

The next step for Alex is to categorize his resources as either bottlenecks or non-bottlenecks. Bottlenecks are defined as:

any resource whose capacity is equal to or less than the demand placed upon it. And a non-bottleneck is any resource whose capacity is greater than the demand placed on it.

Jonah says they should balance flow of product through the plant with market demand. Don’t balance capacity with market demand. And flow through the plant is flow through the bottleneck.

Jonah says you should keep the flow a little bit below the market demand in order to not lose money when the market demand falls.

  • come back to this point. keep flow a little bit below market demand

Bottlenecks are unavoidable. Can you have chain without a weakest link? Not with variance. With variance the bottleneck would change places all the time.

The next step is to find the bottlenecks. How do you find the bottleneck?
One thing would be to look out for where inventory is piling up. That would mean the station that should process that inventory isn’t meeting the demand that is put upon it. The bottleneck could be a material. In that case station that processes the material would underproduce. The demand for the material would be higher than you could get hold of. The material could be a product of some processing. You could look out for where you have a shortage of materials and trace it back to what is limiting the supply of it. Whether that would be due to a station having too little capacity, them buying too little of the material, or there not being workers to work on processing the material. If you trace back from a shortage I think it’s likely there’s also a big inventory ahead of where you find the shortage begins. Since it’s the station that goes too slow and causes shortage the stations before it is likely much faster and therefore inventory should be produced ahead of it.

Ralph the data guy has gathered data to figure out where the bottlenecks are. It turns out the data is very inaccurate.

Their approach was to figure out what the market demand for their products were and then find the resources that doesn’t meet the demand.

Jonah’s definition didn’t said the bottleneck was defined by the demand placed on it. He didn’t say it was specifically market demand. I think every resource could have capacity above market demand but poor organization means they’re not getting that throughput. Because capacity is how much they can do, not how much they’re currently doing, right?

If they’re trying to get a balanced plant it could mean the bottleneck is changing a lot and therefore they can’t just look at data over a large period and figure out the bottleneck. Maybe it’s possible to have every resource look like they’re over performing the market demand yet still have throughput less than market demand as a whole? I think that’s possible except for the last stations which would put the finishing touch on the product. If they met the market demand then the system as a whole would.

Stacey suggest looking searching at the places where they’re missing parts most of the time. Since the parts most often in short supply should be in short supply because they passed through a bottleneck. Alex suggest looking at the biggest inventory pileups.

The biggest pileup doesn’t necessarily mean bottleneck, it just means that it’s the biggest difference in capacity between any two stations.

Could a station which produces lots of inventory be a bottleneck? No, the one in front would be a bottleneck ahead of that station.

They identify the NCX machine, one of the most efficient machines, as one bottleneck. It took the role 3 other machines by being able to do all the 3 different operations the other machines did. The other machines together spent 14 minutes per part and the NCX spends 10 minutes. The single machine takes less people to maintain. The machine results in being one of the most cost-effective machines they have. With the old setup they have more machines for each type so they could produce more parts in total.

They identify heat-treat as a second bottleneck. It can process many parts at a time but takes a long time on each batch. The problem is that they’re running half empty because they get lots of emergency requests for less than a full amount of parts.

Alex realizes he can’t move the bottlenecks to the front of production. The stations are in the order they are because of the dependencies between them. It’s not like the hike where he could move the kids around any way.

Increasing capacity on the machines by buying upgrades to them is too costly.

ch 19

How else can they increase the capacity of the bottlenecks?

I can’t see what you could do with the NCX, that is if it’s running constantly. They said it was one of the most efficient machines. Did they mean by idle time or cost per part? If it was by idle time then that means there is still some idle time to get rid of. So have it running 100% of if it isn’t already. If I remember from my last reading I think that’s what they did.

For the heat-treat, schedule the whole production such that you won’t have to expedite it for small batches.

Jonah visits the plant since the quick fix of offloading work from the bottlenecks isn’t possible.

I remembered correctly. The NCX was idle when the workers had a break. Jonah says they should keep all bottlenecks working as much as possible. Production lost on the bottleneck is production lost forever for the entire plant. It’s production you can’t make up for later in the production.

Jonah suggests using the old machines. Lou objects that they’ll be very expensive to operate. The new NCX machine was bought in to be cost effective. They had gotten rid of one of the necessary machines.

Jonah asks if they could not heat-treat some of the material they currently are. Bob says engineering won’t like approving the change and Alex says it might take a month to get it approved.

Jonah suggests outsourcing some of the heat-treat. Stacey says that would increase cost-per-part. Jonah asks how much the piled up inventory is worth. Lou answers what it’s worth as material. Jonah says that’s not what it’s worth to them, to them it means throughput which means what’s the value of all the products you could produce with these parts.

They’re not sure how much it would be because some of the things they’re processing won’t end up as finished products immediately, some will be finished goods inventory or become spare parts. The bottlenecks need to work on the most urgent orders. They need to work on what becomes throughput as fast as possible.

They determine the products being held up by the bottlenecks are worth around a million dollars.

Jonah asks to inspect the quality control on bottlenecked parts. He asks what the impact was of rejecting the bottlenecked parts. Alex realizes it means lost time on the bottleneck and therefore lost throughput. Alex suggests putting QC in front of the bottleneck such that the bottleneck will only work on parts that are non-defective. Scrapping parts before the bottleneck means scrapping something you have excess capacity in making, but scrapping something that went through a bottleneck means lost time on the bottleneck. Jonah also says process control on the bottleneck parts should be very good. Bottleneck parts becoming defective also means lost time on the bottleneck.

Jonah says their calculations on what it costs to run the bottlenecks are wrong because of wrong assumptions. He said people who came to ask for errors in their calculation of physics problems weren’t wrong because of the numbers but because of wrong assumptions.

This is doing global optima cost calculation as opposed to local optima cost calculation which I suppose cost accounting does.

Jonah says losing an hour on a bottleneck is the same as losing an hour on the whole plant.

Jonah says there are two main themes of optimizing bottlenecks:

  • don’t waste the time of bottlenecks
    • don’t let it sit idle
    • don’t process defective parts with it
    • don’t make it work on parts not in current demand
      • make them work on things that lead to immediate throughput
        • working on spare parts is sacrificing current throughput for future throughput which hurts cash flow
  • offload work from the bottlenecks
    • shift some parts to be processed by nonbottlenecks
    • can a different machine do the same processing? offload work to it

Project Notes

I spent 4 hours and 10 minutes today. Yesterday all I did was program the simulation.

If I want to do the sub-projects and mastering ideas I think variation and dependent events could be a sub-project. I think that’s what I’ll do tomorrow.

It’s your first reading of any Goldratt book I take it?

Why do you like it more than expected?

I thought I had written this in this topic. It should be here too.

I thought it would feel burdensome to do this much analysis of a book. It gets tiring after a while but in itself it doesn’t feel burdensome. I actually think doing more analysis and thinking has made the book more enjoyable than my first read. I can’t really remember how I felt about the book 3 years ago though.


I’ve been sick the past 3 days. I wasn’t proper sick until yesterday and I could’ve done some work yesterday too.

That text doesn’t say that. I think you’re assuming that only pro-explanation people use the word “explain” positively, but that’s incorrect.

You’re reading too much into it. It’s unclear from this text what his epistemology is.

Someone could easily write those Goldratt sentences while disagreeing with Popper and Rand about conjectural knowledge and contextual knowledge.