Print Story I can only receive
By TheophileEscargot (Mon Jan 02, 2012 at 01:23:03 PM EST) Reading, Links, Me (all tags)
Reading: "Toyota Production System", "Econned". Me. Web.

What I'm Reading
Finished Toyota Production System by Taiichi Ohno. Slightly different type of cult book: Ohno was the Toyota manager who invented just-in-time production: the book describes how it works and how he introduced it.

The book is very short, only 176 pages including appendices, as befits a book about a minimalist system of eliminating the unnecessary. At £27.54 for the hardback it doesn't exemplify good value though.

Just-in-time production considers most inventory holding to be waste. It costs floor space, warehouse rent, time spent inventorying, and suggests spare capacity in whatever process produced this inventory. Ohno combated this by reversing the usual process. At each stage in the factory, the later stage submits a work order (a kanban worksheet enclosed in clear plastic) to the earlier stage, which makes only the quantity requested.

This reduces cost and waste, and also makes it possible to produce a large number of different items. It involves less central planning than traditional Ford-ist "large scale production": the decisions are taken at a lower level. Was interested to see that they alternate production at the lowest level, so the same line might make a Corolla and a Celica alternately all day.

This means the workers have to be multiskilled: the same worker might move from a lathe to a drill to a die-punch machine throughout the day. Interestingly, Ohno claims concern about workers being alienated from their labour, and suggests positioning workers so they are not isolated but can form meaningful teams. However, he is resolute that the number of workers per job should always and continuously be reduced to the minimum, and is vague about what happens to the Reduced.

Also interesting: it was only in the economic crisis of the late Seventies that the flexibility of the system led to it being adopted outside Toyota: it wasn't a product of a boom.

Ohno differentiates the Toyota Production System itself from kanban, the system of worksheets. The Toyota Production System is made up of kanban, just-in-time production, zero-defects (where a line is stopped by anyone if there's a problem) and autonomation where machines automatically stop if there's a risk of a problem.

One good idea seems to be the Five Whys. Whenever something goes wrong you should ask Why five times, tracing each cause back to a previous cause, to make sure your solution addresses the root of the problem.

The translated prose is simple and occasionally awkward, not sure how it compares to the original Japanese. Has mostly Japanese metaphors: I liked the idea of Ninjutsu Management, silently and near-magically making constant improvements.

Overall, a useful book if you run a manufacturing plant; and quite interesting if you don't but are interested in how the modern world makes stuff.

What I'm Reading 2
Finished Econned by Yves Smith, of the Naked Capitalism blog. Excellent, detailed book on both the specific origin of the financial crisis, and general problems with economic policy, financial regulation and the profession of economics. It's a couple of years old, so doesn't have anything on the Eurozone crisis or the failed recovery, and is also pretty focused on the US.

Despite that, and despite the plethora of other books on the origins of the crisis, the book is highly worthwhile. Smith is a financial insider with a good knowledge of economics. She's very good on explaining how economic models fail to reflect reality; and how weak, disinterested managers and ideological regulators had little grasp of what was going on.

This book is more thorough and better thought out than Robert Peston's "Who Runs Britain: The Economic Mess We're In", or Vince Cable's "The Storm", and more specific than Taleb's "Black Swan" and "Fooled by Randomness" and better than "The New Paradigm for Financial Markets: The Credit Crash of 2008 and What It Means" by George Soros.

Overall, presents a disturbing picture of how the formerly-useful banking sector has essentially been looted by profiteers, abetted by ideological zealots, who then raided the public purse for bailouts.

Sadly the more pessimistic worries of the book seem to have come true. Smith warned that propping up failed banks with bailouts and easing of regulators, has a history of making banking crises worse. It's better to let them fail and use the public purse to shelter the public from the consequences.

What we seem to be left with today is a zombie financial sector, incapable of its supposed role of lending and servicing the real private sector economy, endlessly supported and pandered to by the state. In turn, the private sector can't pull us out of recession since it's starved of investment capital and lending by the zombies. The state can't pull us out because all its resources have gone to propping up the zombies. So we seem to be stuck in a recession without a clear end.

Back at home now after spending the holiays with the parents. Back to work tomorrow, will probably be busy from the start. Girl B still doesn't get back for a while, won't see her till the 9th.

Feels good to have all my stuff around: gave myself a haircut and had a dumb-bell workout, feeling better now.

Don't think I've put on that much weight for once, though I'll need to weigh myself a couple of times at my normal time (first thing in the morning) to be sure. Weighed myself when I got back and it claimed I was 11st6, same as my last measurement on 2nd December, but I think that must be dehydration at work. I think I ate enough to gain a couple of pounds at least, but I might have lost a couple of pounds over December and put them on again.

Socioeconomics. Feminism's uneven success. Heterodox economics roundup. IP Feudalism and the Shrinking of the Public Domain.

Pics. Wooden lizard.

Sci/Tech. Hipster programmers.

Politics. Polls basically unchanged throughout 2011.

Stoicism. The lectures of Musonius Rufus now online in English translation for the first time, thanks to Michel Daw of the stoic forum.

< It's 2012. All good things must come to an end... | New Baby >
I can only receive | 24 comments (24 topical, 0 hidden) | Trackback
Just-in-time production ig reat until... by wiredog (4.00 / 2) #1 Mon Jan 02, 2012 at 01:57:22 PM EST
There's an earthquake in one part of one country that shuts down your entire worldwide supply line because there's no slack.

Earth First!
(We can strip mine the rest later.)

I don't think that's a JIT specific problem by lm (4.00 / 1) #2 Mon Jan 02, 2012 at 03:24:53 PM EST
That's a problem with globalization, not with just-in-time manufacturing.  After all, if you look at the old-school way of doing things, at best, you're given a bit of buffer before one order of a part runs out and the next arrives late because of the earthquake (or flood, et cetera).

Back in the seventies, the rust-belt in the US probably could have effectively organized a JIT system that was almost entirely regionally self-sufficient. Between the steel towns, the rubber towns, the car towns, etc., all the suppliers were pretty much right there.

There is no more degenerate kind of state than that in which the richest are supposed to be the best.
Cicero, The Republic
[ Parent ]
Dell used a JIT method by technician (4.00 / 1) #3 Mon Jan 02, 2012 at 06:41:12 PM EST
and may still. When I worked there they prided themselves on having no more than 18 hours of parts on the floor in the enterprise line.

That fouled them up a couple of times, most notably with hard drive, memory, and processor shortages. At one point, we emptied half our test lab of processors to help maintain (all of our processor stock was brand new, waiting for testing). Dell was just recently adversely affected by hard drive supply, and they were caught short by suppliers wanting tons of cash for scarce commodity where folks like HP and others had stock that they'd purchased at half the going price. Dell's margins are as low as tens of dollars for certain consumer systems, so that sort of thing hurts.

It's an interesting thing, too, to note that Michael Dell has in the last few years put some cash and effort behind understanding global climate change, since climate change spurred weather events can really muck with their entire process. To that end, Dell is now Just In Time Sort Of, as they count OTR and vendor stock as part of their capacity. They're trying to mitigate floods in Asia with a few days extra pipeline product, without actually putting it on the books.

Economic Crisis by brokkr (4.00 / 1) #4 Mon Jan 02, 2012 at 07:25:10 PM EST
At least we have zombies now, so it's got that going for it.
Deyr fé, deyja frændr, deyr sjalfr it sama,
ek veit einn, at aldrei deyr: dómr um dau∂an hvern.

ninjitsu management by clover kicker (4.00 / 1) #5 Mon Jan 02, 2012 at 07:50:47 PM EST
Is that where they're fucking invisible, show up out of nowhere to sabotage shit or gank someone for unknowable reasons and then disappear again?

Sounds like the only ninjutsu by wumpus (4.00 / 1) #6 Tue Jan 03, 2012 at 06:26:29 AM EST
management I've ever seen (or even heard of).


[ Parent ]
Zombie banks by Scrymarch (4.00 / 1) #7 Tue Jan 03, 2012 at 09:50:44 AM EST
The description is not wrong, and not good, but by comparison with Japan it's not entirely stagnant. Some big banks were let fail of course - perhaps inconsistently and insufficiently but still something. The worrying European element is basically another variant of the same thing, creditor bailouts where they are being shielded from picking bad risks to lend to.

Another more dynamic element is that bits are falling off the zombies. The financial sector is shrinking. Lazy nonstat link

It is happening most obviously in employment. Banks are still laying people off.

Iambic Web Certified

Toyota thing by Herring (4.00 / 1) #8 Tue Jan 03, 2012 at 11:35:43 AM EST
OK, I am deeply cynical, but a couple of things that don't seem to sit with "the British way of doing things":
This means the workers have to be multiskilled - A lot of my work seems to be about de-skilling processes

It involves less central planning than ... Well then there's less for managers to do if the system runs itself. How do they justify their existence?

Interestingly, turning the process backwards - starting with desired outcome (e.g. Celica, more profitable product) and working backwards methodically strikes me as a far better way of doing analysis than some smart-arse saying "We need a system that does X,Y,Z". I think I'm putting myself across badly but I know what I mean.

At the moment, the work for, say, team A is carefully planned (and we know what our estimates are worth). But their tasks depend upon tasks from teams B,C and D. All of their work is carefully planned too. If teams were able to just ask for things when needed, then a) we wouldn't have teams tied up doing work that wont actually be needed for months when there are urgent jobs and b) there wouldn't have to be this massive planning bureaucracy. It certainly wouldn't be less accurate in forecasting than the current system.

But then I have a feeling that applying "manufacturing type methods" to things like software development never ends well.

You can't inspire people with facts
- Small Gods

Starting at the end by Merekat (4.00 / 2) #9 Tue Jan 03, 2012 at 11:39:13 AM EST
It is called product based planning. I like it but I can't get anyone to understand it and they prefer to revert to a neverending cycle of pointless 'progress' meetings and filling in of templates, as if they were the work, not the end thing.

[ Parent ]
Seven by Herring (4.00 / 1) #10 Tue Jan 03, 2012 at 11:40:55 AM EST
At least.

Progress meetings are more important than progress. Aparently.

You can't inspire people with facts
- Small Gods

[ Parent ]
There have been efforts by TheophileEscargot (4.00 / 1) #14 Tue Jan 03, 2012 at 03:49:41 PM EST
To apply these ideas to software development. One of the reasons I was interested in the book.
It is unlikely that the good of a snail should reside in its shell: so is it likely that the good of a man should?
[ Parent ]
In my experience by ucblockhead (4.00 / 2) #17 Tue Jan 03, 2012 at 06:04:49 PM EST
Too often the application of these ideas is the buzzwords not in the actual process.
[ucblockhead is] useless and subhuman
[ Parent ]
I have worked on teams by Kellnerin (4.00 / 3) #21 Wed Jan 04, 2012 at 08:29:40 PM EST
that used Kanban methodology. This is in a Scrum shop where teams were given enough autonomy to do their own thing in terms of process and project management tools and some of them drifted in the direction of the new cool thing.

In practice it's a lighter-weight version of Scrum, without the iteration-based planning. (Scrumban is apparently a thing, and that's more like what we're doing.) You still line things up on your backlog but you pull on them when capacity opens up (not necessarily due to some strict limit on how many stories can be in progress at the same time, just when you run out of parallelizable work that the team already has in progress). The system is designed to be flexible but most teams seem to use "Think" as the first phase, where someone looks at the problem in enough detail to decide a reasonable way to approach it and then identifies tasks, which the team reviews together. The way I've seen it work, the tasks generated at this point include implementation, testing, or documentation instead of having specific phases like "Do" or "Test" (there are of course the usual sequencing considerations).

You don't estimate up front how long you think it will take to do things (How accurate do those tend to be anyway? Are they worth spending a day every iteration working out?) Instead you assign story points and determine the team's velocity based on past performance, and you make rough estimates about how much can get done by some future point (say, a desired release date) based on back-of-the-envelope story point estimates, which the team later refines after they've "Thunk" the story. (Inaccuracies in the back-of-the-envelope estimates tend to come out in the wash.)

Some benefits in relation to Scrum are that the team as a whole spends less time in planning meetings (without much decrease in long-term planning accuracy) and doesn't feel as much guilt for not closing off every story they "committed" to for an iteration. (Iterations don't really apply to Kanban, though you can still choose to have regular reviews for business stakeholders and team retrospectives for course correction.) It's also somewhat more flexible since you can move something to the top of the queue mid-"iteration" and the team can pick it up when someone frees up from their current work, rather than waiting till the next iteration's planning to see if the team can commit to it. That can (and maybe sometimes actually does) help with resolving inter-team dependencies.

"Plans aren't check lists, they are loose frameworks for what's going to go wrong." -- technician

[ Parent ]
One extra problem with the banking system... by Metatone (4.00 / 1) #11 Tue Jan 03, 2012 at 12:32:04 PM EST
is that even now, it's more profitable to engage in various trading and  strategies than to do boring old "lending to the real economy."

So, I agree with Yves Smith's diagnosis, but I think there's an extra layer of problems in getting money into the real economy via banks. And to some degree, maybe that explains the gap between Scrymarch and Yves observations of what is going on.

Getting money into the real economy by Herring (4.00 / 2) #12 Tue Jan 03, 2012 at 12:52:40 PM EST
Does the real economy want it? I mean, I don't want to increase my borrowing. Companies producing stuff don't want to borrow to increase production of stuff when demand is low...

You can't inspire people with facts
- Small Gods

[ Parent ]
There is some talk by TheophileEscargot (2.00 / 0) #15 Tue Jan 03, 2012 at 03:50:01 PM EST
Of a lending drought harming the real economy. But it's very hard to prove which way the causation goes. It's hard to know to what degree a lack of business lending is hurting the real economy, and to what degree a weak real economy means there is no need for business lending.
It is unlikely that the good of a snail should reside in its shell: so is it likely that the good of a man should?
[ Parent ]
Anecdotally speaking... by Metatone (4.00 / 1) #16 Tue Jan 03, 2012 at 04:45:31 PM EST
I know plenty of small businesses who have had their credit terms radically changed by the bank and as such are forgoing bank financing while they explore other options. For now they put things on hold, sack a few people, etc.

Now whether the SME sector is really the important to the economy, compared to the larger corporations, many of whom seem to be sitting on cash reserves (possibly waiting for demand to pick up, as Herring says) I don't know... 

[ Parent ]
lending droughts by ucblockhead (4.00 / 1) #18 Tue Jan 03, 2012 at 06:11:25 PM EST
Speaking as someone in a company that employees 180,000 people worldwide, I never, ever have heard comments about how we'd expand if we could just borrow. On the other hand, it's blatantly obvious from our balance sheet that higher unemployment means that there are fewer people out there in the world who can buy game consoles and 3D TVs. One under appreciated effect of the shrinking middle class is that it means lower sales for a lot of businesses. A guy with $5 million in the bank does not buy a thousand times as many TVs as someone with $5,000 in the bank. The rich spend fewer of their dollars, and when they do spend them, it is often on luxury goods that employ fewer people.
[ucblockhead is] useless and subhuman
[ Parent ]
This mostly applies by TheophileEscargot (2.00 / 0) #20 Wed Jan 04, 2012 at 01:26:10 AM EST
To SMEs (Small / Medium-sized Enterprises) not large companies, which can raise money relatively easily. But there's an argument that most growth in the economy comes from SMEs, not large companies which have often stabilized in size.

I think we've discussed this before, but a healthy capitalism is a process of creative destruction: the problem in a recession isn't so much that destruction rises, more that the creation stops happening.
It is unlikely that the good of a snail should reside in its shell: so is it likely that the good of a man should?

[ Parent ]
Just in time by jump the ladder (4.00 / 1) #13 Tue Jan 03, 2012 at 01:12:22 PM EST
Great for Toyota but it just shifts the warehousing to the supplier companies. The traffic is horrendous around the major Japanese conurbations so the "In time" aspect of JIT is managed by suppliers trucks being parked around the corner of the Toyota plant with the drivers frequently sleeping on their cabs until the order comes through from Toyota.

Other side effects... by wumpus (4.00 / 1) #19 Tue Jan 03, 2012 at 08:43:48 PM EST
In the book Speed Tribes (a book about Japanese youth toward the end of the boom), one example of effects of JIT methods in building motorcycles is discussed. When a sufficiently small order for parts goes down the line to the mom & pop companies that actually make the parts comes up, they simply request the parts from even smaller contractors. In this case, our heroes are skateboarders who bring back the parts as complete motorcycles.


[ Parent ]
Productivity and Operations Management circa 1990 by jimgon (4.00 / 1) #22 Thu Jan 05, 2012 at 07:49:01 PM EST
We had to study JIT when I was in college.  Lots of fun stuff trying to figure out the exact point in time in which you need to order your replacement stock in order to keep the assembly areas from running out of raw materials and halting production.  That was back when management degrees taught you how to run a factory.  The class also covered quality metrics.  IT is just starting to catch up on the quality end to what I learned when I was trained to run a factory and I just shake my head and think back to college.  I never did work in a factory, but the management classes have made being a PM much easier. 

Zombies back is about right.  Sucking all the investment dollars out of the system preventing economic recovery I don't entirely agree with.  The zombie banks are indeed draining capital, especially into the derivative market, but they aren't the prime thing holding the economy back.  I'm squarely on the demand side of things.  There's no need to invest in capacity since there isn't sufficient demand to require addditional capacity.  You don't increase expenses unless you have sufficient demand to cover the expenses through additional sales and thus increase profit. 

Technician - "We can't even get decent physical health care. Mental health is like witchcraft here."
Quality by Herring (4.00 / 1) #23 Fri Jan 06, 2012 at 10:06:17 AM EST
I worked in engineering for a bit. I learned some stuff about quality systems and process capability studies and all that. Great for looking at repeatable processes and materials traceability - we used to make bits for nuclear reactors and the like so quite important.

I get the feeling that a bunch of fuckwit MBAs came and looked at all that and though it could be applied to anything - including software engineering (that's probably about when the term came to be used over "programming"). This is clearly bollocks. Producing a peice of software is a one-off task concerned with managing complexity. Nothing like producing the 1,000,000th widget.

You can't inspire people with facts
- Small Gods

[ Parent ]
Six Sigma is the worse by jimgon (2.00 / 0) #24 Sat Jan 07, 2012 at 06:38:33 PM EST
I had to sit through a presentation at work, new employee orientation stuff, and there was a presentation on quality that included a high level and brief discussion of the concept of Six Sigma.  Just another process improvement scam.  I've been the victim of every major management fad on process quality since the early 90s.

Technician - "We can't even get decent physical health care. Mental health is like witchcraft here."
[ Parent ]
I can only receive | 24 comments (24 topical, 0 hidden) | Trackback