- il y a 2 jours
The best-seller “The Goal” by Eliyahu Goldratt described in the form of a novel how the Theory Of Constraints (TOC) could save a factory by quickly boosting its performance. But how exactly should one implement TOC (combined with Lean) in production? Philip Marris, the CEO of Marris Consulting, presents his experience in implementing TOC and Lean over 250 times in many different industries.
He presents the guiding principles: the 5 focusing steps. He presents 2 difference examples: one in the automotive industry and another in the aeronautics industry.
This is an 43 minute edited version of the full presentation of 56 minutes.
Have a look at our website: https://www.marris-consulting.com/en/
Have a look at our online courses: https://e-learning.marris-consulting.com/
Check out our LinkedIn company page: https://www.linkedin.com/company/marris-consulting
Subscribe to our newsletter: https://marrisconsulting.substack.com/
Marris Consulting is a management consulting firm.
Our motto: Factories, People & Results.
He presents the guiding principles: the 5 focusing steps. He presents 2 difference examples: one in the automotive industry and another in the aeronautics industry.
This is an 43 minute edited version of the full presentation of 56 minutes.
Have a look at our website: https://www.marris-consulting.com/en/
Have a look at our online courses: https://e-learning.marris-consulting.com/
Check out our LinkedIn company page: https://www.linkedin.com/company/marris-consulting
Subscribe to our newsletter: https://marrisconsulting.substack.com/
Marris Consulting is a management consulting firm.
Our motto: Factories, People & Results.
Catégorie
📚
ÉducationTranscription
00:04to present myself I my name is Philip Maris I'm the CEO of Maris Consulting I'm a horrible
00:11consultant so as a precaution you must assume that everything I'm saying is either a lie or
00:18an exaggeration okay if I can say that as easily as that it's because if you look around at for
00:24instance all the videos we've had published there are quite a few clients that confirm that what
00:30I'm saying is the truth and I take that precaution because the theory of constraints when applied
00:37properly can produce extraordinary results unbelievable results so you might want to
00:43check by looking at the videos of the various client testimonies because results the speed
00:49and the size of the results you get with proper theory constraints just hard to believe unless
00:55you've understood how it's done and that's what I'm going to try to do today by explaining not
01:00only the results you see in the real world but what the underlying logic was and I'm going to
01:05be talking mostly about theory constraints today but are in our world in the way we do things with
01:11our clients it's to keep it simple a mixture of about 50% lean and 50% theory constraints it's
01:17one
01:17of the trademarks of our company where we fight against the silos or the the sects of one way
01:23of doing things and we believe there's a lot of value in mixing lean and theory constraints in
01:30particular but why not some six sigma some agile some trees and other things right and I've been
01:35trying to apply lean many many different kinds of businesses adapting the ideas from Toyota to making
01:42hamburgers or rockets or planes or whatever adapting to various environments and that's one
01:47of my fascinations okay to keep it very simple I'll come back to it the theory constraints tells
01:52you where you should use lean first I joined the L.E.L. Goldratt just after we published the goal
01:59and I've also been fascinated by applying the theory constraints ever since so we've been applying this
02:06cocktail of theory constraints and lean for many years I founded Marys Consulting 15 years ago in a few days
02:13it's hard to describe exactly our perimeter in terms of kinds of business because as long as there's a
02:19physical object in there somewhere we're involved as you can see it can be in fast food or it can
02:26be
02:26making rockets and between the two you've got cars trains planes luxury watches expensive handbags and many other
02:36things so the theory constraints uh gold ratism that is L.E.L. Goldratt and uh what he invented created
02:47or adapted from previous ideas what he built was theory constraint it would probably have disappeared by
02:54now if there hadn't been uh one remarkable initiative by Eli Goldratt himself uh which was to write a novel
03:02um the goal
03:03and that book is now uh by far the biggest bestseller in terms of management books if you want to
03:10accept
03:11it as a management book because it certainly takes the form of a business thriller or a novel since
03:16it's it's just about a hit eight million copies sold it's now translated into 32 languages inside
03:21it's in universities it's still pertinent even though it was written before internet before china
03:26before sap and so forth but what's frightening is it's very fresh today right so for you some of you
03:33who haven't uh read the goal uh do it uh now it's also a twist because of course uh the
03:42goal being a novel
03:43is not a textbook and so maybe another pitch or another subtitle for today would have been you know
03:49you like the goal how do i implement it and that's what i'm trying to try and explain right if
03:53i can try and
03:54summarize in just one page the theory of constraints let me start with the symbol the pillar of the
04:00annual budget of organizations when you've finished it you sign it and you validate your budget for
04:05next year uh you have if you've done your work properly determined what capacities you need in the
04:12various parts of your organization right and you decide to hire people or buy new machines or not
04:18depending on on one side your forecasted your estimated demand your guesstimate to the future
04:2412 months what am i going to sell right but of course as you well know uh it's harder and
04:30harder to
04:30forecast the future right there's so many things going on that we're living in a vuka world it's it's
04:36volatile it's uncertain and so forth right we don't really know what is going to be happening in 12 months
04:40time do we um so you've got this great uncertainty about what you're going to actually have to produce and
04:47sell
04:47right and it's just going to get worse and worse right there's no there's no end to that where you
04:51think that it's terrible because you only have two months visibility uh in 20 years time you'll only
04:56have two weeks visibility right it's just going to get faster and faster and faster so that's one side
05:01of the equation and on the other side of the equation you have your capacities and your capabilities
05:06and the inertia of those capacities and capabilities has not really changed over the past 50 years
05:12that is to say it takes you about the same amount of time as uh 30 years ago to buy
05:18a new machine and
05:19install it or to hire somebody and train them right it's a question of months or years depending on
05:24what kind of thing we're talking about right so how can we possibly succeed in getting our budgets right
05:31and distribute the work equitably if we don't quite know what we're going to be doing next year and the
05:36speed at which we will adapt our capacities and our capabilities is limited it's much too slow right
05:42so the main message for me in theory constraints is watch out your budgets have guesstimated something
05:48about distributing the work equitably in your organization but it is wrong right don't believe
05:54when you sign it that you've done it because it can't be done anymore right and in fact you tried
06:00your
06:00best right okay but somewhere in the system somebody's going to take the short straw right
06:06they're going to have a larger workload than everybody else in the system okay that will
06:12be the constraint of the organization and it will be surrounded by people where they have excess capacity
06:18right you have not distributed the work equitably whatever you think of your budget whatever your
06:23budget says you've played around for the numbers to hide that fact maybe the truth is there is
06:30a resource in there that's got a greater workload than the others that is the constraint in your
06:34system okay and that's the starting point of the theory of constraints and what fascinates me
06:39is what i've just said is obviously just going to be truer and truer year by year as visibility decreases
06:45and our inertia continue to be uh significant okay so if we have that we have a constraint in the
06:53system
06:54as you'll see i i often present things as a series of tanks with water flowing through them
06:59that's the logo of the company and all that the quantity of water that's going to flow through
07:05this thing is determined by the bottleneck and so rule number one it is the constraint that determines
07:10the performance of the system right and rule number two which is just as difficult non-constraints
07:16there's no point in improving their performance or increasing their capacity because they will make
07:20no difference to the overall performance right so an hour gained on a non-bottle deck
07:26is a mirage it will absolutely not change the performance in the system okay which also means
07:33that uh trying to measure the efficiencies of all resources in your system is dangerous because if you
07:41do that you'll make sure that all machines working all the time and as we can see in this uh
07:46picture
07:47that means that the the stuff will pile up up to the bottleneck uh and people will just be keeping
07:54busy
07:54without being able to sell more okay so we have to learn to to stop working intelligently uh and
08:01this by the way is true of lean just as it is fear constraints can man system assume and ensures
08:07synchronization by telling a workstation that it could it should stop because its client its downstream
08:14operation has not got any demand anymore right so lean and field constraints do the same the only way
08:19of synchronizing things in today's world is uh to learn to stop intelligently at the right point at
08:27the right time to keep everything synchronized okay the way they do it lean and talk it's different but
08:32the idea is exactly the same okay so we get to the sentence that is and i want to come
08:37and go around
08:38the sum of local optimums is not equal to the global optimum right it is not a good idea to
08:44seek full
08:44utilization of one's modern bottlenecks and about 80 90 95 percent of your organization is a non-bottleneck
08:52so it's quite a big cultural change or mindset change that's involved in the theory of constraints
08:57right you don't want to keep everybody working all the time because if you do you get the results that
09:03you have in this picture where everything's piled up in front of the bottleneck that means a lot of
09:07working process which means long lead times and a big mess and the way one implements the
09:14theory constraints uh the five steps of the theory constraints the first one is to identify the
09:19system constraint it's quite easy to do in production as we will see today uh it's a little
09:25bit more subtle in uh projects environments but that's subject to other webinars once you've identified
09:31it you decide how to exploit it now step two and step four are close cousins exploit is what you
09:37can do
09:37immediately without wasting time without spending money right what can i do in the next hour in the next day
09:43in the next week maximum uh to improve the performance of the bottleneck and it's all about focusing
09:48rather than trying to keep everything working everywhere you just go to that resource that bottom
09:53that source and you ask it what it needs and very quickly normally you can get 20 40 100 percent
09:59more out of that resource simply because you're focusing all the energies for all the services on
10:04that resource and uh you'll no doubt find uh it's quite easy to improve i repeat at least 20 percent
10:12and more likely much more than that subordinate is uh to make sure that you don't launch more into the
10:20system than uh the constraint can digest and handle okay um and it's difficult because that's where we
10:28have to learn to stop working or stack a new light that impression everybody must keep making noise
10:34with their machines everywhere all the time and that's the best way to make money okay we'll come
10:38back to it elevate the close cousin to step two is you continue increasing the capacity of your
10:44constraint but this time you can spend time and money okay if you do that little by little the
10:50constraint will produce more and more and the constraint will go somewhere else in the system right you
10:55have turned them what was a non bottleneck into a bottleneck you go back identify it and start
11:00again an overview of all the different dimensions of the theory constraints is here we're going to be
11:05talking about the red box today which is the historical origin of the theory constraints and
11:10novel the goal another important uh solution as some call it or element of the theory constraints is how
11:17you manage projects with something called critical chain project management there's a lot of stuff
11:20uh in on our websites about about this it's about 40 of our business is using critical chain particularly
11:28in new product development or other areas such as uh mro of airplanes and stuff anyway not the subject for
11:34today then there's a packet around uh supply chain management uh not very clear what is the official
11:43net label name uh within the theory constraints about this uh often called replenishment uh and a very close
11:51cousin called demand driven mrp uh the the founders of uh dd mrp uh chad smith uh did uh 20
12:01or 30 years
12:02of uh theory constraints before uh creating the demand driven mrp so there's a big overlap in the way what
12:09they're
12:09proposing and what you find in uh supply chain management of the theory constraints this last
12:15point i wish i could i could talk about more today because it's it's one of my passions it's when
12:20uh you
12:21have a constraint in marketing and sales i'll keep it very simple before crying about the fact that you
12:27don't have enough work in your factory uh because of the market or taxes or the price of labor or
12:32whatever
12:33make sure that in fact your marketing and sales is not the constraint are you sure that you're making
12:39enough noise marketing uh and uh your sales process is efficient because quite a few organizations
12:46uh in fact before the market being too small it's because nobody knows that you exist so you're not
12:52getting the phone calls and you're signing the orders and therefore you don't have enough work
12:55but the constraint was uh in theory constraints the view of things uh probably a marketing and sales
13:01constraint okay so before saying that the market your market is too small can please confirm that
13:08you're doing enough marketing the line below is slightly different different kind of objects uh
13:13uh the this one here throughput accounting watch out the word is is uh unfortunate but it's the
13:19historical one it is not an accounting system okay uh the accounting system you have to do according to
13:26the laws of your country and so forth um it's a financial decision making system uh with things like
13:31throughput inventory and operating expense and uh based on the existence of bottlenecks it leads you
13:38to take different decisions for instance on the cost of things and the good sales price or margin of
13:44one product compared to another so uh it will encourage you to rethink your your product mix for
13:50instance okay value added computing very quickly because it's in gray it's my personal opinion you're
13:55going to often find it or never find it in other people's presentations as a theory of constraints
13:59it's just that i think that ellie goldratt wrote a lot of very interesting things about data and
14:03information thinking processes is another special beast because uh that's why there's this arrow
14:10it's goldrattism ellie goldratt was passionate about teaching the world to think or be more logical
14:16in when thinking and just gave the thinking processes it's not theory constraints per se if you
14:24because it doesn't involve the existence of bottlenecks and capacity constraints and stuff right
14:29it's more about how you analyze things how you decide things uh how you take decisions and try
14:35to make that much more logical because we're humans are not robots and uh logic is a lot less present
14:43than we think right when i say that as an engineer i thought i was logical but i discovered all
14:48that
14:48thinking process stuff the uh scheduling mechanism the flow management mechanism of theory constraints
14:56uh version one there are there are variants but this is the historical route it's called drum buffer rope
15:04the drum is to say that uh the first thing you do is to manage uh the factory according to
15:11the capacity
15:11to your bottleneck so it will give you the rhythm hence the name the drum and that should be the
15:17case on all horizons of planning whether you're doing strategic planning medium-term planning or
15:21day-to-day execution right the constraint gives you the tempo of your organization that's the first
15:26thing you do is decide how i'm going to use the bottleneck right um once you've done that you have
15:32the
15:32rope the rope is there to avoid uh launching things into the system too early to keep everybody busy
15:39uh for instance we know that before getting to the bottleneck right between the product the raw
15:43material launch or the batch launch and the bottleneck there are three operations to keep it simple
15:48each of these operations takes one day okay well if i have uh scheduled a batch here of a particular
15:55part in 10 days time and i know it takes me three days to get there i need to launch
16:00that batch into
16:01the factory seven days before right 10 minus three so that's just to ensure that you only launch into the
16:07system what the drum or the bottleneck can uh manage right if you do that the problem you will have
16:14is that any kind of problem whether it be quality downtime absenteeism or overloading local overloading
16:21for a few hours of one of these resources will make that batch late and therefore you protect your
16:27bottleneck uh by ensuring that it always has work in front of it and you do that by having a
16:32buffer
16:32measured in time so you would schedule products to arrive with one hour one day one week depending
16:39on the kind of industry you're in um so the bottleneck is protected against variability and problems
16:46of the non bottlenecks okay so if you do that you have a system where you have maximum throughput because
16:54you've managed the bottleneck accordingly it's scheduled accordingly and you have minimum inventory
17:00because you're only pulling the stuff into the system just at the right moment okay so that's
17:05you have you've put the dangerous part of your system under control after the constraints between
17:10the constraint and finished goods and the work being finished not much can go wrong because you have
17:15excess capacity right so uh that's not going to be a big problem you will have a second kind of
17:21buffer
17:22which is to protect you against variability and and problems in these resources they have capacity but they
17:27might break down and so again you have a second buffer whose role is there to protect your due date
17:33performance so that you meet your promises for your clients and it's the same it's measured in time
17:37you will say for instance i must finish all my orders one day early right i've added one day lag
17:43here in my system and i aim to finish everything one day early so that i can absorb up to
17:48one day's problems
17:49so let's uh start now with uh an example from the automotive industry this is one of the largest oems
17:57in
17:57the world with over 120 factories now i believe it was one of the first to begin its lean journey
18:03in the
18:04end of the 1970s i've worked in many organizations it's the leanest outside of Toyota that i've ever seen
18:11uh so they've got you know all the culture all the tools it's uh they're really good okay uh and
18:19we're
18:19going to talk about a factory uh which has a thousand people it manufactures alternators um and
18:27these thousand people are broken up into 20 autonomous production units okay the problem was there was a
18:35steep increase in demand for cars on the american continent and this factory couldn't keep up
18:40and it put in difficulty or all the assembly uh uh of cars in on the american continent so everything
18:49was doing wild uh they were already working uh 24 7 and they were already investing in new machines to
18:56double the factory but uh they've been working at it for the past six months and they were still uh
19:01late
19:02and stopping detroit as i've also there was a huge uh tension on in this factory in the middle of
19:09uh the middle of mexico we were asked to implement the field constraint i wish i'd give you the longer
19:14version about that discussion at the boardroom level of a company that swore by lean uh that decided to
19:21transgress and try something like the theory constraints but that's what happened you know
19:25we tried everything in our lean toolbox we're still getting killed by by general motors uh we have to
19:30find a solution so off we went um we looked at the factory within the factory we identified which of
19:38the autonomous production units were the bottlenecks okay there were two basically uh but both were
19:44photocopy one of each other so i'll just look at one of these that we just did this twice okay
19:49they
19:50made the rotor that's to say the in the rotating part within an alternator uh which you have the picture
19:56here okay and this autonomous production unit was basically a u-cell okay uh with products going in
20:05here going around on the conveyor belt and coming out here with about 10 different operations about 15
20:12people um and a cycle time of he was 17 seconds initially making about 6 000 units a day so
20:22i spent some
20:23time on the shop floor uh over four hours in fact just making sure that i understood what
20:30this uh unit was doing uh in this company's system they already have a symbol on a one of the
20:39machines in such a line which says watch out this is the bottleneck okay so and this balancing
20:44machine here uh had that sign so this was meant to be the bottleneck balancing machine by the
20:50way is once you assemble the shaft the bits and pieces right of this thing uh go through the
20:57impregnation in the oven and you do various things it's nearly finished here it the thing's
21:02going to spin around and yet it's got forged parts so the material is not equitably distributed
21:06around and so just like you uh balance a car tire a car wheel uh the balancing machine spins it
21:14around
21:14very fast measures the vibrations and can work out where to remove a little bit of mass okay and it
21:19will drill one or two holes in a specific place uh to remove the mass and make sure that central
21:26gravity is well centered it's the bottleneck and it does that right the question is how could you
21:31increase the throughput and the productivity of this line by 15 in 15 minutes and 15 minutes why 15 minutes
21:41because in 15 minutes you can't get out the toolbox right all you can change is a rule what rule
21:46can
21:46you possibly change well think back to what i said a few minutes ago uh the drum buffer rope okay
21:52well
21:53drum no interest here since they were just producing uh uh high or high volumes of a uh of a
22:01rotor uh so
22:02that was taken care of what the uh they didn't have in the system was they didn't have a buffer
22:08right we
22:09have a system with one piece flow with a conveyor belt all around right and these parts are about the
22:15size of a pineapple no but smaller uh and go go around and get assembled by little okay going around
22:25on the conveyor belt okay and what i'd noticed after spending four hours just watching was that this
22:31resource was quite regularly uh stopping because there were no um parts uh feeding it right because
22:41this machine although it had excess capacity regularly had problems not for very long sometimes 15 seconds
22:4745 seconds two minutes and so forth okay but as a result it was stopping this machine quite regularly
22:54okay and you could see that it was seconds right it wasn't a quarter of an hour and it wasn't
23:00apparent in any
23:00of the data because the system here uh to uh declare stoppages starts when you have a three minute
23:07stoppage less than three minutes you don't load it down so in three minutes and more stuff so it wasn't
23:12in any of the data but you could see it you just waited a long time watching the machine seeing
23:17how
23:18many times it happened okay so um we install the buffer now let's be clear what did i do to
23:26install a
23:26buffer that's why i'm using this example i find it fascinating okay normally right what i've said
23:32previously is that because there's excess capacity up to the bottleneck right things should pile up in
23:38front of the bottleneck right that's what we have in the logo of the company right you should have a
23:42big
23:43pile of stuff here it wasn't there there wasn't a big pile there on the contrary i repeat i saw
23:48regularly
23:49this machine stopped because there were no parts right and that was because their understanding of one
23:55piece flow is that you should avoid having piles of things everywhere right and so magically everybody
24:02had in fact slowed down to the speed of the bottleneck right already and of course they had because
24:11if everybody produced according to their own capacity some of local optimums right they would keep
24:16producing and producing more than this and parts would start falling off the conveyor belt
24:22right so in fact everybody in this system had already slowed down to the speed of the bottleneck
24:29right and this is true of all organizations i just like this example it's so obvious when you see it
24:35there uh drawn out like this but it's true of any factory right look carefully the logo of the company
24:42right then the quantity of inventory that is piled up in front of the bottleneck does not go to infinity
24:48right it stops at a certain level that certain level depends on a number of things uh the number
24:54of the amount of money in the in the bank account to invest in inventory or physical storage space
25:01all those sorts of things right it doesn't keep going on for infinity at one stage everybody learns
25:06in an organization when they are feeding a bottleneck how to slow down to get to their speed right which
25:12is
25:12why i insist here because sometimes people think that implementing theory of constraints is going to get
25:17people to work less not really right they've already slowed down it's just they've since you don't
25:23want them to just to to stop working they've just cheated with all the numbers uh to make it look
25:28as
25:28if they're working all the time right but they have already slowed down to the speed of the bottleneck
25:34otherwise you would go to infinite capacity right put that in mind please because people don't realize
25:39that very often right they think and we're going to stop using our non-bottlenecks uh yes uh but you
25:46already have okay um so that's what i wanted to do in this particular example uh the the company
25:56the organization already slows down the speed of its bottleneck i was inventory would go to infinity
26:01it would be right just to complete this story uh we that was what we did in 15 minutes but
26:09we didn't
26:10stop there uh we then improved even further the performance of the bottleneck by making sure it
26:16worked 25 hours or 24 hours a day right 100.00 uh of the time right we had to deal
26:23with reoccurring
26:24small breakdowns right they had like any machine if you look very closely they had lots of uh micro stoppages
26:30and stuff so we got the uh the maintenance teams to focus on this machine and improve it
26:37it was the priority of all support services right including maintenance right both preventive and
26:43curative so if the maintenance team was working at the other end of the factory uh it pulled apart
26:48a machine that had a significant fault and they'd been working at it for eight hours and it was just
26:53about to finish to put the machine back online the bottleneck would break down they would drop their tools
26:58run over to the bottleneck and fix the bottleneck and the guy in charge of the other autonomous
27:02production unit would scream and yell come back yeah you've nearly finished you stay another quarter
27:06of an hour and i can have my machine back no priority number one is the bottleneck and all support
27:11functions whether it's engineering maintenance human resources or whatever that should consider that
27:18resource as the priority we also did a lot of improvements to the production process as i say this
27:25balancing machine it's span and then it drilled a couple of holes uh just by optimizing the distance
27:31that the drill bit went back and forth it was going back too far and we just changed the programming
27:36so
27:36it would uh waste less in in movement i think we gained there so i can't remember it was half
27:41a second
27:41or point two of a second but it didn't matter it meant that the whole of that one thousand person
27:46factory
27:46was uh improving by that much right it is the bottom so you really focus on it uh completely
27:53and uh when you do that of course anybody can get more out that was done okay and we did
27:59other stuff
28:00all over the place reducing the scrap rates uh the long breakdowns uh make sure that there were
28:06all the proper critical spare parts for the bottlenecks so that they wouldn't have
28:10um a serious failure because it would be a big would be a heart failure right if when the bottleneck
28:15is
28:16stopped uh you lose those hours forever okay you can see lots of different videos on this case on
28:22the youtube channel in french in english uh short and long okay point i will uh underline here is that
28:30uh what we had to change was a the sacrosanct one piece flow rule uh and implement a buffer uh
28:38the
28:40idea there is that i think it leads to a lot of quarreling between consultants that leads nowhere
28:45uh both lean and theory constraints again trying to get to the same goal they have very fast flow uh
28:52and
28:53to uterus all showed us that if you do it right uh there's no compromise between speed quality and cost
28:59if you get it right get it right first time you get all three at the same time and theory
29:03constraints
29:03is the same right it's just that in lean you keep the system under tension by not having any kind
29:10of
29:10buffering any kind of stock anywhere okay and that keeps the pressure on people to improve the theory
29:15constraints puts a buffer in to protect production and therefore uh you're less stressed and the danger
29:22with the theory of constraint system of course that you fall asleep on your buffers as i
29:25i'll tell you that's why in many ways taishi owner was right because he keeps the pressure on people
29:30by continuing to to to reduce the inventory and and have very little security left right and center
29:36which will keep pressure on the people to get rid of the root corners of your various problems because
29:41they're trying to do exactly the same thing it's just that the approach uh to get to that goal is
29:46different he did lots of other work lean work right now i say i'm talking about theory of constraints
29:52today but the theory of constraints was telling us where to act and then we've got out the lean
29:56tool box uh to to improve things right let's get back to those five focusing steps trying to use
30:01the backbone of today uh the first thing you want to do is to identify uh constraints uh step one
30:10identify right a critical step is it easy or not obviously people like me who've done it so many
30:16times i find it easy to do but it's true that many people hesitate and it's true that 80 percent
30:22in fact we've read the calculations recently it's more like 75 percent but the majority of companies
30:26are wrong with regards to where their bottlenecks are okay so let's have a look why uh the kind of
30:34reasons why people might get it wrong uh firstly don't look at your computer it doesn't have a clue
30:40uh go down to the shop floor and have a look okay and look for cues of work uh it's
30:46it's just physics
30:47right things will pile up in front of your capacity constraint okay and uh if you need confirmation speak
30:54more to the operators than the middle management why because middle management is paid not to have
30:59bottlenecks right so it's going to get very constipated about this question of bottlenecks because
31:03it's paid basically to to to get rid of them and so it's uncomfortable question where your bottleneck is
31:10the best they will be able to do say oh it wanders around right but what we're looking for is
31:15the
31:15structural bottleneck that's there year in year out or at least many many months um and not the waves of
31:21work okay and please be aware of if your bottleneck is too good to be true what i mean by
31:27that is if
31:27somebody comes and says hi mr mares come and have a look we're doing theory of constraints this is the
31:31bottleneck and the bottleneck is the most expensive beautiful machine in the factory in fact that is where it
31:37should be right in ultimately uh but a lot of the time it's not right so that's where it should
31:45have
31:45been but in fact what we're finding is it's nearly always in somewhere silly elsewhere you'll see some
31:50examples in them many many in our you know videos and stuff okay um second you exploit the constraint
31:58that's fairly easy uh or it's and it's easier in in in emerging nations than it is in in the
32:06old
32:06first world as i call it in england and france in germany and america and stuff because uh if you
32:11really
32:11want to get to 100 right you need to no stopping during breaks no stopping during shift changes first
32:16priority for labor shortages and stuff right uh and i can hear those in in the old first world saying
32:23this is not possible with our unions and stuff well frankly we do it and do it and do it
32:27again and uh
32:28for instance we've done it one of those unionized organizations in the world well i think the most
32:33unionized organization organization in the world which is the french railways right and uh that wasn't
32:39the problem so it can be done subordinate everything step three okay you have to learn to stop working
32:46and again i'm going to have a point of view on this which is slightly different from a lot of
32:49other
32:50um people who talk about this in the theory of constraints world where they say it's the fault of
32:55cost accounting where you try to get local optimums it's true that's there i think that more often one
33:02of the reasons that people have to have difficulty with learning to stop working intelligently is that
33:08they don't have the right attitude and the right relationship with their front line direct labor and
33:15they are treating them as people who can only produce and don't have a brain it's very close to
33:19respect for people in need what i'm saying and therefore if they stop drilling holes since you
33:25don't think they've got anything between their ears what are they what are you going to do with them
33:28right well uh if you want to get around that uh what you do is that you admit that they
33:35can't do
33:36anything for today's sales right they can't drill there are any holes anymore they've done that
33:41so what can they do they can improve things right and they can do maintenance they can
33:47uh uh do some problem solving some quite wise whatever whatever okay they can help contribute
33:52to making more money in the future they can't make more money now because that they've done the order
33:56book right and so uh it's a different attitude to people making sure that uh when they don't have
34:03any more holes to drill uh you you ask them to use their their brains to improve the company okay
34:09very
34:09simple something i could talk a long time about but it's my my opinion of very often why they are
34:15unhappy about people stopping to work elevates is very easy because uh the only difference between a
34:23child and adult is the price of toys so you're allowed to spend money at last it's easy to justify
34:28the return on investment because it's the bottleneck and therefore it will increase the sales of the
34:33company uh what's interesting is it can take time though if it's a long investment process or
34:38rare human capability uh that has to be trained up for many many years and stuff now or you're limited
34:45by regulatory improvements and stuff okay and uh the the other trick or or thing that you have to
34:55watch out for when you do this of course is if you over invest in bottleneck uh and increase double
35:01its capacity or whatever you're very likely to find that the bottlenecks move elsewhere and therefore your
35:05calculation was wrong so which leads me to what i consider to be uh the the second level of the
35:13theory of constraints if you think about the five focusing steps that i presented right you go back
35:17to step one right it's an intuitive process your bottleneck hunting uh i don't have a problem with that
35:23initially because i pointed out very often people have the wrong bottlenecks so you should get rid of them
35:29okay but it gets to a stage depending on the company in the first year in the fifth year i
35:35don't know
35:35where you stop step back and think okay i accept i can't distribute the work equitably in my organization
35:41but why don't i choose the best or the least worst bottleneck least worst constraint in my system and add
35:49excess capacity before and after it so that it remains the bottleneck i have a stable system okay i think
35:55this is
35:56level two theory constraints and i described uh my definition of the best constraint which is the
36:03resource which would take the most money and or time to turn into a non-constraint right so if it
36:08costs
36:08a fortune to have excess capacity somewhere of course that could well mean that's not a bad idea for it
36:14to
36:14be the constraint maris uh fit maris point of view ultimately decide what your constraints should be
36:21and uh organize your your capacities according to you right the second case study aeronautics industry
36:28uh often when i present this one people think ah we need to improve uh we're not that bad okay
36:36as far
36:37as i'm concerned these are just perfectly normal clients with a well-known company uh you can see
36:41the videos and there's nothing unusual they were not worse when they started than other companies
36:46that industry so if you're thinking at the end of this case study that that was easy because we're not
36:50like that watch out um the as far as i'm saying a perfectly normal average company um they were producing
36:59flight control systems that make your airplane go up or down uh start the report okay they were they
37:05were late they were stopping uh european airplane aircraft manufacturing because of their poor due date
37:12performance um we were asked to step in and have a look when we arrived they said yes we have
37:18bottlenecks
37:19we know all about the theory of constraints we have i think it was not seven it was 12 in
37:23fact
37:23machines that are bottlenecks and they would say look it's not a problem because if you look here on
37:27these graphs where the red line is the capacity of the machine and the blue line is the workload
37:32well our machine number five is going to be okay in one month's time machine number four is going
37:37be okay in two months time um and several right so come back in three months time and we solve
37:43the
37:43problem the reason why i was there was the top management had got sick and tired of hearing the
37:48same explanation or excuse for the past three years and they didn't really a word anymore and they
37:53wanted a fresh interview so when we go walk down into the factory more or less immediately within
37:59less than two hours certainly maybe even within half an hour i can't remember um we find a big queue
38:07of
38:07work of hiding in plain sight down the main corridor of the factory a huge pile of products i'd say
38:13about three million dollars here uh something like that um and just hidden in boxes but it's very expensive
38:19stuff there were 370 of the 1000 something work orders 35 of the inventory was blocked in front of this
38:28nearly all the urgent orders were blocked here where was it in front of quality control okay i've
38:33changed this example on purpose because it is extraordinary and i cannot insist enough within
38:39this webinar to get this idea through the number of companies today whatever their business aeronautical
38:46medical food etc etc etc it where the bottleneck is in quality control or testing okay and it is the
38:55worst place to have the bottleneck that's why i repeat people are often getting it wrong saying look at
39:00my expensive machine this is the bottleneck when in reality it is in quality control right that's why
39:04i chose to have the second example in this webinar today so often uh it is quality control i did
39:12a
39:12teaching course intercompany for two weeks ago with five different companies and all five their bottleneck
39:19was in quality control or testing okay that's why this example is here watch out it's just the worst
39:25place in the world right to have quality control as a bottleneck it's just madness well so what we did
39:30was exploit that constraint uh in this city example photocopying machine of the quality control uh
39:38department of 15 people had broken down several months before so whenever they need to photocopy
39:43their documents for their quality reports and stuff they have to go and borrow the machine from the
39:46accounting department uh 500 yards away they would have to wait especially at the end of the month
39:52and they were wasting time going back and forth to the photocopying machine so we just took the
39:56photocopying machine brought it to the bottleneck right the bottleneck is the most important thing
40:00in the company and we invest the roles uh accounting now to come and ask permission if you could possibly
40:05use the uh photocopying machine of the uh quality control and uh we did some transfers and stuff easy
40:14stuff we're just focusing out of a 400 person factory we're just focusing on 15 people we asked them you
40:19what do you need how can we help you how could we get help you to produce more and we
40:23did a lot of
40:24things as you can see from the slide um and we also had to reduce inventory because they had much
40:31much
40:31much too much inventory and i'm assuming that many of you have much much much too much inventory this
40:38rather standard is a starting point with our company so how to reduce the quantity of work in the
40:43factory we use the two for one rule we'll have a look at the video it's nice and simple and
40:48it's
40:48beautiful and again i think it's one of the trademarks of marriage consulting if you have too
40:53much stuff in your factory you want more stuff to come out that's going in and so the two for
40:58one rule
40:58is for you need to get two things to leave the factory before and one thing can go into the
41:03factory
41:03you can use the unit you want for instance work orders that's what we did here they have to finish
41:08two
41:08work orders before they're allowed to launch a new one okay have a look at the video it explains it
41:12in more
41:12detail uh it's beautiful everybody loves it once they've understood and it's simple common sense
41:19so that's what happened uh that was before right with piles of stuff that was after or just before
41:25we of course got rid of the shells because shells are bad for people's health right they want to put
41:29things on them uh we basically uh had a huge impact on productivity overall throughput and productivity
41:37increased 30 percent in two weeks and went on in two years they in fact double productivity
41:42and throughput and the the problems of shortages more or less disappeared at least logistically
41:48speaking what was left why there was still shortages because of quality problems okay which
41:52took longer to fix things they they went on uh and on and on and on when we started it
42:00was the
42:00greatest loss making uh factory within the 60 of the uh saffron group uh very quickly they became the most
42:08profitable uh they uh massive numbers like uh labor productivity they reduced lead times initially from
42:17nine months to 2.8 months they won the 2016 award in the company and they won the best supplier
42:22award from
42:23airbus etc etc etc etc okay and uh this slide here just to show that we didn't just do theory
42:31constraints
42:32right you see in there there's some there's some smed there's some reorganization there's some layouts
42:38there's some stuff like that so we were also using lean the theory constraints was helping us to decide
42:43where to apply uh those ideas okay and that's just the reduction in inventory as as time goes on right
42:50that's the two-for-one rule showing you its power as inventory dropped massively massively
Commentaires