Episode 24

Managing job shops – Sustainable solutions to resolve core conflicts

Category :  Manufacturing & Supply Chain

In the last episode, we had looked at the troubles a job shop faces day-in-day-out. We also looked at the core conflict behind these problems and explored the reasons why some of the common solutions add fuel to the fire.

This episode is the second part of the same discussion to discuss the direction of a sustainable solution using principles of flow management . Tune in!

Transcript
Shubham Agarwal : Hello and welcome to the CounterPoint podcast. Today we are going discuss about job shops in the manufacturing domain. Managing job shops can be very difficult and complex. Many of them suffer from chronic issues like continuous expediting, month-end or quarter-end pressures, fluctuating delivery lead times, shifting bottlenecks and ever increasing demand for more and more resources and all of this which and all of this translates into poor visibility into operations for the top management and last but not the least, the stress it brings it with itself for each and every one in the company. So you know there are many challenges that the job shops across face in order to make sure they deliver the required material to the customers in time every time.

Today we have with us , Karthik eyan who will help us reveal the inherent simplicity behind the visible complexity of managing job shops. Karthikeyan has more than 8 years in transforming variety of engineered to order, job shop environments. Hi Karthik. Welcome to the CounterPoint podcast. How are you?

Kartikayan Ravichandran : Hi Shubham morning. I’m good. How are you?
Shubham Agarwal : I’m great. Thank you so much. So Karthik, before we discuss, you know about the kind of complexities that we can see in a job shop environment, it is important that we first understand what is really meant by a job shop. So could you help us give us a quick view of what a job shop is?
Kartikayan Ravichandran : That’s right Shubham. So, the simplest way to understand the job shop is that it is a kind of a manufacturing system like you said where a set of distinct machines are clubbed together in different work centers based on their technological capabilities Okay, they work on one set of jobs or customer orders at a time and upon the completion of which they move on to another set of jobs either from the same customer or different customers Okay, that is why they they are called job shops. So, if I have to give an example, say consider an equipment manufacturing company, where the final equipment happens after an assembly operation which consists of both in house and say procured components. Now, the in house production of these components are done in job shops, which may be consisting of a cluster of drill machines, CNC machines, press machines, and so on. Now, let me tell you the beauty of these job shops. The beauty is that the same set of machines which are present are capable of working on and producing a tremendous variety of end products. So the amount of customization that is possible is huge. And that is also one reason why these job shops are also termed as custom manufacturing environments
Shubham Agarwal : Yeah that is the good part. Tell the other side of the story
Most of the job shops because of the way that they operate, they actually struggle to grow comfortably year on year, actually, many of them struggle to even generate profits. Not just this. In fact, if you ask me, many of these job shops are unreliable, they don’t enjoy the customers loyalty. In fact, many of them end up paying penalties to the customers on account of late deliveries, which which creates a very bad relationship with the customers, you can imagine that
Shubham Agarwal : Right. Yeah.
Kartikayan Ravichandran : Apart from this, they also struggle with a very low capacity utilization. They struggle with a very high system inventory, both work in process as well as finished goods. And at the same time, the relationship with the suppliers are also actually not good. So so you can actually imagine the life in a job shop for the people who are working there is actually pretty, pretty difficult. It’s not simple. It’s pretty challenging
Shubham Agarwal : For a better understanding of the complexity , can you elaborate the issues from eyes of various managers involved in operations
Kartikayan Ravichandran : For a better understanding of the complexity , can you elaborate the issues from eyes of various managers involved in operations
Kartikayan Ravichandran : let me sort of give you a glimpse of how a day in a typical job shop looks like. Okay, sure. So we’ll start with the planning manager, the planning managers start getting frantic calls from the customers right from the morning. These customers are demanding to know about why their urgent orders are so delayed and still not delivered to them. Apart from this, they want to know how much more delay they need to tolerate. And many a times Let me tell you these customers don’t forget to tell the planning guy that if the orders the delayed orders are not dispatched ASAP, the future orders will be at stake. So you can imagine Yeah so you can imagine the kind of pressure that the planning manager is having on him Okay. So, under such situation I mean what do you think is going to happen?
Shubham Agarwal : I’m sure the planning manager does not ever give the right you know time or date to the customers and it only keeps building up I think
Kartikayan Ravichandran : yes the pressure builds up and what he will try to do is he will try to pass on this pressure and urgency to the production guy right. So, what will happen next in production is all the workstations wherever these orders are stuck these delayed orders are stuck the production manager will take an unplanned setup he will try to keep aside whatever was already running and they will try to take these urgent orders in production okay. So they will basically try to expedite this orders. Now, let me tell you in a job shop there are at least there can be as much as high as three to four unplanned setups in a shift and this happens in almost every shift okay. So you can imagine the extent of capacity loss that is actually undergoing in this job shops just because of this unplanned setups right and let me tell you what happens next because because of these urgencies, the orders which are kept aside and now they will sit in the WIP for a longer period of time that will actually go and increase the WIP in the shop floor. Okay. So this is as far as production is concerned. Now let’s move ahead a little bit and let’s assume for a moment that the orders are completed and they reach the dispatch area Okay, okay. Now, the dispatch manager once he looks at these orders, he realizes that all the items are all the products which are required to be dispatched together, which is actually not present some procured items are now missing because of this a dispatch cannot happen.
Shubham Agarwal : Okay, so even if the dispatches are complete, they’re only complete in bits and pieces
Kartikayan Ravichandran : Correct. Correct. So, the realization hits it hits the production dispatch guy that the full kit sort of required to complete the dispatch is not available, some say procured components are missing. So, naturally, he will now pass on the urgency to the procurement guy saying that you try to bring these materials as soon as possible Okay. Now, the procurement guy is in the same position what the planning manager was when we did some time back right now he will try to pass on these urgencies to be suppliers in turn, okay. Yeah. Okay. So, the suppliers now are being faced with the pressure to change whatever they were running they will change their setups, they will try to keep aside whatever they were trying to plan for that day and accommodate these urgency thereby sending what is required now and keeping aside something which was supposed to be sent today, but probably not not has been sent and it will go and hit the urgencies of the dispatch area probably at a later point of time
Shubham Agarwal : This sounds so chaotic
Kartikayan Ravichandran : it is just the beginning there is much much more than that,
Shubham Agarwal : okay. Sure. Let’s go Yes. So,
Kartikayan Ravichandran : you can see this kind of one thing is leading to the other and this cycle keeps on continuing Okay. Now, let us try to take a look from the customer side show when a customer is placing an order he doesn’t place some quantities of one product, but any customer order will consist of several quantities of several products okay. Most of this time these orders every product in the order needs to be dispatched together it cannot be dispatched right that is a customer requirement in many cases it is true most of the cases it is true that the customer wants the entire order to be dispatched together right. So, when this is true, and also what is true is some products of an order reach the finished goods earlier and the other products of the same order are reaching the finished goods area a little later, then what is a very common observation is that these job shops end up with very high finished goods inventory.
Shubham Agarwal : So I keep piling up whatever is ready, but then I do not have the entire order and hence I cannot dispatch it right.
Kartikayan Ravichandran : Exactly that is what happened. And that is why the system inventory in these jobs especially in terms of the finished goods inventory. It is pretty high. Right. Now, let me tell you how the production team adds on to this problem. Okay, it is very important. So I think you will, you will really appreciate how the production team is contributing. Now, any production environment, the work centers have a production target to meet. I hope you noticed this Right, right.
Shubham Agarwal : Yeah, everyone has targets everyone is subordinating to their own KPIs at every point.
Kartikayan Ravichandran : Yes, everyone has daily shift wise production targets. Daily production that they need to meet. In fact, in many companies, they go a step beyond and they link their incentives with these production targets, you’re right, correct. So, what this basically means is if the production team is unable to meet this target, a good chunk of money is being is going to be lost for them. Right? So, what do you think is going to be the behavior of the production team in such conditions,
Shubham Agarwal : I remember the famous quote – people behave in the way they are measured!
Kartikayan Ravichandran : If I was in that position, I will try to do the same thing that I will try to not lose the incentive. So, what they will, what they will end up doing is every work center will actually try to work on such jobs, such orders, which increases their productivity, while ensuring certain orders, which actually pulls it down. Very, this is very important. shubham you remember this, because we’ll come back to this again when we discuss the solution aspect towards the end, okay, sure. All right, okay. So, so, what this basically means is any order which increases the productivity of the work centers, it should flow right through the shop floor pretty fast. It sounds logical, right? Yeah. Correct. But in reality doesn’t happen you know, why
Okay, why is that
Yeah, in reality, what happens is the same job the same order, which impacts the productivity of one work center in certain way impacts the productivity of another work center in a different way. Okay, okay. So, suppose. So, yeah, so, the production productivity of different work centers are basically affected differently by the same orders because of which the work centers have a different batching considerations different work centers have different batching considerations
Shubham Agarwal : because every work center is concerned about how their work centers can be run the most efficiently the most you know, in the most fast manner
Kartikayan Ravichandran : Correct. Correct. because every work center we just discussed some time, but they have their own targets to meet right right. So, they will not follow a sequence which increases the productivity of some other work center they will actually try to follow a sequence for their own work center which increases their own productivity right right. So, because of this, what really happens is the same customer experiences a very varied and high lead times in different points of time for very similar orders. Okay, at the same time, the plant also experiences a very high WIP work in process inventory.
Shubham Agarwal : But this way of working, the WIP will keep building up and nothing will come out ever. It seems no-one if bothered about the entire flow of orders. It can’t be so bad
Kartikayan Ravichandran : as soon as the month end starts approaching, there is a tremendous pressure on both the production team and the sales team to meet their respective targets okay
Shubham Agarwal : Now you are refering to the total dispatch targets and not the individual work center targets
So, as soon as this target pressure starts piling up, what they will try to do is they will try to see all possibilities of dispatch which is available out of the current WIP out of the current orders which are pending and they will try to expedite and close all these orders towards the month end Okay, okay. So, as soon as this starts happening, what we will observe this month on month towards the month end there is a production skew and then there is a dispatch skew that is a tremendous amount of WIP will be sucked in and they will be converted into finished goods and they will be getting dispatched as well towards the end of the month because of which this queue had arises at the same time the beginning of the month, next month what we will observe this because the WIP has been depleted to a significant extent there is a significant starvation towards the beginning of the next month
Shubham Agarwal : This is so ironic, the work center managers are bothered about their local efficiencies and in month beginning they actually score low on this KPI.
Kartikayan Ravichandran : Yes, so the towards the beginning of the next month, many of the work centers if you go to a shop floor on a job towards the beginning of a month, and you will see that many work centers are actually sitting idle, they don’t have any material to work upon. Okay, okay. All this while there are still orders mind you which are actually delayed and they are pending to be dispatched.
Shubham Agarwal : So why are they not making them now? If they are starved?
Kartikayan Ravichandran : because they don’t have WIP they have been sucked in these WIP has been sucked in to produce some other orders. Then you can talk about okay. So everything sort of culminates towards the month-end where they try to steal material from other orders they will try to do something which keeps on getting dispatched. Okay. And the beginning of the next month they end up with very low WIP because of this, this starvation occurs. Okay, so are you able to appreciate the amount of complexity that is there in this environment.
Shubham Agarwal : Yeah, I think it’s all haphazard as in, I didn’t know that this is so bizarre. But I hope this this is all is there more to it?
Kartikayan Ravichandran : No, no, there’s more. So we discussed the beginning that job shops work with custom orders, right? every order is sort of custom for that customer. And no two orders are alike in production. Okay. So based on the kind of orders that has been received, and the extent of unplanned setups that is happening in the system, the bottleneck of the system will keep shifting from one workstation to the other. Right, right. So I’m not going to define bottlenecks, because I think it has it would have been discussed in one of the earlier episodes.
Shubham Agarwal : Yeah, but you could quickly quickly give a brief description of what a bottleneck is, that’ll really help the listeners,
Kartikayan Ravichandran : okay. So, bottleneck is any resource on which the demand on any particular day is placed more than its capacity for that day. Okay, because of this, there is a backlog that is getting created and the orders are getting stuck.
Shubham Agarwal : So in month end many reources turn bottleneck , while month beginning, there seems to be ample capacity
Kartikayan Ravichandran : So what happens is the bottleneck keeps shifting. And that happens because the kind of orders that are received are different, some orders may take more time in one workstation, the same order some other orders may take more time in some other workstations. So when the product mix changes, when there are some variability or Murphy that has hit the system, so there are chances that the bottleneck may keep shifting from one workstation to another. Right. So when this happens, and this happens time, and again, the production manager will invariably end up blaming capacity as the main reason why all this is happening. And sooner or later, they will decide to add capacity as the only way to get out of this mess.
Shubham Agarwal : But adding capex in this chaotic environment is never going give adequate returns.
Kartikayan Ravichandran : capacity enhancement measures will actually help a company when capacity was the real problem that needed to be solved, okay. But when capacity is not the real problem, and job shops keep adding to the capacity, all this while they are still struggling to grow. Remember where we started, many of the jobs are struggling to grow and now they have started adding capacity. What happens is the added burden of the capex further erodes the bottom line. Okay, yeah. Yeah. So, you can imagine that life in a job shop is actually not easy, especially when they are diagnosing the problem very incorrectly.
Shubham Agarwal : Right. So, you’re saying capex enhancement is not really a solution. The other way of solve the problem is to do precise scheduling of work-centers as per capacity , give due dates and and just follow the schedules in execution. Many software solutions promise that. Why not buy those tools and deploy them?
Kartikayan Ravichandran : . There are some more sometimes these jobs just go for advanced schedulers. So, these are usually used by companies to plan the manufacturing process accurately in such a way that optimizes their capacity utilization, okay. However,any variability would mean that the entire scheduler would need to be rerun again. And likewise, the planning also changes.
Shubham Agarwal : That would mean the original due dates are rescheduled again with every murphy that hits the plant. Why would a customer accept these kind of variations in original commitment with time the planning measures soon realizes that such advanced schedulers don’t really work. So there use may continue on paper, but in reality, they end up getting discontinued, so it doesn’t.
Shubham Agarwal : But, I am sure, people would have surely tried other solutions as well
Yeah, so that is one solution that I can think of another is usually companies can also go for line dedication where they dedicate a set of machines for a particular set of products and so on. The assumption is that this will improve the reliability because of this line dedication, but it but it is not really very simple to define the capacity very accurately, okay. The capacity that you have the capacity is dependent on several factors like the product mix which is running in the system at that point of time the variability in the processes, the manpower availability at that point of time, so, the capacity keeps changing, okay. So, line dedication also is not the solution which will actually solve the problem for this job shops.
Shubham Agarwal : We have reached a dead-end. So many problems and solutions don’t seem to work. What is the way out of this?
Kartikayan Ravichandran : So, if you try to take a step back, and we try to visualize all the problems that we discussed some time back, then what we will realize is no two problems are existing independently, one problem is actually leading to the other no problem is existing independently. So, what this effectively means is, if we are able to solve that one critical lever upon which we can work upon and can also be called as say, the root cause, you are able to identify that and solve everything else should fall into place automatically. So we don’t need 10 different solutions to take care of the 10 different problems that we are seeing, that is what it actually means. Okay, so, in this case, if you realize that if you’re able to improve the reliability of a system to such a significant extent, that is so high, that everything else will actually fall into place automatically.
Shubham Agarwal : So okay, if the production department becomes reliable, if the dispatch manager becomes reliable, if the planning manager becomes reliable, obviously, that shift from like, we saw, the connection would become reliable, and hence the job shop becomes a reliable job shop. Is that right?
Kartikayan Ravichandran : That’s right, just one difference. It’s not about individual departments building becoming reliable, the job shop overall should be reliable. That means when we are committing the date to the customer, it’s not a planning manager or a production team or a dispatch team that is committing it is the job shop the company as whole, which is committing that date. So even though the planning may do interesting, well, right, the production may do his job well, but the dispatch doesn’t do the job well did they are working on some other priorities, it will ultimately not end up serving the purpose. So, the company as a whole should become reliable, the company should improve its reliability.
Shubham Agarwal : Still, I have not got a clear answer to the problem. It sounds very generic. Let me ask you, why many companies are still struggling – if this was so simple
Kartikayan Ravichandran : even though people know that reliability is a problem that needs to be solved, they actually diagnose the problem incorrectly. So, what it means is they attribute the situation to a capacity problem rather than a flow problem.
Shubham Agarwal : I think the last part is not very clear you’re saying the situation might be they attribute the situation to a capacity issue rather than a flow issue. Could you elaborate on that a bit more
Kartikayan Ravichandran : correct. So, like I explained some time back job shops towards the end of the month, they undergo production in a dispatch queue and towards the beginning of every month, they will undergo they will experience material starvation because of which there is a under-utilization of workstations in the beginning of the month. under utilization means basically there is capacity loss okay. So in a system, if there is capacity loss, then how can capacity be an issue? capacity in job shops are already there. It is just being mismanaged by these job shops.
Shubham Agarwal : Okay, so That means when there’s capacity mismanagement, adding further capacity is not going to obviously resolve the mismanagement, you’re saying,
Kartikayan Ravichandran : correct. When there is already a mismanagement of capacity, adding further capacity is actually not going to help preventing the mismanagement by focusing on improving the flow of orders the shop floor, is real solution that we should be aiming for.
Shubham Agarwal : We have a new word called flow now. What creates bad capacity management and hence a bad flow in these environment?
Kartikayan Ravichandran : So, let me tell you something about this job shops, these job shops work in a monthly planning system, what basically means is in the beginning of every month, they will take out the entire order book, which is pending and then look at the commitment data to customers and they will create a monthly plan for both the production team and the procurement team for them to plan their processes accordingly. Okay. So, they will effectively make a plan the planning team will make a monthly plan and they will give that plan to the production and the procurement team.So, now, just remember the point that I mentioned some time back that the production teams have targets to achieve right, yeah, now, put these two things together. So, what do you think is going to happen the production team has a target pressure and they also has, they also have a visibility of the entire next one month plan. So, if you were in their position, what will you do?
Shubham Agarwal : So, I think I choose you know, what to work on and what not to work on on my own choice, I mean, which is also known as cherry picking, right?
Kartikayan Ravichandran : Correct correct, that term is called cherry picking. So, what they will try to do is they will try to batch similar products across multiple orders together or they will try to follow a sequence of production which will enable very minimum a set of changes okay. So, what this basically means is some products of one order may reach the finished goods much faster than the other products required for the same order a little later, okay, and this cycle of problems starts that we discussed this problem starts not not just in production, even in procurement from one month plant goes which is then shared with the suppliers. Now, the suppliers will also try to cherry pick and batch to improve their productivity in turn thereby sending certain items earlier which may or may not be required in production at that point of time, while delaying certain other items, which are actually very urgent
Shubham Agarwal : okay. So, obviously, the supplier is also doing the same what what you know, how they are managing the their own systems is what is how almost every system is managing their own work. So, the same problems that
Kartikayan Ravichandran : Coreect. So the problems are very similar, because they also have targets, they also have to increase their efficiencies, everyone is trying to improve their efficiencies, which is actually driving this behavior. Okay.
Shubham Agarwal : Right. So I think we have looked at the problem quite in detail. And we also understood the you know, the core reasons behind it. Why don’t we look at how do we get out of this mess? I’m sure there is a solution out of this as well.
Kartikayan Ravichandran : Yes, the solution is actually pretty simple and logical, if you think about it. So, in this in this environment, all we need to do is to throttle the inflow of orders into the shop floor based on the output of the constraint, okay, so what is this? Yeah, so, what did effectively means is, you introduce only that much WIP into the system, which is actually moving out of the system daily. Okay, okay. So, this is also known as the pull system of manufacturing, because ultimately, the output of the system is determined by the constraints output, okay, output of the constraint. So doesn’t matter how much you keep feeding into the system, if the constraint is producing suppose x quality, the output of the system will be x quantity.
Shubham Agarwal : Right? But then, you saying just by controlling the WIP, that goes into the system daily, the problem would be solved. How is that?
Kartikayan Ravichandran : Yes, you’re right. throttling the inflow of orders, and controlling the WIP level is just the first step. But it is also the most critical step of the solution. Simply by executing this step. The reliability of the orders can actually show an improvement of more than 50 60%.
Shubham Agarwal : Wow, that’s quite a bit, but how do we take it to the 100% mark? How do we achieve complete reliability in the system?
Kartikayan Ravichandran : Yes. So once WIP is controlled the next step would be to manage the priorities of this WIP which are present in the shop floor. So, what do we mean by managing the priorities is each and every workstation should subordinate to just one priority and that is the due date of the order which effectively means, if a workstation has 10 orders in front of it, it will first try to plan and work on such orders for which the due dates are in the past that means, it’s already delayed or the due dates are due today, which is immediate, then it will take up certain orders which for which the due dates maybe a little into the future and finally, if there is still capacity left for that day, then it will start working on orders for which the due dates are furthest into the future. So, what happens is when every workstation starts falling this single priority they start working on the same priority the flow of orders which are required that actually increases tremendously and thereby it directly impacts the reliability of the system and takes it to the very high 90s not just this there is one more benefit out of it. Now since all workstations start working on single priority okay on what is actually required today and whatever is not required today if they don’t have capacity maybe they will not work on that today or try to work on that today okay. So, what it effectively leads to is earlier the bottlenecks which were arising from time to time, today, we may realize that such workstations actually have spare capacity also the real bottleneck of the system will be exposed well the maximum WIP pile-up will be happening today. So the production manager today is in a much much better position to actually identify which workstations we need to start implement projects and increase possibly the increase the output much further
Shubham Agarwal : okay right. But can it be done on a daily basis and out I mean, is this sustainable?
Kartikayan Ravichandran : Yes, the solution is sustainable what it actually requires is it requires a very robust WIP control mechanism in place. So, what do we mean by a WIP control mechanism? One should be first clearly defining and maintaining the amount of WIP that needs to be managed in the shop floor. Okay. Now, once that is very clearly defined that is a trigger mechanism which actually gives the triggers when the WIP goes up and when the WIP comes down okay. Based on this triggers, the planning manager will actually take the decision to further choke the inflow of orders into the system or to release more orders into the system Okay, okay. So, because of which, the system actually ends up achieving very high utilization of capacity, because it is able to prevent both the periods of starvation at the same time it also leads to a situation where the unplanned setups which are happening earlier that stops happening. So so the system utilization of capacity goes up very high. So this is also one reason why the pull system of manufacturing as we call it stands head and shoulders above the push system of manufacturing, which requires a very precise definition of capacity. But remember, we just discussed capacity is not so easy to define. It’s affected by a number of factors. So yeah, so it actually we realized sooner or later that the push system has a lot of problems, it doesn’t work well. So pull system is the best way for this job shops to move ahead.
Shubham Agarwal : Right I understand now that you know why the push system is something which is not very effective, and could create problems on a daily basis while the pull system is much more reliable and sustainable.
But pull systems cannot help give due dates. We need the push system to give due dates to customer, which means capacity scheduling is a must
Kartikayan Ravichandran : We need to differentiate push and pull systems. Whil giving due dates, we need to model the capacity to give a proper due date which is feasible but in execution, we switch to a complete pull model.
Also remember just decoupling these push-pull approaches will not give high reliability. One has to work in productivity improvements as well , particularly in areas where flow is chronically obstructed. In my experience, because of lack of management bandwidth, many work-centers have hidden wastages. As chaos comes down, once can identify the chronic issues of quality or available capacity for special improvement projects. This journey of improvments will take us to a highly reliable plant which stands to deliver as per commitment, while keeping lead times down.Kartik just to end with, Can you tell us what kind of results one can expect from such implementations. Can you share your experience
Kartikayan Ravichandran : Yes, actually, there are many examples by implementing our flow solution or the pull system of manufacturing they not just increased the reliability of the system. But they also ended up increasing their output and reducing their WIP and FG to a significant extent. So the system inventory also went down. Yes, apart from this, the sales obviously, they also increased. But the biggest benefit, but the biggest benefit out of all this, in my opinion, is that the life of the people who work in these companies who worked in these Job Shops, that improved. And you know why? Nice? Because we managed to remove the complexity from their day to day life. What more does one want?
Shubham Agarwal : Yeah, I think we’ve said this enough and more time. You know that peace of mind does not have a cost to it at all. There is no cost of peace of mind. And I think that’s true. Great. I think that is a wonderful discussion. Karthik, thank you so much for your time. And in case for all the listeners, if you have any queries on this topic, any any doubts that you want to ask us, please write to us on our social media handles, or you can also write to us on our email on our website. The link is in the details. Until the next time, this is Shubham signing off. Bye bye.
Comments

Your Comment

Your email address will not be published.