Production Planning in Microsoft Dynamics 365 Business Central
Production planning is an important part of inventory planning. In this video, ArcherPoint’s Robert Grainger discusses production planning in Microsoft Dynamics 365 Business Central, including a high-level overview of the master data for production planning, and some advice on the strategy around the planning process—like advising users to hold on releasing production orders early because it can reduce the flexibility of the plan and create chaos on the shop floor.
Production planning and scheduling are the hardest things to get right in manufacturing. But by taking the proper steps and getting clarity on your internal strategies and verifying correct setup in your ERP system, you can more efficiently plan, schedule, and produce in your manufacturing organization.
The transcript of this video recording on production planning follows
This is probably my favorite topic, production planning and scheduling, because as I have said, I’ve always felt that the hardest thing to do in manufacturing is to effectively plan and schedule without a doubt…and in many cases, even harder than building the parts themselves. So, this topic is quite important to me and always has been, and it’s something that I’ve presented at previous discrete manufacturing conferences.
To start, the manufacturing software system you’re using is not important in production planning. All MRPs and MPSs generally work the same way. It’s a planning engine. Business Central is more modern than most, which is very nice, and has some functionality that differs from a lot of other ERPs.
But before we get into the planning worksheet, which is what Business Central calls a Worksheet, I’m going to go first to an item because with planning in general, it really starts with data. It’s the old GIGO philosophy: garbage, in garbage out. If your master data as inaccurate, the output of your plan will be inaccurate. It’s really as simple as that. But it’s often very difficult for companies to maintain a very accurate plan because of the needed data accuracy. They can’t keep up with it because it touches many functional areas.
When you talk about production planning, it affects sourcing, procurement, receiving, and everybody within production—manufacturing engineers, scheduling—many different functional areas play a part, and it’s all based on master data, but everything that’s recorded in WIP. So, whether they’re recording labor in real time, backflushing labor, or backflushing material, all these business decisions play a critical role in how they’re going to develop and maintain a plan. There’s the strategy involved here.
But to start from the beginning, I’ll go to an item, just any one of them here. And on this bike here I selected, we’ll just go down to in Business Central, which is actually quite nice, and what’s very different is that we have two different categories here, one called Replenishment and the other called Planning. They are different because manufacturing in Business Central is an option. It’s not part of the base subscription, it’s an additional subscription, part of the premium package. it’s actually quite nice that they allow users to differentiate these two categories.
First with production planning, it all starts with the item setup, the master data. The master data includes your item master here that we’re looking at, your routing, your standard routing, your standard bill of materials, and then your order policies, which in Business Central reside right here on the item card. It’s not a separate extension. It’s not anywhere else in the system. It’s all right here in your item master, which is unique.
With replenishment, this is really just the source of the item. What’s the item source? It’s right here. It’s just called the Replenishment System in Business Central. Is it manufactured, is it purchased, or is it assembled? The replenishment system is where we have to start. We have to tell the system what the source of the item is the planning engine can then tell the user how to replenish the item. Is it going to be a purchase signal, a manufacturing signal, or an assembly signal?
It all starts right here. There’s a lead time. That is very critical, especially for purchase components. If the item is purchased, we need to tell the system how long it takes to get in house. If there’s a primary vendor or default vendor, we can list that right here, all within replenishment. And then if it is manufactured, is it make to stock or make to order?
Make to stock is interchangeable. We’re technically building that to inventory, not to support any single sales order. Stock is interchangeable. So, if I’m building this to inventory, it could support one sales order tomorrow, and then the planning engine has the ability to repurpose it to support a different sales order the next day based on demand, new demand, changes in demand, and demand date.
With stock, you have the ultimate flexibility. You make to order as we’re building this bike in this example for this specific customer on this specific sales order. There’s that hard link. Typically, you do not stock make to order. Maybe it’s too costly and you just want the sales order demand to drive the production for the finished good. You do not want to have any stock policies on the item. And then same thing on the assembly side.
You can have assemble the order or assemble the stock. Typically, if you’re using manufacturing though in Business Central, you would not use assemblies. Manufacturers like to stick with manufacturing. If you do not have manufacturing in Business Central, then you certainly have the capability to just use assemblies and assembly orders. And again, assemblies are just a collection of items that do not require a routing. There’s no series of steps. You do not account for labor, and there’s no WIP tracking. It’s just quickly building a kit or something like that.
And now we move down to the planning section here on the item card. This is where we have our order policies. In the industry, we call all of this order policies or lot sizing requirements on the manufacturing side. Order policies are more on the purchasing side. This is where we define all that, and in Business Central, there are four different types of order policies. Fixed reorder quantity and maximum order quantity have to do with inventory and managing inventory levels since inventory is really the first consideration for effective planning. The planning engine looks at the top level and explodes downward when it plans, but in fact, we work the opposite. We actually work from the bottom up. We build from the bottom up.
This is where the production planning strategy comes into play. It depends on if we’ve got, let’s say, a 5-level product structure. Usually at the raw material level, there’s commonality. So, if I’m taking a 12-gauge sheet of steel, I’m using that sheet to cut many different parts. That’s commonality. So, I might have a safety stock level on that raw material, and I want the system to tell me when I need to replenish the inventory because the turnover is high. I can use that same strategy on subassemblies, maybe somewhere above the raw material level, maybe somewhere in the middle of the product structure. We don’t typically use these with finished goods because it’s the highest risk and it’s typically the costliest. If you safety stock finish goods, you’re carrying finished goods inventory, which most companies do not want to do.
But the first two order policies usually have to do with inventory, meaning your stock components like your raw materials. How do we want to manage that? Fixed order quantity policy is really just saying every time I need to replenish this item, I’m going to do it in this particular lot size. In this case, it looks like it’s 100 pieces each. Let’s say I have a reorder point of 100. As soon as my projected inventory level dips below 100, the system planning is will fire off a signal to me to replenish my inventory in a quantity of 100. It will not deviate from that because that’s my fixed reorder quantity. So, it’s automatic inventory management based on the strategy you have in place.
Now, with inventoried items, it’s not as easy as it might sound. Reorder point and safety stock down here can be considered the same, but they actually are not. Business Central sees reorder point as looking at your projected inventory and saying, when it dips below this input, I am then going to recommend we replenish that inventory based on this quantity here. what that means is it’s going to send a signal the next time bucket into the future to replenish the inventory. there’s going to be a lag, a delay there. it’s that next time bucket plus lead time.
it will allow you to deplete your inventory lower than the reorder point. Where the strategy comes into play is you need to make sure this reorder point is high enough for you to not fully exhaust your inventory level, then you’re in trouble. That’s where safety stock can come into play, where safety stock, the system sees that as a little different. It’s a hard stop. if I’ve got safety stock of, let’s say, 100 in here, the system looks at it differently. It’s actually looking at your projected available balance, not your projected balance. it’s looking at what obviously your inventory level is plus what might be pegged to other demand orders. what might be reserved already into the future based on your time buckets.
And the system does not want you to dip below your safety stock. If it’s 100, the system wants you to have 100 pieces in inventory at all times to account for that odd situations where we have severe demand spikes because by definition, safety stock is added to demand. if you’re constantly draining your reorder point or your safety stock for that matter, then they’re not working for you. Your strategy is not working for you, and we need to reassess that. Typically, reorder point and/or safety stock, depending on which one you use, you need to do some usage analysis on your inventory.
And what I always recommend doing when I used to do this in the plant was figure out what my average usage is throughout the year, keep a rolling 12 month, then add two standard deviation to that that my safety stock is actually covering me 97% of the time, not 50% of the time. If your safety stock and reorder point is only covering you 50% of the time, it’s not working. It’s as simple as that. you need to do some math, some of analysis behind that to figure out what these inputs should be [inaudible 00:15:13] going to be different for every company and whether or not they actually use them. Just depends on the environment.
With your max order quantity, another inventory policy here which is quite nice. A lot of systems don’t offer this, but Business Central does. This is another way to manage inventory with a min-max. You can have your reorder point of 100 and your max at 200, just in case you want to keep your inventory level between two points, which is quite nice that the system offers this. A lot of the times, you see this in medical device manufacturing because you might have agreements with your customers to keep a min-max inventory level. This is very common in certain industries where we have contracts where we have to keep a min-max. Here’s where you can do that. You can tell the system what those min-max inputs are, and the system will maintain an inventory level between the two.
And then of course, you can have an order policy in addition to that. If it dips below your max, which is equivalent to your overflow level right here, let’s say that is 200. As soon as it gets to, let’s say, 175, the planning engine will fire off a signal to the planner to replenish your inventory by a certain amount. You define that certain amount here, and that’s usually your minimum order quantity, your maximum order quantity, or simply just a multiple. You can use any combination of the three. It just depends on the part.
A very simple example is a box. I cannot buy one screw. I need to buy a box of 100. Therefore, your order multiple in this case might be 100, or your min-max might be 100-100. The system never deviates. If you need 200, it’s going to fire off two signals for two boxes of 100 screws each if you use min-max. Order multiples of 100 would say, okay, I’m going to fire one purchasing signal to buy 500 screws. That would be five boxes. It just depends on the part and the planners in the business to make that decision.
Same thing on the manufacturing side. If we need to build any certain lot size because it’s economical for us to do that, we identify that right here. If you do not want the system to deviate on manufacturing production order, we can do that right here. If we can have 100 and 100 as our min-max, the system’s only going to recommend production orders in quantities of 100. That’s it. Perhaps we don’t tie up machines or work centers for very long, maintain that manufacturing flexibility out there on the floor rather than have one production order of maybe 2,000. It’s going to sit in on a machine for too long. It just depends, again, on the part.
The other two order policies here are, order-for-order and lot-for-lot. With order-for-order, it goes similarly back to the make to order replenishment policy up here. for those that are using manufacturing and using the planning worksheet, order-for-order means there’s that hard link from supply to demand. The system does not have the flexibility at all to move this production order one bit; this production order has a hard link to this sales order item. The planning engine will not move it or repurpose it to support any other demand even if a new demand is entered sooner, needed sooner. It will not move it. It will create a new order. This is that hard link. And out in the field in the industry, we call that Pegging. Supply demand pegging. In Business Central, it’s called Order Tracking. That’s order-for-order. It’s just that hard link.
Lot-for-lot you can use for make-to-order or even buy-to-order parts. What this is doing is telling the system to look a little bit further out into the future and all the demand I have within this period; just create me one supply order. It’s a way to minimize supply count for make-to-order or buy-to-order items. Lot-for-lot, again, is just saying, I don’t necessarily want inventory on whatever this item is, again, a bike in this example, I don’t want inventory on it, but I don’t want to build just one. I want to tell the system to look out maybe two weeks into the future and for all the bike orders I have, just create me one production order to build them all at once.
And we identified here lot-for-lot and identifying our lot accumulation period. This is the time bucket for non-inventory items. The lot accumulation period is telling the system how far out into the future to look in order to group my supply. In the industry, this is called your day supply or supply days. In Business Central, it’s called lot accumulation period. How far out in the future do we want to look when we create supply? It’s a really nice way to maintain a limit on your supply, even for non-inventory items. It’s beautiful.
It’s a very nice feature that Business Central offers that a lot of other systems simply do not because one of the challenges in production that most companies deal with every day. What’s one of the number one things that production managers hate? A ton of job orders out there on the shop floor. It’s more maintenance; it’s more for them to manage. Operators, cell leaders, and supervisors get overwhelmed in an attempt to sequence that work effectively. They just can’t do it. By minimizing the supply count, it really helps the sequencing of the work by all the folks out there on the shop floor.
The time bucket for inventoried items. If we go back to fixed reorder quantity and maximum order quantity, this is our inventory selections here. The time bucket is simply just time bucket. This is your day’s supply input for inventoried items. And here it looks like we got one week for this item. And again, that’s just telling the system how far out into the future to look for inventory items when it creates a new supply signal. Again, just another way to limit supply count on the inventory side.
You have your safety lead time here, which is nice. This applies to both purchased and manufactured items. In the industry, this is called a Buffer, which gets you to start a little bit earlier than you would otherwise, which is really nice. A lot of people refer this to as a Shipping Buffer because they apply it to top-level finished good items, and then it would trickle down to all of the subassemblies and components underneath that. It’s going to get the work started a little bit earlier to protect you against lateness or disruption and WIP like we experience every day in manufacturing, whether it’s a resource being out, a machine being down, late material, or inaccuracy in routings because routings are always estimated. It all depends on who you might have working on the item, an experienced hire or an apprentice.
This is your buffer. For instance, if you have a two-day safety lead time buffer in here, that’s quite nice. It’s telling the system that if it takes me five days to build something and it’s due in 10 days, rather than start it on day five to finish on day 10, the system would recommend starting it on day three to finish on day eight, ahead of the demand date of day 10. It provides two days of wiggle room to account for all those disturbances I mentioned. That’s critical; I think a lot of manufacturers should strongly consider. And there are other options to us to build in WIP Q and other buffers, but those are in WIP or on purchase items. This is the downstream buffer option right here.
This is where it all starts to form an effective plan—getting everything on the item master input correctly. There’s often strategy in here. What items are we talking about? What level in the product structure? It usually requires a lot of guidance to make these inputs correct so that you have the correct output from the planning engine. That’s the item master.
In addition to that, we need accurate routings. These are estimated; with routings, they’re never going to be perfect. There’s just too much variability with resources that we just want to get them as accurate as possible to provide the best plan to us because the lead time for manufactured parts is the routing. If we’re way off here in our routing, then obviously the plan for the part is going to be way off. It’s really as simple as that.
Having strong cycle times because the cycle time is a combination of four inputs: set up, run, wait, and move time. Having strong cycle times in our routings is obviously very critical. And then I’ll just bring up a BOM, a bill of material. The most important thing here is making sure the structure is accurate, meaning that we have the dependency built in for planning that we have all our subassemblies and components identified with the right low-level code that the system can plan top-down, but we release and build from the bottom up. Therefore, we work on purchasing our raw materials first if we need to if it’s not already in inventory. If it is, then we issue that to the lower-level job. We release that order, work on that, complete that, then release the next production level up from there.
We manage WIP based on how we release orders. We don’t want to flood the pipe. If we release everything once way in advance, all we’re doing is creating chaos. We’re going to overwhelm the shop floor, and then the planning and scheduling systems, if we are using scheduling, is going to not work for us. We need to limit the number of released orders we have out there on the shop floor that the system can work for us. A lot of the times and in a lot of companies, I see that they release orders way too early. As soon as the production order is created, they release it. They release everything all at once. And that’s not good, even if they won’t start the work for weeks or maybe months in some cases.
There’s no reason to release that production order. If we keep that in planned status, the system will maintain the flexibility for us. As soon as we release it, we hamstring ourselves. Then you’ve limited its flexibility to make changes on that order, especially in the case if, say, the sales order was canceled. We’ve got a release production order out there. What if it’s got material issued to it or labor on it? Nothing we can do about it. Now it’s a business decision: do we want to finish the order, put it back in inventory and wait for the next one, or just absorb it and cancel it? We really don’t want to do that. It’s something I always try to advise clients on—don’t jump the gun with releasing orders. Leave them in—in the industry, we call it Open. Leave it open. In Business Central, it’s called Planned. Leave it planned. Only release it when you need to. Only release it when you’re ready to start the work. Most people don’t realize that you lose all the flexibility when you start releasing stuff prematurely.
That really covers it for the master data high level here. You’ve got the item master inputs, routing, order policies, and bill of materials, and now let me just take you to the planning engine itself in Business Central. This is what it looks like. In Business Central, it’s called the Planning Worksheet. It starts with generating a plan. I’ve already ran one here to save time, but under Prepare, we just hit Calculate Plan. You get this window that tells the system how we are planning, and how we want it run.
MPS is the master production schedule. That pertains to top-level finished good items. MRP is what planning in general has been called for many years. In the industry, we call it MRP or Planning. MPS is another component of it. When most people refer to MRP, it’s both. They don’t differentiate the two. Why would we want to plan only the finished good, especially if we’ve got many components underneath it? We should plan everything together. MPS and MRP really should be run together in my mind in almost every case, but you do have the option to run just the components or just the top levels.
And then our start dates. Here’s our horizon. We call this the planning horizon. In Business Central, it’s the Starting and End Date. It’s up to you when you want the plan to start. I suppose this would reference how many late orders you actually have first and foremost? Do you have demand that has not been fulfilled that is in the past? Then our start date should probably be in the past. That’s incorporated because, if you start with today’s date, the system will automatically think that late orders have been fulfilled, which might not be the case. And a struggle for many companies is lateness, and it’s a result of poor planning and poor scheduling. This is what the system aids them in correcting. Identifying your planning horizon. I always recommend this should be your longest purchasing plus manufacturing lead time and then some. If you’re not covering that, you’re not doing yourself justice. Longest purchasing plus manufacturing lead time.
In some environments, I say double that because it all depends on if you forecast. Are we forecasting? At what level are we forecasting? Are we forecasting finished goods? Are we for forecasting subassemblies components? With forecasting, that could be a whole other discussion. It’s another strategy on how to forecast. I always recommend forecasting at the lowest level you can go. There’s less risk, it’s less costly, and you probably have better usage analysis on your components than you do with subassemblies. If you forecast the top level, it’s much higher risk. You run the risk of setting an unfinished goods inventory, which is costly. You always want to forecast parts with commonality, similar to what I mentioned with safety stock. That 12-gauge sheet of steel I mentioned is used to make many finished goods. Forecast that. You’ve got better usage analysis on that. There are higher turns on your raw materials. Forecast the raw material. Have that available inventory and then let sales order demand drive production.
That’s your horizon. And then again, if you’re forecasting, which is very nice, Business Central allows you to upload or enter many different forecasts, and you can tell the system which forecast you want to use with this planning run. A lot of systems do not allow for this. And then of course, if you wanted to just plan a certain item or a range of items, you could. You could limit the plan right here by using your filters. Typically, you run a plan for everything. You want to plan the shop or the location.
And again, it depends on the environment, the business … how often are we planning? I always like to plan every day. You’re entering sales orders every day, aren’t you? You’re modifying inventory every day through adjustments or issues or whatever it is. Why not incorporate that every day? Keep your plan current. When your plan falls behind is when you get yourself into trouble. I like planning every day. It doesn’t mean you need to adjust the horizon, the start date, because your late orders might remain in the past, but we should run the engine every day that our signals are up to date. But again, it’s just up to your company. Some plan everyday schedule once a week. It just depends on the parts—how fast they move through it.
Then we hit okay to run it. Here’s the output. Here’s our plan, and a lot of these messages are for us. There are some warnings, meaning exceptions, attention, emergency you would get. I don’t know if I got any emergencies. It doesn’t look like it, but a lot of it just depends on the state of the item. Emergency would be like more on the demand side. If demand fell into the past and is unsupported or unfulfilled, then it’s an emergency because there’s a customer order sitting out there you should ship. Attention would more be on the supply side. If you have fallout or something like that, and you don’t have enough supply to meet the demand, but it’s not past due, it’s into the future; it would require your attention. The exception is if you ran yourself negative or something at inventory is off. Maybe a user did some poor transactions or incorrect transactions.
But what the system’s trying to do is help you. It wants you to look into every single one of these messages and take action, accept it, ignore it in some instances. Maybe you know something the system does not because it doesn’t have the data, but you know, so you can ignore it if you wanted to. But here are all these warning messages. Let’s just go to some with new supplies. Here, somebody entered a sales order last night and I run planning first thing this morning. Planning creates new supply to support that demand. Somebody entered a demand for a touring bicycle and look at that. The planning engine just created a new production order to build it.
The source of your item is right over here. The reference dock is in here, too; you can peg it. We call it Pegging. In Business Central it’s called Order Tracking. You can find out what the relationship is between supply demand. Why am I building this bike? The system will tell you why you’re building the bike because it was ordered on this sales order number. There’s demand justification with everything within planning. At a high level, you can think of planning as material management. Planning manages the material. It is infinite. Planning’s job is to make sure supply is equal or exceeds demand based on policy. That’s planning’s job. It is infinite; therefore, it can go into the past.
Scheduling is quite a bit more complicated because it is finite. Scheduling realizes true material and capacity constraints. It recognizes what you can and can’t do out there on the shop floor. Scheduling can only work into the future, from this moment forward. An example would be if you have something due in five days, but it takes you 10 days to build it, planning would say just start this five days ago. That’s not possible. What scheduling would say is, we can start this today and finish in 10 days. Therefore, you’re five days late. Lateness is a metric of scheduling, not necessarily planning, but it all starts with planning. You have to plan before you schedule. You have to have a plan.
Planning is managing the materials supply meets demand. Scheduling is then sequencing the plan. It sequences the work based on demand date and time. Within the planning worksheet in Business Central, you can actually see these start and end date times. In Business Central, the plan actually does provide you with infinite scheduling dates. These are ideal scheduling dates. Obviously, if we were to run an advanced schedule, it might modify these to make them more realistic because the system might say you don’t have machine capacity on this day, so you have to work on this three days into the future. Therefore, you might be three days late. Planning doesn’t necessarily know that because it’s infinite, but it still provides you with these targets, which is very nice—a huge reference when releasing orders.
But what the system wants the planner or planners to do is review every single one of these messages and take action. You can take action on these all at once; you don’t do it one by one. You can just say process, carry out message. I won’t do that because I don’t want to lose my data here, but you can say carry out for all of those selected here in this column. Business Central doesn’t default the action message to be selected because it actually wants you to look into them individually, which is good.
A lot of folks might disagree with me or have gripes on this because planners in general have enough on their plate. It’s a hard job. It is. Being a production planner is a very hard job. Those people are rare, and they might not have time to review…in this case, I’ve got a small data set. It might be thousands of messages. Nobody has time for that. The system wants you to review them individually, but planners might not have the time or might not want to do that. They just have to trust it. That’s the reason these don’t come defaulted. The user would have to make the selection there to carry out the message.
But you’re going to have tons of messaging here to cancel. It wants you to cancel this one assembly order here, maybe because stock was found. There’s no reason to assemble it anymore. Somebody made an inventory adjustment. Now we have inventory. We’ve got stock. Maybe the sales order was canceled. We don’t need the supply anymore. There could be…now this is considered redundant. We call this redundant manufacturing. We want to limit the number of redundant orders. We don’t need them. The system is guiding us. It’s helping us. It can see and do a lot more math than any one planner or planners could every single day. Even if you get the best planner on their best day, they cannot do what the system is doing.
That’s why I always say look at all the action messages on my very small data set. Most businesses have thousands. Planning and scheduling are the hardest thing to keep up with in manufacturing. We can drill into order tracking, forecasting, and everything—the fact box, everything you see over here on the right-hand side, which provides high level data on the item itself. We’ve got material availability, which is very nice. Again, there’s demand justification for everything in here; you can drill into it and see based on bucket individual date or demand, what the system is doing. Why did it create this for me? Well, here you go. Here’s your justification. We’re going to run ourselves negative if we don’t create a new supply line to support whatever order this is on this date. This is just by event. By demand line.
You can look at it individually by date. We’ve got all the justification tools in here through our material availability. We call it Material Availability; in Business Central, it’s called Item Availability. And then a nice resource that planners always reference is order tracking right here. Let’s see if I can get one here. I landed on an example here ironically. I just went to the order tracking link through the related future at the ribbon there.
Here we can see this planning item was created to support this sales order. It’s linked. There’s justification for everything, even if it’s stock. It doesn’t have to be make to order. Even though it might be make to stock, there’s a sales order for a stock item that we don’t have in inventory. The system’s going to tell you to build it. Technically, you’re supporting inventory because it’s stock, but it’s just going to be a soft link to this sales order. This is the demand that’s triggering the supply. When we’re done with the production order, we simply move it to inventory and then ship it out when the customer needs it based on the demand date on the sales order line. There’s the soft link, a soft peg, even for all stock. It maintains a relationship throughout the entire product structure.
And then we can always drill in underneath the top level to see what supports this finished good that we’re building. But planning is a big topic. A lot of inputs. Obviously, you see a lot of outputs here. It’s all driven by the individual company. What’s the strategy? How do they manage their inventory today? Is it effective, not effective? Is there a better way to do it? What are their order policies? What’s the accuracy of their master data? How do they plan today? How do they schedule today? A lot of different factors to consider with implementing a planning engine. This is not a plug and play. You can’t just turn this on. It typically takes significant business consulting and guidance to utilize planning in manufacturing and then ongoing training to perfect it. It’s a project in itself for sure.
Contact ArcherPoint today for your Microsoft Dynamics 365 Business Central or Dynamics NAV production planning needs.