From a functional perspective, backflushing automates the issuing of material to the manufacturing floor upon the completion of the production process, which the ERP system defines as the point when the manufactured part is transacted into Finished Goods.
From a backflushing evangelist’s perspective, Backflushing is a way to significantly increase manufacturing efficiencies by eliminating manufacturing Work Orders and its associated task of issuing material to the Work Order.
How does Backflushing Work?
The actual backflushing process is really quite simple and contains a few basic steps:
MANUFACTURING LOGISTICS
Employees use the materials it needs to manufacture the quantity of the Part ID for a particular manufacturing schedule
Scrap is tracked
When the Part ID is ready to be transferred to another Inventory location (e.g., Finished Goods) the following information is entered into the ERP system:
Part ID manufactured
Quantity manufactured
Scrap incurred
ERP SYSTEM CALCULATIONS and AUTOMATED ENTRIES
The ERP system will reduce the amount of raw materials and/or any sub-assemblies for the:
Increase the inventory in Finished Goods for the quantity of the part that was manufactured
One can certainly make an argument that any one of the above examples where the prevention of an error, cost avoidance, and/or process improvement could very easily pay for the cost of an alerts module all by itself.
Follow-up Analysis and the Importance of Cycle Counting
By definition, if the BOM is accurate and all scrap is recorded, the only material manufacturing variance should be for any scrap recorded over/under the yield in the BOM. The only way to verify the accuracy of the backflush process is to establish and follow an efficient and regular cycle count program. If the cycle counts result in significant material quantity variances, then either the BOM is not accurate or scrap is not being recorded properly and corrective action must be taken immediately.
Keys to Success
To have a successful backflush process there are a few important things to ensure:
The BOM must be extremely accurate in terms of quantities, expected yield and structure (raw materials vs. sub-assemblies) and contain few if any variable components
Scrap for the production run are recorded on a timely basis
Production cycles are relatively quick
A strong cycle counting program is up and running
While there is no hard and fast rule on what is acceptable or required, most companies categorize their inventory into three classifications; 1) “A” items which typically represent 15-20% of the quantity and 75-80% of the total inventory value, 2) “B” items which usually comprise 15-25% and 10-20% of the quantity and value respectively and 3) “C” items comprising the balance and therefore having about 65-70% of the quantity, but only around 5% of the value.
This inventory value distribution is used primarily for inventory control purposes. Inventory control can include various forms including how tightly a particular inventory item is secured and the accuracy of purchase quantities and cycle counting. In essence, the main purpose is to focus a company’s inventory control on mainly the “A” items, less so on the “B’ items and even less on the “C” items. For example, a company may decide to cycle count their “A” inventory once a week or once a month, the “B” inventory once a quarter or twice a year and the “C” items once a year. Similarly, the “A” inventory items will get the most review and analysis of the purchasing recommendation provided by the MRP.
The benefit of focus is the reason why most companies do categorize their inventory in this manner. Without this tool manufacturing, purchasing and finance personnel will end up spending an inappropriate amount of time reviewing purchases and cycle counts on parts that do not warrant such attention.
The “D” category (which I chose to define it for this blog) mentioned in the title is a bit of a misnomer as it really is not a category, because it represents items of such little value they should be expensed upon receipt; i.e. not classified as inventory for financial reporting purposes. A good example of such “inventory” would be any inexpensive nuts, bolts or washers. Obviously, for “D” parts there would be no need to cycle count nor spend a lot of time analyzing the inventory as the decision was already made by management to expense it.
With the above as context, a reasonable question that finance and manufacturing should ask is should we reclassify some of our existing “C” inventory to a more appropriate expense item. By definition, there will be a one-time charge to the P&L when the existing inventory is expensed, but the charge should be immaterial given the low-value of the parts being reclassified.
Bottom-line: at a minimum, a once a year review of all your inventory items should be done so that all parts are classified appropriately in order maintain the appropriate level of management and control of the inventory. This review should also include analyzing “C” items to determine if certain items should be expensed upon receipt.
One of the more common axioms in the manufacturing world is “Inventory is Evil”, because too much inventory can only lead to problems. The most common problems often sited are the cash consumed purchasing the inventory, increased warehouse costs to store excess inventory, higher obsolescence write-offs when a particular part is classified as obsolete or inactive, and extra manpower to count, review and manage the inventory.
So, how does a company, even a well-managed one, get into inventory trouble in the first place? Two obvious reasons include poor forecasting by sales and unexpected inventory obsolescence due to a technology change. However, to list all the potential reasons for high inventory levels is beyond the scope of this narrative as the intent is to help you to discover ways to reduce your inventory levels without making overhauling your current basic processes; in other words quick fixes that can have dramatic results.
One of operations’ key responsibilities is to ensure the appropriate level of inventory is on-hand to support sales. With this in mind there are four important points to consider:
The main tool that operations use to recommend inventory purchases is Material Requirements Planning (MRP).
The basic assumptions that drive purchase recommendations by the MRP are forecasted total sales, sales orders already in backlog, bill-of-materials (BOM), vendor and in-house lead times, and safety stock levels.
The basic stereotypical trait of most manufacturing people is risk avoidance
The general philosophy of most companies, particularly companies experiencing high growth (e.g. start-ups), is to not lose a sale due to lack of inventory. In addition, it is my experience from working for manufacturing companies in fast paced Silicon Valley, a manufacturing manager or purchasing employee is much more likely to get fired for missing a sale than ending up with excess inventory. Playing it safe thereby becomes a basic survival strategy.
Since the BOM, including yield percentages, always needs be accurate as possible and backlog is a known quantity, operations have only a few remaining ways to insert some conservatism (i.e., playing it safe) into the MRP output. They can increase the demand side of the calculation by nudging the Sales forecast up a bit or increase the supply requirements by ratcheting up either the lead times or safety stock levels of all purchased and manufactured inventory.
The analysis and impact of safety stock is pretty straightforward so I won’t spend much time here, but lead time is a bit more challenging and therefore probably warrants a more frequent review than safety stock levels. In fact, Supply Chain Digest published an article on May 4, 2006 titled ‘The Impact of Lead Time Variability’ highlighting “Preliminary Research out of Georgia Tech finds there’s a lot more variability on inbound deliveries than many companies may realize” and as Georgia Tech’s Dr. Donald Ratliff noted that “not only does lead time variability impact a variety of supply chain cost and performance metrics, the impact of variability is actually greater the more efficient a company’s supply chain is.”
That is a rather thought provoking mouthful. In essence, even though lead time variability impact is more prevalent than believed, the impact is greater on well-run supply chains.
Given that a very common way to be conservative in inventory planning is to increase lead times and the implications on lead time variability per the Georgia Tech study, it warrants a more frequent review than it probably is receiving today. To complicate matters just a bit, there are numerous places where lead times can be used or “hidden”. However, a great place to start is the lead time for the receiving department to receive inventory and then place the item into its appropriate stock location in the warehouse. Depending on your particular situation, this particular lead time might even be considered superfluous, because if there is an urgent need for specific material, manufacturing will not only be aware of its arrival at the dock, they will find a way to expedite the material out of receiving and onto the production floor.
Coming from my finance background, conservatism does have its place. However, the impact of being conservative should at the very least be understood by all so that any conservatism becomes a conscious decision by the executive team. For example, let’s assume that the annual COGS for a company is $10,000,000 and the inventory level is $2,500,000. This equates to an inventory turn of 4.0 or having 91.25 days of inventory on hand. If the company has a 1 day lead time for material receipts, then by eliminating this lead time, will by definition, reduce inventory to 90.25 days in inventory. This reduction will result in an inventory being reduced by 1.1% / $110,000 as the company’s MRP will now calculate the inventory purchase requirements based on a shorter lead time.
Bottom line: both lead times and safety stock are typically set and reviewed infrequently. However, as their impact can be a significant drain on cash and a cause of an increase in inventory obsolescence expenses when parts are classified as inactive, they both should be reviewed and analyzed for appropriateness and accuracy at least once a year to ensure that all inventory conservatism is known and understood by all so that appropriate action can be taken.
The three main benefits of lean manufacturing, if implemented correctly, are 1) a reduction in inventory levels 2) it exposes inefficiencies on the production floor and 3) it reduces waste. One of the key elements of lean manufacturing is the deployment of a Kanban system.
While there different types of Kanban systems, a simplistic definition is it represents a Demand-Pull production approach where customer orders dictate what is manufactured as opposed to the more traditional Demand-Push where the manufacturing organization is tasked with producing specific quantities for specific parts to meet an approved sales forecast for a given period. One of the critical keys to success of a Kanban system is to have an efficient/effective Just-in-Time inventory system so that inventory can be delivered to the factory quickly to satisfy inventory requirements for an order received.
Technically, a Kanban inventory system uses a Kanban (a graphic or visual signal; e.g. color coded cards or lights) that indicate to manufacturing to produce another unit or to replenish inventory on-the-manufacturing-line. Its intent is to minimize inventory levels on the manufacturing floor and to control the quantity of production for a particular product.
A simple example is described below in Scenario 1:
An order is received for 10 units of Part A
Part A’s top level Bill-of-Material is sub-assemblies B plus one Part C
The manufacturing floor is comprised one line with a 3 step process:
Process 1: Build sub-assembly B
Line 2: Build Part A (Part B + Part C)
Test
Scenario 1
The manufacturing process would be:
Issue a Green card to Line 1 to build 10 units
In-line inventory actions:
If all inventory levels were sufficient at the time the green card was received, use in-line inventory to assemble 10 of Part B and 10 of Part A
If after manufacturing the order, additional in-line inventory is required the impacted manufacturing location(s) will post a Red card.
In comparison, if a Work Order system is used (Scenario 2) then the below manufacturing process would be used. The important element in this scenario is the 10 units for Part A were considered part of the 100 unit forecast that manufacturing used in their production planning.
Scenario 2
The two biggest risks that often are associated with Kanban are 1) if a large order quantity is received, the Kanban system may find it difficult to produce the required quantity in time for the requested delivery date to the customer and 2) If the manufacturing production cycles are long the manufacturing floor space required to keep the production flowing might become extensive.
Recently I was walking through a warehouse with the owner of a small distributor, when one of his employees brought him a widget that had been damaged by a forklift. The owner told him to trash the widget, and tell accounting to write it off; the value of this widget, $170. This got me thinking, “are they aware of the cost to replace this widget?”
Later, I asked the warehouse person how much they would need in sales to replace this widget; he said $170. Well, not exactly, let me explain.
In order to replace the cost of the damaged widget, the money must come from the margin of future sales. This company earns a 2% margin for this widget, so they must sell 50 widgets, just to pay for the widget that was damaged! That is $8,500 in additional sales!
$170/.02 = $8,500 in new sales needed to replace damaged widget
$8,500/$170 = 50 widgets
This example doesn’t take into account carrying costs; if carrying costs were 30%, the replacement cost would be $11,050!
This chart illustrates additional sales required to make up for lost/stolen/damaged inventory:
Gross Margin
Item Value
2%
3%
4%
5%
6%
$25
$1,250
$833
625
$500
$417
$50
$2,500
$1,667
$1,250
$1,000
$833
$100
$5,000
$3,333
$2,500
$2,000
$1,667
$200
$1,0000
$6,667
$5,000
$4,000
$3,333
$500
$2,5000
$1,6667
$12,500
$10,000
$8,333
$1,000
$50,000
$33,333
$25,000
$20,0000
$16,667
Again, to breakeven on lost, damaged or stolen inventory, the replacement cost comes from future profits! How much harder must your sales team work to make up for damaged, lost or stolen inventory?
These little inventory costs that occur daily and weekly add up over time, and at the end of the year, these “little” costs can have a substantial effect on the bottom line. Therefore, it is important for employees to understand costs associated with inventory, and the impact it has on the company’s financials.
