carlsbadgroup

Epicor Enhanced Quality Assurance

Wed, 17 Jun 2020 23:40:01 GMT

Epicor Enhanced Quality Assurance

A Potent Weapon in the War on Defects

Producing and shipping high quality products at a reasonable price is the only way a manufacturer can remain in business. Gone are the days of “good enough” or “we compete on price, not quality”, manufacturers are being held to increasingly higher standards of quality by their competitors. Much of the credit for this goes to W. Edwards Deming, often regarded as the “father of Statistical Quality Control” based mainly on his early work in postwar Japan and who was widely regarded as a hero in Japan for his emphasis on statistics as a tool for improving quality in manufacturing. Statistical Quality Control (SQC) and Statistical Process Control (SPC) are the lifeblood of a manufacturing company that seeks to measure and capture data about their products and to monitor the production process to ensure high quality standards.

Costs of Quality

Deming and others showed that the cost of inferior quality could be measured, giving management an opportunity to evaluate their quality programs in dollar terms. These costs fall into two categories:

Inspection costs, which are easily measured, as the cost of human and machine resources to inspect products during the manufacturing process.
Defect costs, some of which are also easily measured, such as the cost of reworking a defective part or the accumulated cost on a scrapped part. However, there are defect costs that are not easily measured, such as the cost of a late shipment and / or the cost of a lost customer.

Because of the impact of the defect costs there is no “optimal” solution in terms of minimizing costs, the ones that can’t be directly measured will overwhelm the ones that can. However, management can use the information from the measurable costs to evaluate its quality program not with an eye toward cost minimization but output maximization.

An Effective Quality Program

It’s not enough to know that a manufacturing process creates defective parts, or even how many are being created. The real knowledge comes in the determination of why the defects occur, and for this it is necessary to capture process information such as diameter, thickness, tensile strength, etc. This is where the Epicor Enhanced Quality Management (EQA) module comes in. With this module manufacturers are able to create inspection plans for receiving or in-process inspection, and calibration plans for machinery. An inspection plan will:

Define attributes to store measurements and to establish pass / fail criteria. These criteria can be local to an inspection plan or global.
Create specifications and attach documents to the specification for further details as needed. A Specification controls the inspection attributes that are evaluated against the specification and defines the acceptable values for each attribute. You will assign a revision to a specification to controls its effectivity date.
Set equipment calibration criteria.
Use revision control to create multiple revisions of a plan and approve the revisions as appropriate.
Determine sample sizes.
Set inspection intervals using Skip Lot processing.
Store the collected data in database tables available for export to a SPC system and / or a Business Activity Query report or dashboard.

Because all of the above are user-defined, the Epicor Enhanced Quality module can be used by any manufacturer.

Setting Up Inspection Plans

The EQA module uses the Epicor Product Configurator as the base for data collection, although a license for the Configurator module is not required. An Inspection Plan is revision-controlled and may be used with multiple specifications.

The process is as follows:

Define the attributes for the data you wish to collect – what is it you want to measure?
Create the specifications and assign the attributes.
For each attribute, define any parameters and pass / fail values, as appropriate. These can be done in the specification or in the Inspection Plan itself.
Finally, pull all of this together into an Inspection Plan or Calibration Plan. Launch the Product Configurator and bring in the specification and attributes needed in the plan.

Once the Inspection Plan has been created and approved it can be assigned to a part (for receiving inspection or RMA Processing) or to an operation for in-process inspection.

Using an Inspection Plan

In the Inspection Process program in the Epicor Quality Assurance module for a purchase order receipt, the program will recognize when the purchased part has an approved inspection plan and will enable the data collection process. This will start with choosing a sample size; the Product Configurator at this point generates the number of inspection instances to match the sample size and displays the Attributes screens to collect the data.

In a similar fashion, Epicor recognizes when an inspection plan has been assigned to an operation and will provide the same functionality. In both cases the process generates a Nonconformance for any parts that fail inspection. Inspection personnel can use the Nonconformance to create a Corrective Action as needed.

Likewise, the collected data can be exported using a Business Activity Query into a BAQ Report or dashboard to display statistics by part and attribute, such as Mean and Standard Deviation, and to create an X-bar Control Chart to display sample statistics for process control purposes. Crystal Reports is an excellent tool to display the control chart.

Skip Lot Processing

One of the more robust features of the Enhanced Quality Assurance module is the ability to determine the inspection interval for a purchased part to allow receipts of the part to bypass receiving inspection based on past quality results. This is done by defining a Skip Lot Code, which specifies the number of lots to be inspected, the number of lots to be skipped between inspections, and the repeat cycle. Users will define multiple skip lot codes to be used based on inspection results.

For example, a user may decide to inspect two lots of a part and then skip the next three lots if the first two pass inspection, and to repeat this process for a total of three cycles. If the supplier delivers lots that pass inspection for all of the cycles, then Epicor will automatically change the inspection occurrences to another Skip Lot Code that the user has assigned, such as skipping eight lots and inspecting two. If any lots fail inspection then Epicor will likewise assign a different Skip Lot Code for more intensive inspection.

So, What Can I Expect from EQA?

You cannot expect to build and ship high quality parts without knowledge of your manufacturing processes, and EQA provides this knowledge regarding the incidence of defects and where these defects are occurring. As you may have inferred from the above, this module requires a large amount of setup to define the inspection attributes and then use them in specifications and inspection plans. EQA provides a means of gathering production data and storing the data in a format readily available for analysis through SPC software or user-written reports and dashboards, allowing you to see what is going on in your shop floor.

Epicor Product Configurator

Wed, 17 Jun 2020 22:42:03 GMT

Epicor Product Configurator

Business Case

By now most, if not all, of you have heard of the Epicor Product Configurator and some of you may be very familiar with it. For those of you who are not that familiar with the Configurator here is a brief list of things that it can do:

Automate the part configuration, engineering, or quoting process. This is crucial during the sales process to ensure the customers get what they need.
Generate “intelligent” part numbers and descriptions. These can be catalog part numbers or “one off” parts created for a single quote or sales order.
Generate a Bill of Material and Routing to assist in developing product costs.
Prevent invalid configurations and combinations of options.

The Configurator does this by capturing the knowledge of your engineering and design staff and placing it into the Epicor database, and then applying that knowledge during the sales process.

How Does It Work?

The Epicor Product Configurator is a rules-based “expert system” that applies a set of “rules” to user inputs. Basically, the Configurator designer creates input pages which prompt for user inputs and then execute the rule set to accomplish the tasks listed above. The inputs could be very simple, such as “yes / no” questions or a selection from a list, or more complex requiring users to enter measurements or other alpha-numeric data. The output could be a unique product design, a standard configuration, or something in between. The output may also include a “smart string” which can be used as a part number or a means of retrieving a unique configuration of an item.

The “rules” can do many different things, such as:

Specify when to include or exclude an option or subassembly.
Set the value of a field, such as “Quantity Per”, or create a block of text.
Look up values in a table, similar to a spreadsheet, based on several input values.
Retrieve data from the Epicor database tables.
Perform calculations to set values or read a table.

Basically, the Configurator can do as much as you need it to do as a means of generating a sales order or quote, or provide a starting point for further development during the sales process. It all depends on how much “knowledge” you are able to store in the Configurator database tables.

Why Would I Use the Configurator?

If you answer “Yes” to any of the following, you are a good candidate:

Your products are very complex with a large number of options. The Configurator is used by many “Engineer to Order” companies that build unique configurations of machinery. These include a manufacturer of equipment for making and installing doors, a manufacturer of pollution control machinery, and numerous electronic manufacturers that sell high-end servers and peripherals.
Your products have dimensional characteristics, such as a leading manufacturer of sky lights that uses the Configurator to create both catalog part numbers and unique configurations. In addition, many of the window manufacturers in the USA use the Configurator.
You are an “Assemble to Order” manufacturer which creates unique end-item configurations from a base product and a set of options. There are too many combinations to create unique part numbers for each one and you need a way to differentiate the product on a sales order from the last order for the same product.
Your products must meet industry standards for safety, purity, or be in compliance with Federal and State laws.

If it appears that the above list includes almost every manufacturing company in the USA, it’s not a coincidence.

If the Configurator is so Great, Why Doesn’t Everyone Use It?

After telling you all of the great things the Configurator can do, I feel the need to point out that it has some shortcomings as well:

The development time can be lengthy, certainly if your product is overly complex. It took a long time for your engineers and designers to learn everything they need to know about the products and it can take a long time to get that knowledge into Epicor. But in most cases the time saved in the future by using the Configurator more than covers the investment in development time.
Along the same lines, the Configurator may not be able to address all customer situations for very complex products. Or, as is often the case, the additional benefit to be gained from Configurator development is not worth the extra effort. Many Configurator users will create a model that will handle, say, 80 – 90 percent of the development effort and then do the remaining development manually.
A product re-design almost certainly requires re-development of the Configurator model (see number 1, above).

If your products can benefit from the Epicor Configurator then you owe it to yourself to take a look at it. We at Carlsbad Group LLC have implemented the Product Configurator in dozens of companies in many different industries and are ready to provide any assistance we can.

Material Requirements Planning

Wed, 17 Jun 2020 22:03:28 GMT

Material Requirements Planning

Not a day goes by in the life of a manufacturing professional that we don’t hear the term Material Requirements Planning, or its more common abbreviation MRP. We are all familiar with the term, and what it means, but how well do we really understand the inner workings of the tool? Before MRP, most manufacturing companies relied on what was referred to as the “order launch and expedite” system, whereby work orders were released to the floor based on upcoming shipments and expediters worked their magic to get the orders done on time. This led to great inefficiencies in material procurement and resource usage, but there didn’t seem to be any other choice.

The Origins of MRP

No one knows for sure who first put the words together but one of the original proponents of Material Requirements Planning was Dr. Joseph Orlicky, who wrote the first book on MRP while working for the IBM Corporation. Many people had been using the concepts but nobody pulled everything together in an easy to read publication. So, what exactly is MRP?

You’ve heard the jokes: “More Reams of Paper”, referring to the massive computer output of the mainframe devices in use back in the 1060s and 1970s. Or how about “Marketing Runs the Place”, expressing the frustrations of the production managers trying to fulfill the ambitious goals of their sales counterparts. But was MRP that “new” or “radical”? And why the shroud of mystery? After all, it’s an easy to understand concept once you get “under the covers”. Most of the tools were already in place, such as Bills of Material, inventory management, sales orders, etc. The trick, of course, was to pull everything together into a coherent whole.

The main problem was the fact that people were focused on the wrong objectives. Prior to MRP the main focus of inventory management was stock replenishment, which essentially specified a desired inventory level for each product and techniques such as reorder points and Economic Order Quantities (EOQ) to ensure that the inventory was kept at the desired level. This level was dictated by the “service level”, namely the ability to meet demand and avoid a stockout.

These required adherence to unrealistic assumptions: 1) that demand for the products being built and sold was steady and predictable, and 2) the supply of raw materials was likewise known with certainty. These techniques were taught in universities and were considered the “state of the art”. It was not necessary to use a computer; these calculations were simple and could be done on a slide rule or manual calculator. Demand was based on a sales forecast, and the various inventory control techniques insured that there were sufficient materials on hand to build the products. But one piece was missing, namely that of “dependent” demand. Dependent demand states that the demand for a part is based on the demand for another part, as illustrated in the diagram below:

Bills of Material, as depicted above, were already in widespread use as an engineering or design tool, but nobody thought of them as a Planning tool. It’s such a simple concept but it was a breakthrough at the time. The “breakthrough” was not the dependency, which was obvious, but the notion that the demand for the raw materials could be computed based on the production schedule for the finished product. There was no longer a need to maintain a specified stock level of each component part, only to make sure that the materials were available when needed.

MRP Logic

Basically, MRP is designed to answer four questions:

What are we going to build?
What materials do we need to build the items?
What materials do we have?
What do we need and when do we need it?

The term “when needed” brings up a second concept, that of “time phasing” or the process of insuring the materials are available when production begins on the finished goods. Thus, there is an “order by” date for the purchased components based on the “need by date” of the component and its purchase lead time, plus any allowances for receiving inspection, kitting, or material certification processes. These calculations required a computer, as the demand for the finished goods was typically not “known and steady”, which was one of the central tenets of the EOQ model discussed above.

Now that the notion of a “desired” inventory level was dispelled, it became obvious that the desired stock level of each part is zero. There is no need to carry inventory for a Dependent Demand item unless it will be used in production. This led to the Basic MRP Model.

Let’s assume:

Forecasted sales for FG of 100 per period for the next 8 periods

On hand inventory of RM1 of 200 units

Each unit of FG requires one unit of RM1

OK, nothing new here. We all know how this works. Starting with 200 units of RM1 on hand, we would need 100 units per period starting in period 3. So far, we have answered the four questions listed above, but how to satisfy this demand? Let’s add another row to the table. Assume a two-period lead time for part RM1:

With a “Lot For Lot” order policy, we would need to order 100 units of RM1 in each of the first six periods of our planning horizon. Now let’s add some more information, such as open purchase orders for RM1:

This assumes a purchase order quantity of 200 units, or a “Fixed Order Quantity” policy. Does the above look familiar? It should, this is the Time Phased Inquiry in Epicor showing supply and demand. This demonstrates the fundamental objective of MRP, which is to keep supply and demand in balance. Remember that MRP wants to keep inventory at zero, or as close to it as possible, based on supply and demand. To that end, MRP may ask us to do things that may not make sense. For example, look at what happens if the demand in period 3 changes to 150 units:

Please note that MRP will plan based on the greater of sales orders or sales forecast in a given period. In period 3 we see that a surge of orders pushed demand above the 100 units forecasted.

MRP will respond to a net requirement in one of two ways:

Reschedule an open order, or
Create a new planned order.

In this case MRP will generate three suggestions:

Move the purchase order receipt from period 5 to period 4
Change the quantity of the expedited purchase order to 50
Create a purchase order in period 3 for 200

Here is the updated material plan:

Remember that MRP wants to keep inventory at zero, and since the net requirement in period 4 is 50 units then that’s what MRP will tell us to do. This leads to the plethora of suggestions that MRP generates. MRP will always reschedule an open order before suggesting a new one, based on the (usually correct) assumption that our supplier will not accept a new order ahead of an existing order. And, unless we say otherwise, MRP will also tell us to reduce the order quantity to the net requirement of the period. We would need to lock the quantity on the purchase order release to prevent this from happening.

What affects MRP Processing?

Among other things, this would be the Order Policy – how much to order at one time. These come in different forms:

Lot for Lot, also called Discrete, where each planned order is in the exact amount needed.
Fixed, where the order quantity is the same for each planned order.
Min / Max / Multiple, where the order quantity is no less than the stated minimum, no more than the stated maximum, and always a multiple of a given quantity. Note that this order policy, and the previous, can result in multiple orders in the same time period.
Days of Supply, where the order quantity is computed to cover the net requirements for a given period of time.

The choice of order policy for a given item is the result of several factors:

Minimum order quantity (or amount) required by the supplier.
A trade-off between the cost of placing the order and the cost of carrying the inventory. Note that this applies to manufactured components as well if we substitute Setup cost for the cost of placing the purchase order.
Storage and material handling constraints.
Spoilage and / or obsolescence concerns.

Another factor that affects MRP planning and suggestions is the Reorder Point, which is the stock level that triggers MRP to generate a replenishment order. In many ways, the reorder point is a leftover from the EOQ concept introduced above.

It specifies when an order must be placed with a supplier, or when a work order needs to be released to the shop floor. It consists of two parts:

Minimum On Hand, which is the expected demand over the replenishment lead time, and
Safety Stock, the amount held to compensate for variations in the lead time demand or duration.

In most MRP computer systems, the above will generate messages when the projected available stock goes below safety stock and again when it goes below the minimum, and generate planned orders to replenish stock. Basically, the reorder point is another level of demand above any sales orders and / or sales forecasts.

One thing that people often ignore which has an impact on MRP is scrap, both at the material and assembly level. MRP takes planned scrap into consideration, and will increase the manufacturing rate or material usage accordingly. Actual scrap is a different story, as unplanned scrap will result in a shortage of a component part or end item with the consequence of missing a delivery or having to release more material to the shop and rush the items through the production process. MRP must be made aware of the unplanned shortage so it can generate the proper messages to issue more material to a work order or issue more for future orders.

Last, the impact of lead times on MRP cannot be underestimated. MRP to a great extent relies on forecasts, of both demand and supply, and forecasts become less reliable farther into the future. While a two week lead time would not present a problem, what if the lead time is in excess of 26 weeks? This author worked with a manufacturer who was quoted a lead time of 75 weeks for a critical part, making a reliable delivery date for the end item virtually impossible. Worse, MRP uses a planned lead time, the actual lead time can be much different. The actual lead time for a purchased part depends a great deal on the importance of the customer to the supplier, while the actual lead time for a manufactured component depends on the rank of the expediter (do some of your expediters hold the title of Vice President?) It may seem OK if the actual lead time is less than the planned lead time, but that simply means that parts are available that have no immediate use and that other parts may have been sidelined to make way for them. Controlling lead times is no easy task but it is important to the proper workings of MRP.

MRP Output

Revisiting the four questions posed above, MRP provides a list of planned orders and suggestions for changes to open orders; basically to expedite, postpone, increase, or decrease an open order. If there is no demand for an open order, MRP will issue a cancel message. Implicit in these messages is the relative priority of each open order, not only what is needed but when. The main output of MRP is a list of what to make and buy, with the priority specified in the start and due dates for a work order or the order by and due dates of a purchase order. The beauty of MRP lies in its simplicity and its ability to respond instantaneously to changes in supply and demand.

Shop Floor Scheduling

Wed, 17 Jun 2020 21:48:42 GMT

Shop Floor Scheduling

Scheduling of work orders through a factory is an indispensable part of managing a production facility, yet it is typically the last module of any ERP system that is implemented. The reason why this is true may surprise you, although it’s really not at all surprising. Basically, even though everyone understands the concept of scheduling, applying this concept in a production environment is very difficult. There are all kinds of reasons for this, but perhaps the most important is that people don’t understand what a scheduling module is supposed to do. There are the usual problems: suppliers don’t deliver on time, customers change their delivery dates, excess scrap, and so forth; but I have seen over time that many people do not have a good understanding of job scheduling. Of course, this knowledge will not automatically transform your shop floor into a model of efficiency, but it will at least give you the information you need to make informed decisions.

What is Shop Floor Scheduling and what is it supposed to do?

Shop Floor Scheduling is the process of breaking down a work order into its individual operations and assigning each operation a theoretical start and completion time. Typically, this is done by starting with the promised shipment date to the customer and working backwards to determine when the first operation needs to start, commonly known as Backward scheduling. All ERP software will do this, as well as its opposite of Forward scheduling. But there is one major problem, namely that each work order is scheduled as if it is the only one in the shop. Of course, this is never true, which brings us to our next topic of Capacity Management.

What is Capacity Management and why is it important?

The APICS dictionary has an excellent definition of capacity management:

The Function of Establishing, Measuring, and Adjusting Limits or Levels of Capacity. It is the Process of Determining How Much Labor and Machine Resources are Required to Accomplish the Tasks of Production. Open and Planned Orders are Translated into their Hours Demanded at each Work Center.

Basically, this says two things:

how many production hours do we have, and
how many do we need? The first sentence addresses the actual definition of capacity, which is the availability of production resources (usually labor and machine hours). The remaining sentences are the definition of Load, which is the demand on the available capacity (supply). The process of work order scheduling creates the Load on the work centers. So therefore, the objective is to bring the demand for work center capacity in under the availability of that capacity. If it were that easy most ERP consultants would be unemployed. The first thing we need to address is how capacity is determined. There is a Theoretical capacity, which is basically the number of resources times the number of hours the resource is available, and a Calculated capacity that starts with that number and includes allowances for efficiency and utilization. Too many companies use the second measure without regard to the accuracy of these allowances (nobody uses the first measure unless they are overly optimistic). A better measure is the Demonstrated capacity, which is how much you have actually been able to produce (on average) over a period of time. Don’t take your best performance, use an average of several weeks (or months) in order to determine how much you can actually do.

The second step is to now schedule the open orders into that capacity, which we can do in two ways:

assuming “infinite” capacity, or
assuming “finite” capacity. We looked at infinite scheduling above, this is what MRP does when it schedules a planned order, and is also the default technique when scheduling a job / work order. This technique is used to answer the second question posed above in capacity management, namely “how much do we need?” Infinite scheduling will tell us how many labor and machine hours we need to meet the ship schedule, without regard as to whether these hours are actually available. This usually leads to overloaded work centers, but is valuable information in that it points to where we need more capacity. It is now up to management to adjust capacity as needed in order to satisfy this demand, if possible. However, you will not use infinite scheduling to load your shop floor, at least not in the near term.

Finite scheduling takes each operation in a work order and “inserts” the operation into an open block of time on a resource. This assures that the resource will not be overloaded, but it does not assure that the scheduled ship dates will be met. If there is no available time on a resource to perform an operation then Finite scheduling will push this operation out into the future until a block of time is found. If we used Infinite scheduling to determine overall capacity requirements and adjust capacity accordingly then this situation should not occur, but it’s inevitable that it will. Material shortages, machine breakdowns, and worker absences will require changes to the manufacturing plan. Finite scheduling will tell you when you can ship an order based on the available capacity and the other orders in the shop.

So which technique is best? Actually, you need to do both finite and infinite scheduling for the reasons listed above. We will address both of these techniques in a later publication.

OK, so what do we need to make this happen?

Three pieces of information are critical in effective shop floor scheduling:

Accurate due dates on work orders. The due date should be an honest estimate of when the order can ship, not a wish or promise that can’t be kept. Releasing a job to the floor to ship next week, when the normal lead time is 6 weeks, will result in other jobs pushed out of the way and becoming later than they already are with no guarantee that this job will ship on time anyway.
Accurate routings and operation times. This is critical for finite scheduling but also very important for infinite scheduling. Providing adequate capacity is extremely difficult if you don’t know how much you really need.
A valid manufacturing plan. Whether you are using a ship schedule, a master production schedule, or some combination of the two, it is important that the plan includes the correct parts in the correct quantities.

Building larger lots to reduce setup costs and building safety stock to keep machines busy will reduce the capacity available for shipments.

What not to expect

Is your objective to keep your workers and machines busy and maximize efficiency, or satisfy your customers? Hopefully both, but we seldom achieve both objectives. Neither finite nor infinite scheduling will promise level loads or solve capacity problems. If you do not have enough capacity to meet the ship schedule then no job scheduling technique will help. But changing the ship schedule should be a last resort, your goal is to provide the capacity needed to meet the ship schedule.