Warehouse Management

How you pick orders determines your overall warehouse efficiency. It's easy to understand why: warehouses typically source goods in bulk (think pallets) but fulfill orders by disaggregating the bulk packs. As a result there are many more pick transactions than putaway transactions in your warehouse. Therefore an efficient picking process has a big impact on your on-time order fulfillment and overall warehouse budget.

In this post, I will put some of my thoughts about the popular pick strategies in a warehouse such as order picking, cluster picking, pick and pass, etc. I will also discuss which ones are more appropriate for your warehouse (or more importantly which pick strategies are not).

A picking strategy consists of the following three policies:

Pick Method: How do you group the tasks for execution?

• Order Pick: All tasks for an order are grouped and assigned to the operator. Order pick implies one or more pass of the pick zone for each individual order. Order pick is ideal for those orders with pick volume that exceeds the equipment capacity. Order pick implies one or more pass of the pick zone for each inividual order.  Therefore order pick involves higher travel time. However since each order is picked individually, the time to sort items for an order and time to consolidate order with other picks is less. Since order pick ensures that an individual order is picked entirely, it ensures that a given order is fulfilled quickest. It is a preferred pick method for emergency orders or other "hot" orders. 

• Wave: No grouping of tasks is done. The most optimum task among a pool of eligible tasks is dispatched to the operator. Wave pick is ideal for picking those items that are shippable units with their shipping label and hence do not need packing into another LPN. This method is ideally suited for case and pallet picks. Wave pick results in the least travel time. However wave picks may require additional work if consolidation with other order lines and palletization is needed.
  

• Bulk: When a small number of items are ordered by a large number of orders, it's possible to improve the picking efficiency by bulking the multiple pick tasks for an item and locator combination into a single task. The pick execution occurs by performing a single pick load for the bulk task and performing the drop operation for each individual order. Consider bulk picking those items that occur in a significant proportion of orders in a wave.

• Carton Pick: This means that a group of lines in an order are cartonized and a suitable container is identified for them to pick. The carton is also assigned an LPN and the LPN label is assigned to the piking operator at the start of picking. Some warehouses actually pick the lines into a final shipping carton and eliminate an intermediate packing step. However don't go looking for Carton Pick in Oracle WMS. You wont find it. Instead look for "Pick by Label" in Oracle WMS task menu.

• Other Grouping Criteria: While the above grouping criteria are common, Warehouses can adopt any other flexible grouping options e.g. Delivery or Delivery and Zone (Sub-Inventory). Oracle WMS allows the use of pick slip grouping rule to perform this type of groping and the use of "Discrete Pick" to actually execute the picks.
  

Cluster Size: Cluster size indicates how many groupings of tasks are being picked simultaneously.

If cluster size is 1 and grouping criteria is "Order", you are essentially performing picking of one order at a time. However if cluster size is 2 and grouping criteria is "Order", picks are being performed for two orders at a time. All tasks for these 2 orders are grouped and executed as a single group by one or more operators (see pick options). In Oracle this is commonly referred to as Cluster Pick. Some warehouses also refer to it as a batch pick. Because cluster pick allows more than one order to be picked simultaneously in a single pass of the warehouse, it often results in significantly less travel time. Therefore cluster pick is often recommended when a warehouse has to fulfill a large number of smaller orders and each order has a pick volume that is significantly below pick equipment capacity.

Pick Options: A given order may be split across zones for picking. The picking process across zones may occur in one of two ways:

• Parallel Pick in Zones (Zone Pick): Parallel pick involves picking the order in parallel in multiple zones. The picks from multiple zones are consolidated in staging area.

• Sequential Pick across Zones (Pick and Pass): Order and Cluster pick can be performed in a pick and pass mode. This involves picking for the same order sequentially across multiple zones.

Here is a comparison of these two options:

How should you go about selecting an optimum pick strategy?

The answer is that there is no one optimum picking strategy. Depending upon items, type of demand, frequency of items in order, storage and equipment used, different stratgies might be better for each situation. More on this topic in my next post.

Last week I blogged about picking strategies in the warehouse. The choice picking strategy is dependent on many things such as attributes of picking zone, items, orders and picking wave. Depending upon your objectives the choice of picking strategy could be different.

In order to analyze the Pick Stratgies in the warehouse, I have created this Excel Spreadsheet. The excel sheet requires item, warehouse layout, demand and task information. Subsequently it calculates overall load on warehouse and load per resource. Using this information warehouse planner can analyse the interplay of various picking parameters on picking efficiency. The spread sheet helps you evaluate pick strategies in your warehouse and provides a "what-if" type of analysis based on the choice of pick strategy such as:

• What is the total time required for a specific type of strategy in the warehouse

• What is the load profile on different resources in the warehouse

The objective of this tool is not to perform accurate calculations for pick load and execution time. It is a rough-cut tool to analyze picking stratgies and to faciliatate picks strategy selection and load balancing with minimum data input. Data input requirements are minimized by grouping items into item categories, locators into zones and orders into order types. The assumption is that objects in a group are similar and therefore pick information can be entered at the group level. However like most other analysis tools, the output from this tool would depend on the quality of the input. Therefore user must make appropriate trade-off between accuracy of results vs. the time and effort required to gather accurate data.

By no means do I claim this to be a "perfect" or "tested" tool. Please exercise caution while making any conclusions. Let me know if you find it useful or find something that is obviously wrong.

Cartonization is the process that suggests a container for packing items based on packing constraints such as cubic volume of items and container volume. Cartonization logic is most often invoked prior to picking when the orders are pick released. Typically cartonization is used in this fashion to suggest an appropriate sized box or carton for loose item picks in the forward pick area. Oracle WMS can also generate a UCC-128 shipping label for the suggested carton. Subsequent picking in WMS can be done using "Pick by label" i.e. operator scans the shipping label and applies the label to the suggest container. The picks for that carton are now dispatched to the operator on the hand held device. Some use cases of cartonization are as follows:

• Pick eaches into a pregenerated carton LPN: This type of cartonization is frequently used when a warehouse operator picks eaches into an either a tote or a shipping carton. Cartonization should be restricted to warehouse area that stocks loose items. Cartonization should allow item commingling. This cartonization technique is ideal for order pick with "pick and pass" option.

• Cartonization with shippable cases: This type of cartonization is typically used when cases or full pallets with standard pack quantity are picked and a pre-generated shipping label is applied at the time of picking. Cartonization should be restricted to or "Pallet" areas of the warehouse. In addition, cartonization should merely suuggest packing the standard pack quantity of each case or a pallet into the cartonized LPN.  This cartonization technique is not suitable for "Pick and Pass" option. The picked cases should ideally be moved to staging area on a conveyor where they should be consolidated and palletized.

• Pick cases into a pregenerated pallet LPN: This type of cartonization is used when a warehouse operator picks cases into a pallet. Cartonization should be restricted to area of the warehouse. In addition, item commingling can be allowed such that a mixed pallet can be built during picking. This cartonization technique is ideal for case picks using order pick or discrete pick. "Pick and pass" option can also be used with this cartonization technique.

The biggest benefit of cartonization is the ability to perform one step pick and pack of items directly into a shipping container. If the order to be picked spans a number of warehouse zones, you can use either cluster picking or "Pick and pass" option to decrease the travel time as well. The alternative to cartonization would be a 2 step process where you pick items into a tote followed by a downstream packing operation into the final shipping container. Clearly cartonization saves an additional pack operation. However you will also miss-out on one advantage that the 2 step process offers: additional validation during packing operation. If you plan to pick high value items using cartonization you could possibly have an exception based validation process using a weighing scale interface.

Warehouse management system (WMS) is an important element of any supply chain execution backbone. Therefore a successful WMS implementation is crucial for keeping your customers satisfied, keeping the supply chain costs low and complying with industry standards and documentation norms. WMS implementation failures can be catastrophic for the company's balance sheet as well as for the careers of those involved in the project. Some well publicized WMS implementation failure that can be found on the internet are as follows:

1. Adidas: WMS implementation at the new DC is full of glitches. Adidas can not fulfill 80% of the orders and looses market share that persists for a long time

2. Toys R' Us.com: WMS can not fulfill orders in time during Christmas even though sufficient inventoy was available. Eventually decides to outsource efullfillment to amazon.com

3. J. Sainsbury:  $700 million WMS automation project is a failure and most of the supply chain team is shown the door

Clearly a WMS implementation project carries severe consequences in the event of a failure. What are some of the things that one could do to make implementation of WMS successful? In the next post I will try to put together a list of things that are common among successful WMS implementations.

In Part 1 of this post I blogged about WMS implementation projects and the severe consequences of failure. What is the mantra for a sucessful WMS implementation? What makes a WMS implementation project a failure or a success? Here are some thoughts:

Build a Team for Success

A failed WMS project has organizational issues from its inception. The project team either does not have the authority to make decisions about the project or lacks the expertise in WMS and/or warehouse processes. Absence of executive sponsorship also hampers the project.

A successful WMS project gets the team building part right. This is the first critical step for a WMS implementation project. Most successful projects have an executive sponsor,  usually a key executive from the operational side of the business who has a stake in the success of the project. The executive involvement is necessary to get necessary resources, resolve conflicts as well as handle contingencies. Besides its also important to dedicate resources for important roles such as project manager, WMS experts and business champions. While the project manager could be an internal resource, it may be necessary to staff the WMS experts from outside. They could be either consultants or new employees who have successfully "been there and done that" elsewhere. However it's very important to staff this team with internal business champions. These are end users who have the trust and respect of the warehouse employs and is familiar with the warehouse environment. Chosen carefully they can act as powerful change agents. These are the people who will own the system after go-live when the consultants and contractors have departed.

Begin with the end in mind

A failed WMS project has the characteristics of old wine in a new bottle. People responsible for the project succumb to the natural tendency of implementing the new system the old way. The project team either lacks the foresight or has other agenda besides success of the project.

A successful WMS project team knows that gaining operational efficiency is after all the #1 reason for implementing WMS. Their approach is forward looking. The new WMS processes are designed after careful scrutiny of the current processes. The team looks out for inefficiencies in the current process and how WMS can help resolve it. A successful WMS implementation team does not rely on WMS features alone to deliver the benefits. While looking at software features is a good idea, it's also important to pay attention on things such as warehouse layout, warehouse storage policies, work assignments, resources and automation equipments, etc. WMS implementation project as an opportunity to get things right from the very beginning. This opportunity should not be squandered. A successful project also considers the future growth of the enterprise while designing processes today.

Manage Expectations

A failed WMS project starts with unrealistic expectations about the project. Other times these projects start with no clear-cut criteria for project success. Either way it's a recipe for failure.

A successful WMS project starts with manageable set of expectations. Most successful WMS projects start with modest goals. The project team does not oversell the benefits early on as it's so much better to underpromise and over deliver than the other way around. While some features may appear to be cool, it's important to rationalize if they are feasible for your warehouse. Do you have all the data that is needed, is the date accurate, what will be the impact on productivity if additional data input is needed, does the technological infrastructure exist to support the feature, how reliable is it?

Minimize Customization

A failed WMS project attempts to customize the product to suit its current processes. While customization itself is not a bad thing as many WMS projects need some degree of customization. It's the customization that works around the best practices ingrained in WMS or compromises the maintenance or upgrade aspects of the system that is bad.

A successful WMS project has minimal customization. A successful WMS project treats customization as the last resort. When nothing else such as change in processes or work around would be feasible. The customization is also carefully planned. Only the public APIs or Open interface tables are used. Customization is authorized only after a careful cost benefit analysis.

Document Everything

A failed WMS project has poor knowledge management policies. The warehouse policies, procedures and process are not documented while the project is on-going. When an important project member leaves the project, critical information about the project also walks out the door.

A successful WMS project treats documentation of procedures with utmost importance. Documents are prepared for warehouse processes and policies, configuration document, technical architecture, change management and patching policy and user training. These documents are formally assigned to project team members who are responsible for maintaining it.

Formulate a Change Management Policy

A failed WMS project does not have a well laid out change management and patching policy. Configuration changes and patches are often applied without testing. Worse the patching may occur with total disregard to warehouse schedules.

A successful WMS project has a 3 system approach. Changes are rolled from development instance to test instance for QA. Only after changes or patches pass muster are they rolled into production. A successful WMS project also looks at recommended patch list available on Metalink and applies these patches prior to go-live..

Do not underestimate Testing

A failed WMS project underestimates the importance of testing.

A successful WMS project tests the hell out of WMS before they go live. Any configuration changes such as profile options, rules changes are tested before they are rolled over to production environment. Successful WMS project also do a "day in the life" testing. This is a mock run of an actual go live environment. The testers are the end users themselves. This is a good way to test system stability, ability of the system to withstand volume, technical and network infrastructure. It also tells you if the end users are adequately trained in the system or just shooting the breeze.

Plan Adequately for Go-Live

A failed WMS project does not adequately plan for Go-Live. They either choose a wrong time to go-live, fail to anticipate problems and often start with incorrect inventories.

A successful WMS project diligently plans for the D-date by doing the following:

1. Go-Live Date:  It sets a realistic date for WMS go live. This is done well in advance. The go-live is usually scheduled on a weekend or a holiday (if its not a 24X7 operation). If the warehouse observes seasonal variations in demand, the go-live is scheduled during lean times

2. Facilities Planning: Physically mark your warehouse areas prior to go-live. Use barcode tags and mark the aisles prior to go-live.

3. Set help Desk: It's realistic to expect problems in the first few days. A help desk is setup to resolve these issues. The people manning the help desk are experts from the project team. They know if an issue is user/training issue or a configuration issue or a genuine technical issue for which a TAR needs to be opened.

4. Inform trading partners: It's essential to keep vendors, customers, carriers etc informed about your go-live schedules. This is all about managing expectations. That way any delays or changes will not come as a surprise to them. Its also a good idea to close all open transactions prior to go-live.

5. Perform physical inventory: With a new WMS, you want to start with a clean slate. You don't want your warehouse operators to distrust the system from day 1. Therefore go ahead and do a wall-to-wall physical inventory prior to go live.

6. Equipment Planning: Equipments such as handhelds, label printers, desktops are placed where they should be. Employees are trained to handle them, recharge them and troubleshoot basic issues. 

7. Contingency planning: If shit can happen, it will. Question is what will you do if it does? What are you going to do if network infrastructure is down? Handheld devices are not working? Carousel isn't spinning? System performance is abysmal? It's important to have a contingency plan. It could just be as simple as manual picking and shipping. Important thing is to be prepared for such an eventuality.

Continuous Improvement

A failed WMS project ends with go-live. The next project is initiated only when the current system is unstable, unusable or out of support.

A successful WMS project knows the importance of continuous improvement. Most successful WMS projects start with modest goals and continually refine their usage of WMS. They are up to date with patches and new WMS releases. Features are implemented and system is patched round the year.

Did I miss anything? Your thoughts are welcome.

Inventory inaccuracy is a nightmare for any warehouse. If left unchecked, inventory inaccuracy can lead to a negative feedback cycle of declining productivity and increasing inaccuracies. A downward spiral where warehouse productivity declines and feeds even more inaccuracies in the system. Left to itself, inventory inaccuracy erodes profitability and warehouse efficiency in a big way:

a. Poor customer service when a wrong product is shipped to a customer or a wrong delivery date is promised

b. Increase in Backorders because ATP system thinks there is plenty of stock

c. Lost productivity when operators run around looking for missing products

d. High product obsolescence when the missing products are "found" but too late to be of any use

f. High inventory levels because you need the extra safety stock to hide the inaccuracies

g. Inefficient warehouse usage when you need to stop warehouse operations to carry a physical count in order to satisfy auditing requirements

Here are some steps that you can take in Oracle WMS to improve accuracy:

1. Use RF devices to transact on the spot. This is the easiest way to improve inventory accuracy. When transactions are recorded on the spot in real time, there is less chance of error. This means going paperless and using task management in WMS to convey pick instructions to operators.

2. Train Warehouse Personnel to follow documented procedures when exceptions occur in the warehouse. The warehouse workers should be familiar with the procedure when an exception occurs e.g. if a product is not found as suggested by the system or damaged, the operator should know how to log exception and follow the steps. 

3. Find and fix root cause when exceptions occur in the warehouse. Task execution using RF is a great way to record exceptions in real time as they occur. Analyze exception data in warehouse control board to see where and why exceptions are occurring? Are more exceptions being recorded for certain items or certain employees? Why? If a shipment of wrong product was detected, where did that pick come from and was the inventory corrected for the original item? Was wrong putaway the cause for a pick exception?

4. Storage Policy of items in your warehouse also impacts accuracy. To avoid picking the wrong items, make sure that items similar in appearance are stored apart from each other. Commingling items in the same locator is also a recipe for shipping inaccuracy. You also need to make sure that locators properly marked and physically distinguishable. When slotting items in the locator make sure that the locator corresponds to the item velocity and has enough space to store the maximum quantity of item specified. If a locator stock is overflowing into warehouse aisle, its usually not a good sign. If you have negative inventory allowed flag enabled in your warehouse, you need to question really hard as to why is it needed?

5. A counting policy is a must for every warehouse. Cycle counting a great way to gradually improve inventory accuracy. While eliminating yearly physical count is a noble goal, it can only be achieved when the warehouse has reached a certain threshold of accuracy level.

6. Bar codes or RFID are great auto-ID technologies to improve accuracy. Barcodes have an error rate that is significantly lower than human data entry. Additionally DFI feature in WMS can further improve accurate data entry.

7. Checks in warehouse processes to ensure accuracy. Example of such checks could be an additional packing step to scan items prior to putaway or shipment, a weighing scale linked to a divertor to weigh and compare standard and actual weight of LPNs bound for storage or staging, etc.

8. Check Digits is another way to improve data entry accuracy. When you dispatch an operator to a suggested locator how do you make sure that the picks are coming from the suggested locator and not from any warehouse locator? Locator check digit is a great way of ensuring that operators perform picks and putaway at the same physical locator as the data entered in WMS.

In my previous posting on inventory accuracy in the warehouse, I referred to locator check digit feature in R12. In this blog entry I want to provide more details about this small but powerful feature to improve inventory accuracy in the warehouse. The biggest benefit of locator check digit is that it ensures that warehouse operator physically travels to the locator and performs transaction. Consider this example:

Sounds familiar? How does check digits help? The way check digits work is that each locator is associated with a check digit and transactions can be confirmed only using check digit. This check digit is printed on the locator label physically but it is not suggested to the operator on the mobile device. Meaning that operator can use the locator suggested on the mobile device to travel to the locator but must go to the locator to record its check digit. Besides check digits also helps in faster data entry especially if you have to enter long locator values manually.

Here is a small utility that I have developed in Excel that could be useful for initial locator or warehouse bin definition in your warehouse. To summarize the Excel file does the following:

1. Auto generates locators using range for 3 segments such as Row, Rack and Bin or X, Y and Z coordinates.

2. Generate check digits for each locator using algorithms such as Modulo-10, Alphanumeric, or Unique 4 digit check digit

3. Create locator labels with check digit. You need to have an XML enabled printer such as Zebra 110XiIII, Intermec PM4i, Sato CL412e, Toshiba B-SX6T, Cognitive Advantage LX, Datamax I-4308 or Printronix T5206. Alternatively you need an XML enabled print server such as Loftware, Niceware, Optio or Unibar.

The excel sheet can run a macro to generate the label XML. If you wish to randomize the check digits, there is a macro for it as well.

The locator generation parameters includes a range for each segment. Each segment in turn is split into two sections and you could provide a range for each section using the "From" and "To" parameter. You could for example create locators from EA1-A1-A11 to EA2-D3-Z14 by entering the "From" and "To" parameters as follows:

Some additional thoughts on when to perform cross docking: Depending on the length of the order lifecycle of your business, choosing when to cross dock is an important factor to consider when adopting cross docking. We all know of course that cross docking reduces working capital, reduces labor for putting away inventory etc.

Future Supplies that demands are tagged against need to be monitored for changes to shipment schedules, damages on the way and other reasons, it is a good idea to perform one last round of checking that the crossdocked ASN, PO is indeed on-time just before waving the orders down to the execution system. In a networked fulfillment environment, this may improve fill rates as substitute material or cross docking opportunities may exist in other DC's if there is a problem in supplies arriving.

I am back from Las Vegas after attending Collaborate'07. The conference lived up to expectations with record attendees (6000+) and exhibitors (50+?). The conference was well organized with excellent sessions and well planned demo grounds. From a Warehousing and Distribution perspective there were these sessions:

• WMS SIG: Despite the schedule on a Sunday morning at 9:15 AM, it was very well attended. In the first part of the session, I covered a summary of challenges that are faced by Warehouses today and how Oracle WMS features address those challenges. The second part of the presentation was done by Walt Zipperman from DAZ Systems. Walt did a great presentation on an actual WMS implementation which his team managed to successfully complete in just under 3 months! Isn't that incredible?

• Logistics Strategy: Jon Chorley, Vice President of Supply Chain product strategy delivered this session about the current capabilities and future direction of overall Oracle logistics applications.

• WMS and OTM: Creating an effective logistics foundation: The title of this session was somewhat misleading. The session was mostly about collaboration in a supply chain execution environment. In a WMS world, this means ensuring that compliance mandates are supported in a flexible manner.

• Shipping Directly from Manufacturing with WMS Cross Docking: WMS direct ship feature can be very effective for "Just-in-Time" shipment of assemble to order (ATO) finished goods. This process can be made "contactless" i.e. the finished goods coming off the assembly line can be directly packed and staged to a shipping area. Subsequently the shipping transaction can be completed by a single scan of the LPN bar code.

• Continuous Moves in OTM: Integrating Inbound and Outbound Transportation: This session was about Oracle Transportation Management (a.k.a OTM) capability to synchronize inbound and outbound shipments and thereby increase carrier utilization and lower rates. 

I am not sure if the presentation material is available on-line for those who registered. I will provide a link if I find it.

In one of my previous posts, I blogged about facilitating "Just in Time" crossdocking from your assembly line directly to a shipping dock. The objective is to finish assembly completion into an LPN, move it directly to a staging area and then ship it. This process has obvious benefits such as:

• Reduction in inventory, storage and thereby carrying costs

• Minimal material handling and therefore reduction in labor and equipment costs

• Reduction in lead time and therefore faster shipment

In order to make this process work, you need to tie the WIP job to a sales order using either the new planned crossdocking functionality in R12 or using ATO feature. Either way your WIP job will be reserved to the sales order. To further optimize the process, you need to minimize "contacts" for the shipping container and you might want to consider the following ideas:

• Shipping Label at Manufacturing: To minimize material handling, you would want to create a shipping label directly at WIP completion for the license plates. The labels can be customer specific that include the customer name, address and purchase order number. By labeling the license plates at WIP completion, the shipping process can be simplified because no additional labeling steps may be required as part of the outbound flows. However the customer data is not available on any labels at WIP completion. You will need to use the "Custom Label Field" feature in R12 to fetch the shipping information e.g. a custom label field can be defined for  order number as follows. In this example the return value of the following SQL is mapped to custom label field defined for order number:

Voice picking in warehouses work in a similar fashion as RF devices. Instead of picking tasks displayed on an RF screen, warehouse operators listen to task information on their headsets through a voice systems connected to WMS through a Wi-Fi network. Voice picking also allows task confirmation through spoken commands. Voice picking has numerous advantages in a warehouse such as:

• Voice Picking makes the data entry operation hands free. You do not need to hold an RF device or scan a barcode to confirm a pick task thereby leaving both hands free for physical movement of goods. This significantly boosts operator productivity.

• Picking by voice improves accuracy and as I indicated in a previous post, importance of accuracy can not be underemined

• Voice Picking  is particularly suitable for environments where punching data on RF devices is not feasible such as freezer section for storing perishables. In such an environment you can imagine the plight of the warehouse operators with gloves fumbling and punching data on keyboard

Does it make sense to use voice picking in the entire warehouse? Possibly not. Voice picking may not make sense for warehouse area with low volume or pallet picks which can be done equally effectively using traditional RF devices.Voice picking is also most effective for repetitive tasks. If the warehouse operators perform a large number of different transactions, voice picking may not be very effective. A careful cost-benefit analysis is needed to determine what areas of the warehouse can benefit from voice picking.

If you wish to enable Voice-picking, Oracle WMS has capabilities to interface with any external voice picking system such as VoCollect. It also gives you the flexibility to have voice picking enabled for certain areas or operations of the warehouse e.g. use voice picking for high volume unit picks and RF pick for pallet picks, putaway, receipt and inspection transactions.

The approach needed to interface with a voice picking system is similar to the interface with any other material handling equipment. The interface with voice picking can be initiated at the time of pick release. Its possible to define a business event at pick release for certain areas of the warehouse to compile the task information and interface it to a voice picking systems. Subsequently voice picking system dispatches the tasks and sends a confirmation back to WMS.

Stocking policy for an item dictates the location and desired inventory levels in the pick face. If you think about it, where you store an item in the warehouse can actually make a profound impact on warehouse efficiency. Just imagine, if a frequently picked item is stored in a location that is further along from the shipping area, how much time would be spent in a non-valude added activity such as extra travel time. How about placing a frequently ordered heavy item in  a lower rack where the operator needs to bend down and physically pick?

Most warehouse pick faces have so called "golden" zones. As you can guess, these are the "sweet spots" for picking. Typically these bins are closest to the end of picking line and usually at a level that is most convenient for picking. Alas such bins in the warehouse are few! They have to be; otherwise the pick face itself would become large and unmanageable. How then does one choose locators for an item with a view to the maximum impact in boosting productivity? Naturally you would want to slot your most active items to be placed in most convenient locations.

A distribution center typically consolidates demand across customers and then fulfills this demand using bulk sourcing. Bulk sourcing implies that items will be received in Pallets and master cases and later shipped to customers as eaches or inner packs. To handle this need, you need a pick face optimized for picking eaches (or another lower UOM) and a reserve area in the warehouse that is optimized for bulk storage. You also need to setup replenishments from reserve area to your pick face. The inventory levels to maintain in the pick face are also equally important. The minimum and maximum quantity to store in the pick face should be optimized to avoid stock outs as well prevent frequent replenishments from the reserve area to the pick face.

In order to determine an appropriate stocking policy for your pick face, you need to analyze the demand pattern of the item from the pick face and determine the location based on other constraints such as locator capacity. The parameters to be looked into are specifically the demand pattern and constraints. The demand pattern could be based on historically data or based on forecasts.

Demand Pattern

• Horizon: This is the time period in future for which you want to analyze the demand and determine the stocking policy. The demand for items is dependent on a number of factors such as seasonality, external events, promotions and stages in the product life cycle. To determine stocking policy you need to have a clear idea for item demand during the period e.g. if you want a stocking policy for holiday season, you need to know the time horizon during which this policy will be in effect and demand pattern expected for an item within this horizon.

• Pick Frequency: If the pick frequency is higher you would want to place the item in a convenient location to pick regardless of the quantity ordered for each pick. What this means is that you would place an item with 10 picks of 1EA at a favorable location as compared to say another item with 2 picks of 5 EA even though the net demand for each individual item is the same.

• Pick Demand: The order quantity for each item determines the stocking levels for that item in the pick face. If the demand is high, you would want to keep a high "Maximum" quantity so that frequent replenishments are avoided.

• Variability of frequency and demand: Does the item exhibit an irregular demand pattern during the time horizon? Is it likely to have a much higher pick frequency on a given day and no activity on some other day? This factor would determine if you want to dedicate a locator for an item in the pick face.

Constraints

• Item Dimensions: Item dimensions are an important factor in identifying pick face locations where the maximum replenishment quantity for an item can fit. You would also want to maximize the cubic volume available to you in the pick face.

• Item Attributes: These are factors such as crushability, weight, etc. If an item is crushable you would want this item to be picked up last. Similarly if an item is heavy you would want to pick this item towards the end of picking. Some warehouses also store items that are similar in appearance further apart to minimize the possibility of pick inaccuracies. For the same reason, it's never a good idea to commingle items in the same bin.
  

• Standard Packs: Most warehouses typically replenish in integer quantities of UOM (e.g. cases, pallets) to minimize material handling. Therefore standard packs determine the replenishment lot size.

• Storage Attributes: These are attributes that dictate locator capacity i.e. if the locator can hold the weight, has enough cubic volume to store the item and has dimensions to fit the item.

In the Part-2 of this post, I will discuss process that can be used to formulate the stocking policy in the warehouse. All that is fine and good but how can Oracle WMS help? That is a good question and will be addressed in Part -3 of this post. So stay tuned!

The previous post discussed parameters that influence stocking policy in warehouse pick face. The stocking policy should answer questions such as:

• What items should be stocked in the pick face?

• How much items should be stocked in the pick face?

• Where should a specific item be stocked in the pick face?

• How do you replenish items to pick face?

To answer these questions we need to simplify our approach. One way to do that is to use profiling i.e. grouping items and locators using Pareto Principle. The typical ABC analysis is an example of Pareto Principle. The idea is that certain items or locators can be grouped together in order to simplify analysis.

Locator Profiling

Profiling locators in pick face implies grouping locators into categories based on pick convenience. Pick convenience is a measure of the distance a pick operator travels to perform  a pick and the pick complexity required to perform a pick. Based on these factors, locators can be grouped into categories. A sample grouping could look like:

If you prefer a more graphical look, here is how the locator can be categorized:

Item Profiling

In order to perform this grouping, we need to analyze item demand patterns expected from the pick face. If the pick face in your warehouse is designed to store only Eaches, you absolutely need to remove full case or pallet picks from this analysis. We are also not concerned with the pick quantity either as we are assuming that the amount of time to pick is not not that dependent on pick quantity. Once the pick frequency is available, its possible to rank the items in descending order of pick frequency. Ideally you will see a small percentage of items constitute a large proportion of pick task.

Update: The following graphs shows this type of "Long Tail" distribution of pick activity. The picks for each item on the X-Axis are plotted against absolute number of picks on Y-Axis.

In this example, 80% of picks are coming from 20% or less items.

If the demand pattern in your warehouse is somewhat uniform across items, you will need a large proportion of items to cover a smaller percent of pick tasks.

Fat Tail Pick Distribution

In this example it takes 40% of items to cover 60% of tasks.

If you have a "Fat Tail" pick distribution, having dedicated locations in the pick face may not be that efficient.

Slotting

Next step is to use this information to slot items based on locator information coming from locator profiles. So for example, if there are 5 "gold" locators, you need to identify the top 5 items that constitute the bulk of pick activity, then identify the remaining items to slot the "silver" locators and so on.

By profiling locators and items, we have significantly simplified the problem. Instead of slotting hundreds or thousands of items into a few locators, we now have to slot a few item categories into a very limited number of locator categories. An obvious solution would appear something like this:

If a given slot could not be assigned to an item due to capacity or other constraints, that slot can be assigned to another item with a lower pick frequency. Not all items may get slotted based on number of locators available in pick face.

Fixed or Floating locators

What if the demand for an item or a set of items is highly volatile? Does it makes sense to assign a "gold" locator to an item on a lean day? Possibly not. Sure you can reslot an item during lean times. However reslotting is costly and could get complex especially when there are not enough empty locators.  A possible solution may be floating locators. In such a situation, the locator in the pick face is dynamically determined based on demand for that item on a particular day or even within a pick wave. Sometimes when demand is very volatile and non-uniform across items, its also possible to have a floating locator for an item with a "pick to zero" policy i.e. determine replenishment needs across all orders within a wave, replenish the total demand to pick face and draw down the locator to empty when the wave is completely picked.

Replenishment Policy

Since items that are stocked in pick face are usually fast movers, its not a good idea to have a stock out in the pick face. Therefore the locator should carry a minimum level of stock during the time it takes to replenish. Thus if it takes 1 day to replenish pick face, the average demand for a day should be the minimum quantity of stock at the pick face. The maximum quantity is usually determined by the locator capacity. To fully utilize pick face capacity, you may want to maximize the cubic volume of the pick face.

To reduce material handling during replenishments, its a good idea to specify replenishment lot size as the UOM that is stored in reserve area e.g. if pick face is replenished from reserve area storing master cases, specify replenishment lot size as the quantity of Eaches in a master case. This would ensure replenishments are done in integer quantity of master cases.

Clearly there are number of other factors that go into making an optimum slotting decision. Factors such as order correlation (slot items usually ordered together in the same aisle), crushability (slot crushable items at the end of pick travel path), item likeness attributes (slot like items apart from each other), etc. are important. However this simple approach should be a good starting point.

In the next post, I will cover how Oracle WMS can help in formulating and implementing a stocking policy for your pick face.

This is part 3 of the blog entry on Stocking Policy for the Warehouse Pick Face. Part-1 was about parameters that influence stocking policy in your pick face and in Part-2, a simple approach to formulating stocking policy for the pick face was discussed.

How does it all fit in if you are using Oracle WMS to manage your distribution centers?

Pick Face Definition

The physical space that constitutes pick face in your warehouse needs to be a separate sub-inventory. The individual storage bins in the pick face can be defined as locators in this sub-inventory. As discussed in my earlier post, you may choose to define locators using a Row, Rack and Bin criteria and assign a check digit for improving inventory accuracy. An excel spread sheet that makes it easier can be downloaded here. If the pick area stores a particular pack configuration e.g. cases, cartons, units, each, etc. you may also configure the pick unit of measure for the sub-inventory corresponding to this pack configuration. The pick unit of measure will let you configure pick rules that minimize material handling in your warehouse by fulfilling orders in integer multiples of pack configuration.

Dedicated and Floating Locators

Next step is to assign items to your pick face sub-inventory. Here you have two options. You can either dedicate specific locators to items or have a "floating" locator for the item in the sub-inventory i.e. assign items to sub-inventory without assigning a specific locator. As we discussed in the earlier post, you will likely dedicate your best locators to items with high pick frequency and uniform demand pattern.

Replenishment Setup

A pick face needs to be replenished from a reserve or a bulk storage area. The parameters for replenishment such as minimum quantity, maximum quantity and replenishment lot size can also be setup for items in the pick face. The replenishment lot size should be the UOM multiple of the higher level pack configuration e.g. if you replenish 2 full cases every time you replenish the pick face and each full case contains 24 EA, your replenishment lot size is 48 EA.

Replenishment Planning

The first  two steps take care of the setup needed for the pick face. In order to keep the items stocked in the pick face, it needs to be replenished at regular intervals. To do so, a min-max replenishment planning can be scheduled to run at regular interval say every 30 minutes. Replenishment planning will compute the replenishment quantity using the following equation:

Once the replenishment quantity is computed, the replenishment move orders are generated using replenishment lot size. The replenishment move orders that are generated at this point, do not have a source and destination locators identified. Pick and putaway rules let you do that.

Pick and Putaway Rules

Pick rule identifies the source locator for replenishment while putaway rules can be configured to identify the destination locator for replenishment. Putaway rule can be very easily used to select a destination locator that is dedicated to a particular item.

What about floating locators? This too can be achieved using putaway rules. If the item already exists in the pick face, a simple putaway rule can be configured to aggregate material at the same location. You can also use locator flex fields and SQL expression in the putaway rule to setup almost any business policy. The Rules Engine Example  (Note: 232247.1) document available on Metalink has sample setup and putaway rules example where locators can be assigned to items based on their ABC classification.

Replenishment Tasks

The pick and putaway rules are invoked during move order allocation. The allocation process also generates replenishment tasks that can be assigned to a warehouse resource for execution. The allocation process itself can either be carried out manually for a move order or allocated in batch using "Move Order Pick Slip report". The replenishment tasks can be dispatched to an eligible warehouse resource as any other pick tasks.

This concludes the 3-part series on stocking policy for your pick face. I will be very interested in knowing your thoughts, any specific challenges that you face or your own perspective on this topic.

This is of interest to all of you who worry about finding storage space for your items. Especially those with large number of items and multiple storage options :)

My previous blog posts (here, here and here) on stocking policy for your pick area can guide you in finding the optimal location for items. However if you have certain types of storage options in the warehouse with fixed cubic volume, what you need is the most cost effective way to slot items into various types of storage options commonly available in the warehouse. Art Avery has tried to do just that in this article. Art was also kind enough to send me the spreadsheet with the number crunching.

Sure, parameters in your warehouse may be different. The important thing is to understand the simple and yet elegant concept behind it and use it in your situation.

The question was asked in the recent past. How are the needs of a WMS in BRIC (Brazil, Russia, India, China) countries different from a developed country such as the US? To answer that question, lets examine why enterprises decide to implement WMS in the first place. What needs do they expect a WMS to address? There are many, but in the end they can all be categorized in one of the three generic buckets:

1. Improve Operating Margins: Improvement of operating margins by reducing logistics cost is one of the biggest motivator for using WMS. To achieve cost reduction a warehouse must efficiently use the warehouse resources. In other words get more by using less of:

• Labor and Equipment

• Warehouse Space

• Inventory on hand

2. Improve Customer Service: Excellent customer service is essential to winning and retaining your customer base. WMS can make a big difference by facilitating:

• Accurate Order Promise Date

• On Time Shipment of Orders

• Minimize shipping inaccuracies

3. Improve Adaptability: A WMS can help you adapt your warehousing infrastructure to changing business environment. At the very least a WMS should enable:

• Flexibility of Operations

• Adherence to compliance and standards

• Supply Chain Collaboration

Clearly, even though this may sound somewhat counter-intuitive, a BRIC enterprise needs WMS for all the same reasons as a company in a developed economy. What is different about BRIC and other developing economies is the operating environment characterized by:

• Fragmented Supply Chain: Relatively poor road and port connectivity implies that there are high lead times and inventories. Complex multi-level supply chains involving wholesalers, stockists, etc. are common.  The intermediaries in the supply chain are usually small with little investment in sophisticated supply chain execution systems.
  

• Low Transaction Volume: Fragmented retailing and complex supply chains imply large number of smaller warehouses with many of them in unorganized sector

• Low labor costs: Therefore less emphasis on warehouse automation

As a result current expectation from WMS are different. However a consumer boom fueled by rising income levels and easy credit is rapidly changing the operating environment in BRIC economies. Competition is becoming more intense and consumers are becoming more discerning. As a result the following trends are fast emerging:

• Shift in the bargaining power from manufacturers to retailers to consumers: This trend would eventually result in consumers getting the upper hand with respect to choices, price and availability of merchandise. A world class warehouse and distribution system would be a key competitive differentiator.

• Consolidation of supply chain: Changes in supply chain will be rapid. The consolidation may involve relying on large logistics service providers (3PLs) for warehousing and distribution needs or retailer owned DCs built to distribute over a large geographical area and to gain economies of scale.

• Supply Chain Collaboration: Collaboration between retailers and manufacturers will gain momentum with more electronic order exchange, ASN, item code standards, electronic manifesting, etc. in warehousing and distribution space.

What is needed is a WMS that can rapidly adapt to these changing business conditions including an exponential increase in business volume. Due to low labor costs, warehouse automation is not critical. However the ability of WMS to improve warehouse efficiency by facilitating lower inventory, faster lead times and better utilization of warehouse floor space gains even more importance. In a developing economy, the volumes often do not justify big IT investments in WMS. Therefore a low overall cost of deployment for WMS is critical. What is needed is a WMS with least amount of additional technology infrastructure and a WMS that is tightly integrated with the corporate information backbone. WMS can also help the enterprises provide superior customer service through perfect order fulfillment. A WMS implemented in a silo-ed environment may not be the most efficient option from this perspective. Due to rapid growth and changes in the business environment, WMS must have technological flexibility to adapt to:

• Higher Volume: WMS must be able to scale very well to higher volumes and growth without compromising existing infrastructure.

• Changing standards: With more emphasis on supply chain collaboration, rapid changes to standards is to be expected. Can WMS adapt to these changes? Does it have open and transparent approach to standards?

• Technology Platform Changes: WMS should be capable of supporting paper based system today however in future it should have the capability to support more advanced Auto-ID technologies such as Barcode and RFID along with with transaction support on wireless hand held devices. As enterprises grow and mature, it should provide flexibility to support automation using voice or other robotic equipments. In effect, enterprises should have the flexibility to move up the sophistication ladder without phasing out their existing WMS infrastructure.

Modern Material Handling's latest issue has an excellent article on WMS Selection (Hat Tip to Bill Reilly for forwarding the link). The seven questions covered by Bob Trebilcock in this article is a good starting point and should provide a lot of clarity to anyone in the market for WMS. From an Oracle WMS perspective the very first question is the most relevant:

Can you get WMS functionality from your ERP system? If you have Oracle Applications installed, can you prove that Oracle WMS won't meet your needs?

What this means is that if you are an Oracle Application customer, you need to have a pretty good reason why Oracle WMS should not be chosen for your warehouse needs. You need to consider the savings from an integrated WMS like Oracle vs the incremental benefits from a 3rd party WMS. How can that be done? WMS Solution Factory has all the tools!

Take a look at the Integrated WMS value calculator at the WMS solution Factory. Essentially you need to compare the savings from an integrated business like Oracle WMS vs the incremental operational benefits from a bolt-on WMS (if at all they do exist). You can use the WMS value calculator at the WMS solution factory to see if the additional benefits are justified.

Besides the 7 questions, the following points are also important:

What are you doing now? What are you doing about your current WMS needs? This is important if you want to reuse your existing IT assets and want to limit the learning curve. If your current WMS needs are causing operational issues, the time available to select a WMS is fairly limited.

Where do you want to go? You need to have a clear idea about your motivation for a new WMS. An important aspect of WMS selection is to identify the long term goals and objective behind new WMS implementation. A WMS implementation is capital intensive and once deployed not easy to switch. Questions to be asked: How is the future growth impact warehouse operations? Is the WMS vendor financially viable? Is the WMS vendor  capable to support your WMS needs for the long term?

How do you want to get there? If the number of WMS vendors seem bewildering, here is a suggestion: While a WMS is important, operational success in your warehouse and WMS deployment has less to do with the software and more to do with how its managed. A clear plan for success is important. I have blogged about it in the past at here and here. Do it and you will be fine.

This topic was covered in detail at the WMS SIG at Collaborate 08 in Denver. Replenishment plays a key role in the warehouse as it ensures that there is a enough stock in the pick area to cover the expected demand. The replenishment in this context means an intra-warehouse replenishment from a reserver or bulk area that is optimized for storage (Think pallet storage, high bay warehouse etc.) to a small and compact pick area optimized for unit picks (think flow racks, shelving, pick to light systems, etc).



This post also covers steps to analyze pick area activity using task history from warehouse control board.

A replenishment policy for pick face involves deciding the following factors:

1. Should there be a pick face? :  The answer clearly depends on the business scenario. If the warehouse is involved in distribution activities where it sources or manufactures goods in bulk (think pallets) and fulfills a large number of smaller orders in lower handling units (think cases or each), a pick area can often boost productivity and order velocity significantly. In Oracle WMS world this pick area should be defined as a sub-inventory.

2. What items to stock in the pick face? : Clearly all items need not exist in pick face at all times. If that were to happen, the resulting bloat in pick face will obliterate any productivity benefits arising from a small and compact pick area. The best practice is to have as many "Fast"  moving items stored in pick face. The definition of "fast" moving is important: an item is fast moving if it is picked from pick face at a higher frequency. Therefore its important to analyze the demand from pick area only. It does not matter what is the total pick frequency of the item across the warehouse or what is the value of the item or how much is the average pick quantity. In Oracle WMS world, the item's task history can provide us with insights about how frequently an item is picked from a given sub-inventory on a historic basis. Using this insight, its possible to make a decision  about whether to stock an item in pick sub-inventory or not. If the item is stocked in pick area, item sub-inventory relationship needs to be defined in Oracle WMS for the item and pick sub-inventory including the replenishment parameters. 
   

3. Where the item should be slotted? Should the locator be fixed or floating?: Clearly an item that is picked often in the pick face should be stored in the most optimum location for picking. However if demand profile is seasonal or highly erratic, its possible to have a floating location for an item. This way Putaway rule in Oracle WMS can dynamically slot an item based on certain characteristics e.g. Slot in the golden zone during "High" season or slot in the high season when item is flagged as a "Promotional" item. However if the item has a high steady demand, a dedicated locator can be identified for the item. In oracle WMS, its possible to dedicate a locator for the item using the item-sub inventory relationship.

4. How much quantity to stock in the pick face?: One of the objectives for having a pick face is to avoid going back to your reserve area often. Therefore a pick face should stock enough to cover 5-10 picks. Therefore the maximum quantity of an item to stock can be 5-10 times an average pick. If you replenish pick face in a standard pack, you can round this quantity to the nearest pack size e.g. 5 cases, 3 boxes, etc.  The maximum quantity is also constrained by locator capacity i.e. the maximum quantity to be stored can not exceed the available cubic volume. For this reason, step 3 and 4 are somewhat iterative in nature.

5. When should replenishment be triggered?: This determines the minimum quantity before a fresh replenishment is triggered. Usually the replenishment should be initiated when the pick area stock has just enough quantity to cover 2-3 picks. Once again the average historical pick quantity can be used to set the replenishment minimum.

6. What should be replenishment lot size? If you replenish pick face in a standard pack, you should set the standard pack quantity as the replenishment lot size.

The replenishment analysis spreadsheet is a tool that can be used to analyze the task information from WMS control board. It allows you to classify items using their pick frequency into "Fast", "Medium" and "Slow" moving. This excel spreadsheet is being provided as-is. Please feel free to use or modify it as you see fit.

Note: The Excel Sheet can be used to Analyze only 5000 items. You must extend the Array Formula on item column in the "Frequency Distribution" worksheet beyond A5001 row to include additional items if you are likely to have additional items.

The following steps must be taken to perform the Pick Frequency analysis:

1. Copy task history from Oracle WMS Warehouse Control Board: The warehouse control board can be used to get information for a specific sub-inventory e.g. EACH, CASE, BULK, etc. In order to view the task history in control board, you must check the Completed status on the task Tab. In addition the shipping date range can be populated on the outbound tab before you press the "Find" button. The results on the Warehouse Control Board shows the task history for completed tasks. In order to select the tasks and copy the task history to the clipboard, select all the rows (use the square on the top left hand corner of the spreadsheet), use the left mouse click and select "Copy All Rows". This action will copy all the rows to the clipboard.

2. Populate Task Data worksheet: The task information from Warehouse Control Board can be pasted in "Task Data"  worksheet using Control+V or using Edit->Paste. (look at the bottom of this Excel window to navigate to this worksheet)

3. Configure Item and Quantity Columns: The columns containing item and quantity needs to be defined. This can be done using Range definition feature in Excel. Use Insert->Name->Define to configure which column on the "Task Data" sheet contains the Item and Pick Quantity information from the tasks. By default the item information is assumed to be in the "A" column (A2:A1500) and Quantity information in the "E" (E2:E1500) column. The column and the range should be modified based on the data.

4. Perform Data Analysis: Press "Analyze Data" button to execute the Excel macro that categorizes the items into Fast, Medium and Slow items. This macro will uniquely identify all the items and count the number of times each item was picked. The macro will sort the items in descending pick frequency and categorize the items based on the "Fast", "Medium" and "Slow" cutoff fences. It will also compute the avaerage quantity per pick for each of the item