Month: March 2011

PIECE PICKING:WHICH METHOD IS BEST?

/ Kurt Hatlevik / 2 Comments

In the following blog posts I will focus on the picking process in a warehouse. There are extremely many combinations in this area, so it is worth understanding what we meet when we will be out there and selling Dynamics AX. There will be follow up postings posts later with picking options in Dynamics AX and also what picking strategies that are available in WM&D. This article shall lay out a 10 step plan for evaluating the multitude of alternatives available in designing a piece picking system.

Area of Concern

Piece picking is one of the most labor intensive processes within a distribution center and as such, it is a major area of concern for operations professionals. As a result, operations professionals are often asked this question… “Which piece picking method is best?”

For the purposes of this article, piece picking is defined as any order selection process that requires the product (units) picked to be individually handled and/or placed into an outer carton, tote, or other container before shipping.

In comparison to full-case or pallet picking operations, piece picking typically has lower unit volume, lower revenue, and higher labor costs. The higher labor costs are accrued due to the need to open vendor cartons, pick SKUs less densely, erect/pack/seal shipping cartons, and apply carton labels while picking less volume per item than full case or pallet picks. The movement towards just-in-time supply chains and the proliferation of direct-to-consumer catalog/internet fulfillment operations is driving a change in order profiles. Smaller orders (those with fewer line and piece counts) are being received more often in many industries and market channels.

To design or re-design a picking operation, one needs to develop a cost versus benefit analysis to include the systems to be considered. In order to fully analyze the nearly innumerable alternatives available, one can follow a 10 step plan to narrow down to the feasible options as illustrated below:

Step 1: Selection of Possible Storage/Picking Types

Mine available historical operational data, apply design year projections, and profile item inventory, movement, & unique characteristics to determine the alternative pick types available (cubic velocity by SKU). In an over-pack picking system high density drawers, bin shelving, carton flow rack, vertical lift modules, and carousels may be among the candidate infrastructure to be considered.

Systemic support and overall capital budget constraints may eliminate some of the more mechanized alternatives. The matrix below provides some high-level rules of thumb for pick type selection.

Step 2: Mode of Order Transportation

There are two common options in selecting the method used to transport the order to and from each pick location:

  1. Pick to cart
  2. Pick to conveyor

A decision is typically made based on the overall size of the piece picking area (travel requirement). SKU velocity, pick density, and related considerations often define the appropriate method.

A combination or hybrid of the pick to cart/pick to conveyor option is commonly used, in which, fast pick items would be picked to conveyor and slow picks would utilize a cart. Conveyor could be used to transport orders to and from each cart pick zone. The picker may load the cart at the beginning the zone and unload to cartons or totes to conveyor at the end of the zone.

 

 

Step 3: Selection of Picking Technology

Operations with a large number of picks in a small area benefit the most from pick to light technology, while operations with picks spread across a large pick area (many pick facings) are better suited for either RF or voice picking.

The benefits of voice versus RF is that voice is “hands free”, the picker never has to put down (or holster) an RF device (scan gun) to either pick or open cartons. In a voice system, picking time can be reduced by the picker listening to the voice directive while traveling to the next location instead of stopping to read the RF device.

The RF device, however, has the ability to capture detailed item information, typically a production batch or serial number, quickly by scanning item barcode labels. Voice may be more cumbersome in these applications.

 

 

 

Step 4: Order Batching

Order batch picking is the process of grouping orders together to create a larger pick order, which reduces the overall pick lines in an operation versus a discrete order pick (picking of one order at a time). Since multiple orders are being picked into one container, in most cases this will result in the requirement for an order sortation station, an area designed to split the pick order into the individual customer orders (discrete orders). If the order batch contains all single unit orders or all orders of the same item, the batch will typically not require a separate sortation station, the items are allocated to a specific order at the time of packing or manifesting. Hardware costs based on number of users, not locations Hardware costs based on number of users, not locations Hardware costs based on number of users, not locations Hardware costs based on number of users, not locations

 

 

Step 5: Batch Sortation Technologies

Much like the picking technologies, batch sortation can be performed utilizing paper, RF, voice, or a put-to-light system (which is a reverse of a pick-to-light system). Since a piece pick batch sortation is typically in a condensed area, usually consisting of sections of carton flow rack or shelving, the methods that are most commonly used and are the most effective are RF and put-to-light. Put-to-light is more expensive, but the productivity rates are considerably higher, 100-150 lines/hr for RF versus 200-over 250 lines/hr for put-to-light (broad industry estimates).

Step 6: Cluster Picking

Cluster picking is a methodology of picking into multiple order containers at one time. The containers could either be either totes containing order batches, discrete order shippers, or discrete order totes. There are two main piece picking systems that benefit from the use of cluster picking:

  1. Pick to cart operations, in which, a cart would be loaded with multiple totes or shippers and the picker will make one pass through the pick zone and sort to the pick containers, thereby avoiding unproductive travel.
  2. Vertical lift modules (VLM) or carousel operations utilizing the independent zone picking technique, the picker would setup the pick container batch, this would initiate the mechanized pick modules, and the items would be sorted to the proper pick container. This technique is used to optimize the VLM & carousel pods by limiting the number of machine cycles.

Step 7: Cluster Picking Sortation Technologies

The same options and reasoning that applied to batch sortation applies here. The same task is being performed on a smaller level, typically sorting into two to four pick containers. The most common cluster picking system contains a put-to-light cart, whereas the picker utilizes RF or voice picking technology.

Step 8: Select a Picking Container

Pick to tote operations typically utilize order batching & sorting techniques that reduce the picking labor greater than the additional labor required for packaging and sorting.
Pick to shipper operations are executed in one of three ways

  1. Picker erects the cartons at time of picking. Requires the replenishment of cartons to multiple areas.

  2. Cartons are erected and transported to the pickers via separate conveyance system (i.e. monorail). Requires added capital equipment expenditure.

  3. Cartons are erected and inducted directly into the system with the contents pre-allocated. Requires extensive systems support.

 

 

 

Step 9: Multiple Pick Zone Order Routing Methodology

Pick to tote operations with order batching lends itself to the independent zo ne pick method, a process in which segments of each order are picked independently in various zones and directed to a consolidation area where the elements of the order are combined to the shipper (batch sortation area).

Pick to shipper or discrete order pick to tote operations are typically most effective in either a pick and pass or zone routed system. This routing method will reduce the overall number of totes or cartons in the system. A zone routed system although more complex and costly, will significantly improve productivity and work flow especially in operations in which an order requires picks from multiple zones. Unlike a pick and pass system, zone routed totes or shippers are only touched by those people in zones where product needs to be picked. All touches are productive as a result.

 

Step 10: Order Consolidation Methodology

There are several methods with which to consolidate an order before shipping. If your operation is shipping only parcel orders, this might not apply if each carton is shipped independently of each other and no physical consolidation is necessary. For operations that must keep order integrity at the shipping dock, the common techniques are as follows:

  1. Consolidate the order before packing occurs when either pre-picking orders before all items are available or discrete order picking to a tote with a 100% order check or to reduce outbound cartons.
    1. Manual sortation on conveyors – human sortation system, a person is assigned the task of sorting the totes to gravity conveyor lanes typically directed by RF scan
    2. Automated sortation on conveyors – requiring controls and diverts
    3. Tote mini-load system AS/RS (Automated Storage & Retrieval System) – stores all totes until order is complete then releases entire order to packing in a “slug” or “train”.
  2. Sortation is commonly required after packing, typically for LTL or TL shipments.
    1. Manual sortation on conveyors – human sortation system, a person is assigned the task of sorting the totes to gravity conveyor lanes typically directed by RF scan
    2. Automated sortation on conveyors – requiring controls and diverts

 

Conclusion

Once the options available for your operation are narrowed down a cost versus benefit analysis must be performed for each remaining alternative. Some questions that need to be answered are:

  1. What is the projected labor cost?
  2. Does the alternative meet the projected throughput requirements?
  3. What are the initial capital costs?
  4. Are there any on-going costs & expenses?
  5. How much space does this system consume?
  6. Does it meet order accuracy goals requirements?
  7. Does the alternative meet the required packaging standards and that of your customers?
  8. Is the system flexible and scalable?
  9. What are the risks associated with deploying such a system?

 

Source : http://www.OPSdesign.com

Cross docking (Three)

/ Kurt Hatlevik / Leave a comment

As I promised in my previous posting I would now like to go through an approach on transportation cross docking in Dynamics AX. This article is meant for experienced Dynamics AX professionals that have deep knowledge both functional and technical to Dynamics Ax, and is a follow up to the last two articles on Cross docking (See Cross docking I (one) and Cross docking II (two))

First a small reminder what transportation cross docking is. Transportation cross docking is used in distribution scenarios, where you have pure cross docking services. The cross docking facilities are often not the legal owner of the goods. The following figure shows the supply flow.

Here the cross docking warehouse is a pure service provider and do only push forward the goods from received from the vendor and internally transport the goods to the correct outbound dock. The warehouse A is not the cost owner of the goods, and it prefers pallets or cartons. At the distribution warehouse B, C the goods are properly received. Some are placed into the warehouse, but some is cross docked again to the outbound dock.

So what do we need to achieve this in Dynamics AX? As far I have seen it is possible, but will require many manual processes. Here an extension to dynamics AX is required. The first steps are to identify the main concepts needed.

Inbound logistics

I would often describe the inbound freight as a trailer or container that delivers goods to a destination or a HUB. If you are using third party bulk transportation, then the cargo is the entity that communication to the carrier is overall based on. The inbound cargo does not have any directly information of the contents, cross docking information, but contains volumetric information, status and delivery time information. Like when it is planned to ship from the vendor/source hub, and when it is planned to arrive to the warehouse/destination hub. There are some industry standards that are commonly accepted, and that is the Cargo 2000 standard, that contains all needed milestones statuses for a carrier. Getting insight into these milestones is vital to be able to plan forward in the supply chain and to be able to plan cross docking. The milestones should be stored as transactions related to the inbound freight to track the goods. Another register called Yard movements should keep track of the goods/trailer/container as it is received at the yard, and scheduled for receive.

Inbound cargo is the first required concept for cross docking, and standard Dynamics AX do not have this term, but in the To-Increase WM&D solution, the Inbound cargo screen looks like this in AX 2009, together with the cargo 2000 milestone statuses:

When receiving goods that should be cross docked, the goods will be directed to the outbound dock, and associated with the right transportation route(See outbound logistics later for a description).

SSCC(Serial Shipping Container Code)

The next requirement is a unique identifier for the goods arriving. This is very often required to be able to handle the incoming goods. With handling we mean being able to plan and execute the needed steps to cross dock. Without a unique identifier this becomes very difficult. What exists in standard AX is the concept “Pallet”. But in AX the pallet is just an identifier, and do not contain any more information than the identifier and the location it is placed. The content of the pallet can be seen as the “On-hand” overview or the transactions related to the pallet.

But in transportation cross docking the warehouse performing the cross docking services do not necessarily have any insight into the purchase order, sales orders or inventory transactions. Another concept is required to support a transportation cross-docking scenario. In WM&D this is the SSCC (also called Outbound Cargo).

The idea behind the outbound cargo, is that it is an entity that do contain all the required information of the SSCC, without having a transactions associated to the SSCC. This is information like ID, status, carrier, addresses and contents information. But there do not need to be any associated inventory transactions related to it. It could be that the SSCC do have reference information at the final receive warehouse/hub. The information on the SSCC is either received as ASN information, or is registered and labeled at the first receive. Labeling is often very important in these scenarios, and must often be carrier compliant.

In WM&D there are also possible to transfer a pallet ID or license plate to a SSCC, and this is automatically done at picking, packing and shipping.

Outbound logistics

Having a good system for outbound freight is vital, because often the planning done here ‘beats the drum‘ for a large portion of the operations. Quite a large portion of the people in a warehouse has tasks related to the picking packing and shipping. In standard AX, the shipments are the tool used for doing outbound logistics. A shipment is a collection of items that are packed in the same container for transport by, for example, ship, rail, truck, or plane. A shipment includes an entire order, a part of an order, or a consolidation of multiple orders. Based on the contents of the shipment, one or more picking routes, one or more internal transports, or both are created. Output order is a request for picking requirements and is the basis of a shipment. From the shipment you can activate an internal transport, a picking route, or both. The shipment status is based on the lowest denominator of the shipment lines status. The shipment is mainly a tool to create picking routes and pallet transports. But more advanced picking, planning and execution is lacking.


Figure: standard AX shipment form.

The architecture in standard AX consists of the following tables :

Figure: standard AX architecture for shipments

As we see here the standard AX do not have any good systems for handling SSCC. It is possible to simulate that the picking routes could “simulate” SSCC, but this leads to many customizations. You may have problems in doing planning and automation, since not much of that exists. Ability to short ship is manually possible, but do require good insight into the transactions. The processes are manual and people must control most of the planning. In WM&D it has been introduced a overlaying system called transportation planning. The transportation planning approach can be visualized like this :


Figure: How transportation planning is related to standard AX processes.

Here all the functionality of std. AX is used, but a planning system has been created. The shipments and picking routes are then the result of the transportation planning.

A transportation route line is the actual route that the transportation needs to go. It can be several pickup addresses and several delivery addresses on a truck route, and each of this combination will have a transport route line. Each transportation route line will have properties like estimated weight, volume and body length/width. A sequence of transport route lines is the destination addresses.

A transport route is in this defined as a planned transportation. The transportation is most often related to a mode of delivery. The transport route has some physical limitations like max payload weight, max volume, max cargo body length/width. The transport route is the “header” of a set of transportation route lines, and is most often represented as a truck or container.

Picture : Transport route and transportation route lines


Picture : Transport route visualized to Google maps. Driving sequence is the numbers on each stop. Truck loading sequence is the reverse of the driving sequence.

The idea here is that all the required output orders will be associated to a transportation route line. These lines can then be planned, either internally in WM&D/AX or in an external route planning tool (or even at the freight forwarder). Then the optimal route can be built, and this is the transportation route. Each line is therefore associated to a transport route.

This gives the possibility to do multisite distribution scenarios with cross-docking capabilities:

Figure : Warehouse D can be set up as a cross-docking warehouse, where items are received from other warehouses or from vendors.

The process will then go through the following steps:

Cross docking II (Two)

/ Kurt Hatlevik / 2 Comments

In the previous posting we looked into the theoretical pure Transportation Cross Docking, where there are dedicated cross docking warehouses. But often cross docking is classified into several categories. Cross docking reduces handling costs, operating costs, and the storage of inventory. That’s why this logistics method is used in various industries, e.g. the retail and grocery distribution. Originally cross docking was solely done by moving goods from one truck directly into another. Nowadays cross docking is able to handle larger and more complex operations. Therefore cross docking distribution centers may include highly efficient automation systems for staging, sorting and consolidating inbound materials for shipping. And even a very short term warehousing function might be included. Here are some cross docking types that is often refered to:

Manufacturing Cross Dock
Use Cross Docking for the receipt, consolidation, and shipment of a pre-defined quantity of raw materials or component parts, typically from many suppliers, to the product plant.

Distributor Cross Docking
Cross Docking is performed on a wide variety of merchandise. Typically it is conducted in operations where various manufacturers, making complementary items, ship their merchandise to a common distributor who assembles the products on a multi=SKU Pallet before delivery to the next level of the supply chain.

Transportation Cross Docking
Transportation companies sort and consolidate parcels and pallet loads based on geographic destination.

Retail Cross Docking
Products are received at these retailers’ distribution centers, moved across the dock, and married with other products bound for the same store. A few types of display-ready-pallet builds for end-users fit the cross dock profile; items are off-loaded but not put-away, the display ready pallet is built, and then released without being put-away. All would have to transpire within 24 hours.

For now, I would like to focus on two main cross docking methods.

Cross Docking I (Transportation Cross Docking) – This process consolidates inbound products from different vendors into a mixed product pallet for the same store, which is delivered to the customer when the final item is received. Logistics companies then deliver the goods to the respective stores. In this scenario the crossdocking warehouse can be a pure provider of cross-docking services. If we compare this to Dynamics Ax, it means that there are no sales orders, purchase orders or inventory transactions. The information on how to cross dock is information either on labels attached to the pallet, or information received as ASN (Advance Shipment Notification)

Cross Docking II
(Distributor Cross Docking) – This process combines shipments from a number of different carriers. All suppliers receive an overall order summarizing the orders of the individual stores. They then transport the goods as one big delivery to the staging area where the goods are repacked onto pallets with other goods from other carriers for the same store. Logistics companies then deliver the goods to the respective stores. This type of Cross Docking is both an advantage for the customer and the supplier as there is only one order or delivery per day or per week.

A Dynamics AX approach to support cross docking II (Distribution cross docking).

Standard Dynamics AX do not have cross docking support out of the box, but I see several companies is using the direct delivery function to achieve cross docking.

By doing this, one purchase line can be linked to a sales order line, and the delivery information is transferred from the sales order to the purchase order. The drawback is that there is a “one-to-one” relationship between the receipt transaction and the issue transaction. There is also an issue that the cross is close to 100% manual. (PS! There is an option to specify a shipment reservation sequence that should look for items in the receiving dock, but this is not a pre determent cross docking ).

But in the real world we see the need of a many-to-many relationship. One receive, should be cross docked to several sales orders, meaning that you may receive a large quantity from a purchase order, and it will fulfill many different sales order to different customers. Often it is even not customers, it could be transfer orders to other warehouses, or retail stores classified as warehouses.

Standard AX do have “traces” of solving this. It is called inventory marking.

Here you do have a many-to-many relation between transactions. It is even documented on the inventory transactions as shown here:

Here the Lot-ID points to the Reference Lot on the inventory transaction. So by doing some manipulation on the inventory marking we can make Dynamics AX keep references between issues and receipts. And this is also something that can be the origin to a cross docking II scenario.

But there is a few more things that needs to be in place to make the cross docking efficient. The first thing is to an efficient way of transporting goods from the receipt dock to staging area/shipping docks. But the transportation system must also support transport to all kinds of areas in the warehouse, like buffer location, picking location, outbound docks or to automated warehouse solutions. This can be visualized like this:

Standard AX do have a pallet transportation logics, even though the looks of it could be improved;

Here the user select the pallet ID, and starts the transport. Arriving to the destination location, the user completes the transport. This screen does the job, but the for a warehouse worker on a forklift, it is difficult to use. A forklift operator wants to use barcodes and/or touch sensitive screens. The other negative element is related to that it only supports full pallets. In a real world scenario we often see that transports are needed on cartons or even on items that do not have a unique identifier. In the To-Increase Warehouse Management & distribution, this transportation system has been extended to support transport on both pallet and carton license plates(also called SSCC), and also transport of goods without having a identifier;

When a good way of identifying goods, and a good transportation method is in place, that it is possible to do cross docking efficiently. As shown in the figure below, the cross docking is then processed in the same way as any other transport:

The screen above is the receive transactions where a single purchase line has been split to 10 inbound transports, where the first is cross docked to picking area, while the others are received into the warehouse. In a cross docking scenario the destination location would be an outbound dock, and the name of the customer would also be visible. To-Increase WM&D also support the ability to create the license plates, and to create the correct labeling for the goods to be cross docked and to the correct customer.

So to conclude in how to make an efficient cross docking with dynamics AX you need:

  1. Build a framework on top of inventory marking to mark how cross docking shoul happen, and that works even when you have a long supply chain.(Going through several warehouses before being shipped out)
  2. Create a receiving system that allows you to receive ASN and the goods, control the inbound transport.
  3. Modify the transportation system to support more scenarios(Pallets, Cartons and items)
  4. Create an efficient system that can be used on forklifts and RF/PDA
  5. Create a label integration to label the goods.
  6. Create an efficient outbound registration system.

My advice is…. Don’t try this. A better solution is to look to proven ISV-solutions that focus on this, and my advice is off course To-Increase Warehouse Management & Distribution.

In my next posting I will look deeper into how to implement cross docking I (transportation cross docking).

Crossdocking I (one)

/ Kurt Hatlevik / 2 Comments

 

Crossdocks are high speed warehouses. If an arriving item has already been requested by a customer there is no need to store it as anticipation inventory; instead, the item can move directly from receiving to shipping, without intermediate storage and retrieval. Thus the item can move much more quickly through the facility and the most costly part of warehouse labor can be avoided.

In a high-volume crossdock the turnover times may be measured in hours. To support this velocity of movement, a crossdock may be nothing more than a slab of concrete with a roof and walls punctuated with doors for trailers. Freight is pulled off arriving trailers, sorted and loaded onto departing trailers without intermediate storage.

There is little or no storage provided in a crossdock because items do not stay long enough; but there is generally a lot of material-handling equipment, such as forklifts and pallet jacks, to move freight. Labor is frequently the main cost and it is devoted to unloading incoming trailers, moving the freight to the appropriate outgoing trailers, and loading. Consequently, the issues within a crossdock are those of material-handling and product flow rather than location and retrieval.

Why crossdock?

The biggest reason to have a crossdock is to reduce transportation costs.

This can be achieved by consolidating multiple shipments so that full truck loads can be sent. The Home Depot is a major retailer and the largest user of Less-than-Truck-Load (LTL) shipping in North America. (LTL means sending shipments that do not fill a trailer and so are not economical to send by themselves. Instead, an LTL freight company consolidates many such shipments and so achieves efficiencies.) At the present writing, LTL costs about twice the cost of Truck Load (TL) shipping, so there is a strong incentive to fill trailers. The Home Depot has begun doing this by having vendors ship full trailers to its crossdock. (The trailers are full because they hold product for many stores.) At the crossdock the product is sorted out for individual stores and consolidated with product from other vendors bound for the same store. The result is that each store has enough freight that it or it and a few close neighbors generate a full truck load from the crossdock. The result can be considerable savings.

Additional benefits include less inventory (because all product flows right through) and less labor (because product does not have to be put away and later retrieved).

Operations

Most crossdocking freight terminals are laid out as long, narrow warehouses with doors around the perimeter.

Figure : View from above of a typical high-volume crossdock, which receives freight, sorts, and disgorges it. Each door is devoted to either arriving trailers, which are unloaded, or to departing trailers, which are loaded. Ideally, freight should flow directly across the dock rather than along its length.

A typical terminal, where the small shaded rectangles represent incoming trailers with freight to be unloaded, and small clear rectangles represent (empty) outgoing trailers. Terminals range in size from fewer than 10 doors to more than 500 doors.

Inside a terminal, a variety of material handling methods is used to transport freight. Forklifts and palletjacks carry heavy or bulky items, and carts transport smaller items. In addition, large terminals may have draglines, which circulate carts around the inside perimeter of the dock.

There are two types of doors in a terminal: receiving, or strip, doors, where full trailers are parked to be unloaded, and shipping, or stack, doors, where empty trailers are put to collect freight for specific destinations. Once established, the designations of these doors do not change, although the trailers parked at them will. A shipping door always receives freight for the same destination. A receiving door may be occupied by any incoming trailer, regardless of its origin or contents.

Arriving trucks may deliver their trailers directly to an unoccupied receiving door; or, if none is available, they may place them in a queue. After the trailer is backed into a receiving door, a worker unloads the freight. After unloading items of a shipment onto a cart, the worker walks to the destination trailer and loads the items into that trailer; or he places the cart on the dragline, if the terminal is so equipped. To handle pallet loads, the worker uses a pallet jack, or hails a forklift driver, or finds a forklift and delivers the load himself.

After a trailer has been completely stripped, a driver replaces it with another incoming trailer from the queue of trailers waiting to be stripped. After an outgoing trailer has been filled, a driver replaces it with an empty trailer to be filled with freight for the same destination.

Freight flow

The patterns of freight flow within a terminal—and therefore the work—are determined by:

Layout
by which we mean the specification of doors as either receiving or shipping doors and the assignment of destinations to the shipping doors.

Geometry
The shape of a terminal determines the travel distances between doors and the susceptibility to congestion. (For example, narrow docks tend to be more congested because workers have less room to manoeuver.)

Material handling systems
For example, pallet jacks are slower than forklifts, but they may be more available; draglines reduce walking time, but can impede forklift travel.

Freight mix
For example, terminals having a higher mix of pallet freight require more forklift travel than those receiving a majority of carton freight.

Scheduling In real time, the dock supervisor determines freight flow patterns by assigning incoming trailers to receiving doors.

Changing the geometry or material handling systems of a terminal is expensive; changing the freight mix is a marketing decision with implications outside the terminal. The two remaining ways to take work out of the system—change the layout or change the scheduling—are inexpensive. In particular, the layout can be changed simply by changing the labels on the doors of the crossdock.

There are two kinds of doors on a typical crossdock: Those reserved for outgoing trailers (for example, the “Miami trailer”) and those reserved for incoming trailers. The outbound doors are reserved for specific destinations but the incoming doors are not so specific and may be used by any incoming trailer (because, while departures are scheduled to specific destinations, the terminal does not have full control over arrivals).

Congestion

As more freight flows across a dock, congestion increases, which interferes with the flow.
There are several distinct types of congestion on a crossdock:

Competition for floor space: Freight may be docked outside a receiving door if, for example, it consists of many unpalletized cartons going to the same shipping door. Then there is an incentive to accumulate it all so that fewer carts must travel to the destination door. On the other hand, freight is very likely to be docked outside a shipping door while the loader figures out how to pack the trailer tightly. When several nearby doors compete for space to dock freight, some invariably interferes with other traffic. At the very least, it takes longer for a worker to manoeveur through the docked freight.

The effects of docked freight are most severe near the inside corners of the dock, where there is less space per door, as shown in this figure.

Figure: There is less floor space per door at outside corners and therefore more likely to be congestion that retards movement of freight.

The need to dock freight suggests that busy outgoing trailers be parked away from the corners of the dock.

Interference among fork lifts: Despite the intention of moving freight simply “across the dock”, most doors will be to the left or right of a door with an incoming trailer and so a significant amount of freight must travel along the length of the dock. Most crossdocks set up two forklift “highways”, one along each long side of the dock. (It is a good idea to set up two so that, when one is blocked, some freight can still flow.) However, the flow of forklifts back and forth along the length of the dock may be interrupted by forklifts making left hand turns into doors with outgoing trailers. This effect can be reduced by parking busy outgoing trailers away from the very middle of the dock (which is also the most convenient location). Note that this works opposite to convenience, which tends to push busy outgoing doors towards the middle of the dock.

Competition for drag line capacity: Each door receiving arriving trailers will need empty carts from the dragline and, after loading a cart, will need empty cart positions on the dragline. This means that there will be diminished dragline capacity downstream of this door. If the door is far from a busy outgoing door then the region of diminished capacity can be large. This creates an incentive to intersperse incoming doors with outgoing doors. In particular, this suggests that current practice, which is to create large banks of incoming doors, reduces the capacity of the dragline.

Design

The first decision in designing a crossdock is “how many doors?”.

Generally doors are devoted to one of two types of trailers:

  • Incoming, from which freight must be removed; and
  • Outgoing, in which freight must be loaded

It is easier to unload than to load. A loader must try to get a tight pack and so may have to dock freight and this double-handling slows him down. A good rule of thumb is that it takes twice as much work to load a trailer as to unload one.

To achieve frictionless flow, the capacity for flowing freight into the dock must be balanced with the capacity to flow freight out of the dock. Accordingly, one should plan to have twice as many outgoing doors as incoming doors. Alternatively, one can balance the rates of flow by assignment or workers. For example, if there are equal numbers of incoming and outgoing doors, balance can be achieved by assigning twice as many workers to load. Note, however, that crossdocks with many doors are generally less efficient than crossdocks with fewer doors. The reasons are as follows. A door can only have a few near neighbors on a dock and so a dock with more doors means that each door is likely to have few more near neighbors but many more distant neighbors. This means that in general freight must move farther across a large dock. Consequently, labor costs are generally higher at larger docks.

An additional factor is that on larger docks more freight flows past the central doors, which are the most important because they tend to be close to many doors. In fact, the total flow of freight past a centrally-located door tends to be proportional to the square of the total number of doors. Therefore a dock with twice the doors tends to have 4 times the congestion in front of its central doors, which diminishes their value.

This follows from the following simple model: Imagine a rectilinear dock as a line with 2n doors (numbered from left to right), and assume that equal amounts of freight move between every pair of doors. Then the flow into any door is of intensity O(n). But the total flow passing the area between door i and i + 1 is i(2n – i), which means that the greatest total flow passes by the middle of the dock, door n, past which flows O(n2) units. But these central doors are exactly those that are nearest to most other doors and therefore are the best locations! Thus, as a dock design grows in length, the lengthwise traffic past the central doors increases rapidly while traffic directly across the dock remains unchanged. Increased traffic means congestion, which helps explain why docks can lose their efficiency as they grow. There are few docks larger than about 200 doors. Most are 80–120 doors long.

Do not forget to allow enough parking space in the yard for two trailers for every door. This means that for each origin or destination you can have a trailer at the door plus one full and one empty in the yard. This helps you handle surges in freight flow.

Geometry
What is a good shape for a crossdock? In general, one wants to enable efficient flow of freight from incoming trailers to outgoing trailers.

Typically, a crossdock is a long rectangle, with doors for trailers around it. The capacity of a dock is increased if it has many doors, but without being too close together so that trailers (outside) or freight (inside) interfere with one another.

Figure : A typical crossdock is built in the shape of the letter I (actually, an elongated rectangle), so that freight can flow across from incoming trailers to outgoing trailers.

A typical dock, such as illustrated in the picture, is generally around 120 feet wide (36.6 meters). This is to allow freight to be staged on the floor. A standard (large) trailer is 48 or 53 feet long (14.6 or 16.2 meters) and a “pup” is 28 feet long (8.5 meters); all are 9 feet wide (2.7 meters). The width of the dock should include enough space for the trailer on each side of the dock to stage its freight (about 100 feet total, or 30.8 meters) plus allow space for travel along the length of the dock (for example, two aisles, each 10 feet wide, or about 3.0 meters). We have seen docks as narrow as 80 feet (24.4 meters), but this is practical only when it is possible to avoid staging most freight, such as when the material is palletized and also easily stackable and may be loaded in any order. If a dock is much wider than this, it just adds to the travel time to move the product from incoming trailer to outgoing trailer.

A dock does not have to be shaped like the letter I. For example, shapes of an L, U, T, and H. But every corner in a dock reduces effective capacity:

Crossdocks have been built in a variety of shapes. Clockwise from upper left: An L-shaped terminal of Yellow Transport; a U-shaped terminal of Consolidated Freightways; a T-shaped terminal of American Freightways; an H-shaped terminal of Central Freight

On the outside of a corner you lose floor space per door on which to dock freight.

  • This increases congestion on the dock, which interferes with the flow of freight.
  • On the inside of a corner, you lose door positions because trailers will interfere with each other in the yard. Because doors are lost, the dock must be longer to accommodate a given number of doors, which means that on average freight will have to travel farther to cross the dock. Thus, for example, freight has to travel farther to cross an H-shaped dock, with four inside corners, than to cross an I shaped dock. (Because the door positions will be lost anyway, inside corners are a good place to locate administrative spaces or hazardous materials storage.)

It is hard to make generalizations independent of specific bills of lading; but in general an L-shaped crossdock is inferior: It incurs the costs of one inside and one outside corner but without getting anything in return. The result is that freight must travel farther because the dock must be longer from end to end to make up for lost doors at the inside corner. Furthermore, there is congestion at the outside corner. The same observations hold even more strongly for a U-shaped dock.

An X-shaped or a T-shaped dock also incur corner costs but they have a compensating benefit: The longest distance from door-to-door is less than that for an I-shaped or L-shaped dock with the same number of doors.

Trailer management

One can reduce labor costs in a crossdocking freight terminal by parking incoming and outgoing trailers so that freight can be efficiently moved across the dock. For example, if much of the freight flowing through the terminal is bound for Miami, the Miami trailers should probably be parked in a convenient location. The challenge is to formalize the notion of “convenient”; then labor-reducing door assignments can be made with optimization models based on the geometry of the terminal, the material handling systems within, and the mix of freight passing through.

References : Copyright 1998–2011 John J. BARTHOLDI, III and Steven T. HACKMAN. All rights reserved. This material may be freely copied for educational purposes—but not for resale—as long as the authors’ names and the copyright notice appear clearly on the copies. Corresponding author: john.bartholdi@gatech.edu