Achieving Leading Distribution Sytems by Developing Warehousing Procesess

We never cease to be bombarded with yet more methods for business improvement. And there is an implicit expectation that we should not only understand them all but be implementing them if we want to keep up-to-date and competitive. Certainly they keep consultants in business. You will no doubt be familiar with a number of these:

  • Total Quality Management
  • Business Process Re-Engineering
  • Continuous Replenishment Programmes
  • Co-Managed Inventory
  • Vendor Managed Inventory
  • Supply Chain Re-engineering
  • Efficient Consumer Response
  • Lean Thinking
  • The Agile Company.

And they can all be applied to the supply chain.

However, distribution networks invariably contain one or more warehouses which act as a buffer between manufacturing and customer requirements and it is essential for these to play their part towards overall goals of meeting customer requirements efficiently and accurately.

Warehouses should no longer be regarded as places for storing goods but rather as an extension of the production process. It is in warehouses that goods are converted from the form in which they are output from factories, typically palletised, into forms acceptable to wholesalers, retailers and individual consumers. Goods may need to be delivered in cages, returnable despatch boxes, cardboard boxes or even jiffy bags. And customers rightly expect goods to be delivered in the correct quantity and on-time, increasingly the next day.

Not only that but orders are becoming smaller, their frequency increasing and permissible delivery times shortening – resulting in demands for improved service level. At the same time as this, there is internal pressure from management to reduce costs through:

  • Reduced administration overheads
  • Smaller warehouses
  • Fewer operatives and trucks.

We also live in a world in which change is becoming ever more rapid. Business and customer demands are changing at an accelerating rate. Today’s solutions may well be inadequate to meet the demands of the next century.

The question is: How can we manage the process of warehousing better?

To address this, attention must be paid to 4 primary areas:

  1. Visibility
  2. Accuracy
  3. Productivity
  4. People.

You may think that the answer is obvious – automate it. In terms of minimising the cost of personnel and overall running costs there is no doubt that this can be an ideal solution. Why is it then that there are so few fully automated warehouses?

The main reason is cost. Equipment requirements are greater and software requirements more extensive. Whereas payback for equipping and controlling a manually operated warehouse is typically in the order of 2 years, for an automated warehouse it can be 5 years or more, and in the UK in particular, it is payback that most frequently takes priority.

Other factors are the difficulty or impossibility of automating picking. Whilst A-frame or cartridge picking has been applied in certain high volume businesses such as Avon Cosmetics and AAH Pharmaceuticals, it remains a relative rarity. By comparison automated storage and retrieval systems for full pallets are relatively common.

The third factor is inflexibility. There are several examples of automated warehouses which were fine for what they were designed to do but which are no longer suitable for current business demands, such as a shift from pallet handling to case picking. In consequence well over 90% of warehouses in the UK remain manually operated.

How then can we achieve leading processes within manually operated warehouses in which picking is a significant activity?

Looking at the 4 areas in turn:


By visibility I mean knowing precisely what you have got, where it is and exactly how much of it you have at any point in time. If you don’t know what you’ve got, or where it is, then you can’t you control it. It is fundamental.

Yet how many warehouses do you know where the stock figures relate to last night’s closing position, or are 2 or more hours out of date because order picking confirmations or pallet movements have not yet been keyed into the computer system.

If there is an urgent order, how can you be sure that the stock is available in the pick face? And how can you co-ordinate replenishments so that they arrive at the pick face neither too early, with consequent additional handling, nor too late resulting in a stock out during picking? To operate efficiently stock must always be in the right place when it is needed.

Visibility of stock, future work, current progress and past movements is also essential to enable:

  • Planning of work
  • Monitoring of progress
  • Analysis of performance

Two other critical requirements are:

  • Traceability of stock – essential in the food, drink and pharmaceutical industries

and, often overlooked when considering requirements:

  • the ability to prove that what you have done is correct.

How often do you hear people complaining about the number of customer claims – customer queries as they are euphemistically called. If you add up the cost of the supposedly missing stock or credit notes it can amount to a significant sum – 10’s of £000’s per month in some instances. Until you can prove that you delivered the correct quantity of product with the correct sell-by date these claims will keep on coming. Unfortunately not all customers are as honest as they should be.

The key to visibility of stock and processes is REAL TIME CONTROL, where all events are known when they happen and all planning takes account of the situation NOW, at the time when the movement is required, and not the situation that may have existed 30 minutes or 2 hours ago.

For example, if a pallet is to be put away, real time control implies that we select the most suitable location when the fork truck operator picks it up. To select the location earlier means reserving empty locations that are potentially further from the ideal location at the time when the pallet is actually moved.

And in a picking scenario, real time control implies that pick instructions are only issued for stock which is known to be in that particular pick location, and not in reserve or awaiting a replenishment move.


Accuracy links with Visibility. Clearly there would be little benefit in having visibility of stock if the information were inaccurate.

Accuracy is important not only in knowing what you have got and where it is but also in every stock transaction, whether a movement or simply a change of status. Historically a wide range of methods has been used to improve accuracy. Invariably these are labour intensive, such as checking processes.

Although effective, being labour intensive means that they are costly and being retrospective means that you cannot be sure picking is correct until checking has been completed. Surely the aim should be to make picking and other stock transactions accurate by trapping any potential errors at source.

Is there an alternative? Is 100% accuracy achievable? In simple terms – YES.

The most commonly used method is by use of radio data terminals (RDT’s) which are typically truck mounted or hand-held. Voice terminals are also attracting an increasing level of publicity and I regard the technology as having improved to a level where a good ergonomic solution can be offered.  Indeed in a number of applications higher pick rates are being achieved than would be the case with an equivalent radio terminal system.

Radio and voice terminals themselves, however, are not a panacea, any more than any other technology – it depends how they are used. In particular they are not a simple substitute for paper and anyone who designs a system to do the same job as the paper previously in use will not be getting the best out of it.

With paper based picking it is typically the case that there are errors in:

  • The identity of product
  • The quantity picked
  • And the quantity in the pick face.

By appropriate procedures using radio or voice terminals all of these errors can be eliminated. But they should also be seen as an ENABLING technology, by which I mean that they enable us to achieve methods of operation and efficiencies that would not otherwise be possible.

Examples of these include:

Dynamic picking In which the contents of the pick face vary according to demand.
Line picking or mobile order assembly In which the pick face is eliminated and loads assembled directly from bulk.
Problem solving In which instructions to compensate for “location occupied” or “insufficient good stock” can be given to the operative on the spot without the need for supervisor intervention
Co-ordination of movements Such as co-ordination of replenishment movements with picking or double cycling of truck movements.
P I checking In which inventory checks can take place at any time, during normal operations or integrated as apart of picking.
Cross docking In which goods for outstanding orders are identified as a part of the Goods In checking process.
Splitting of receipts In which receipts may be split between outstanding orders, the pick face and bulk stock locations


Achievement of optimum productivity requires the application of a number of general principles:

1.    Design for efficient operation

2.    Minimise unproductive movements/activities

3.    Minimise access times

4.    Aid the next process

5.    Minimise interruptions

6.    Retain flexibility

The above list is generalised and can be applied to any warehouse.

1. Design for Efficient Operation

It sounds obvious yet there are so many examples of warehouses which are inefficiently laid out or have the wrong type of storage for what is being stored.

As warehouses become more and more a part of the distribution process, as opposed to a place for simply holding stock, they must be laid out and equipped for movement and not storage. Whenever I visit warehouses containing drive-in racking, I invariably find mixed products in some of the storage lanes. To access products at the back, all those in front must be moved. What would be more appropriate is a quantity of shorter, smaller capacity lanes and some racking in which singles and small quantities of pallets can be placed.

At another VNA warehouse I visited recently, pallets awaiting putaway into narrow aisle locations were placed on the floor and extended well into the marshalling area. The use of multi-level pick and drop locations at the end of each run of racking would have avoided this.

2. Minimise Unproductive Movements and Activities

This ties in with concepts such as the Japanese “muda” – meaning “waste”, i.e. any human activity which absorbs resources but creates no value. Examples within warehousing include:

  • Storing stock which is unlikely ever to be sold
  • Holding more stock than necessary
  • Trucks travelling unladen
  • Keying in of data when EDI or bar code scanning could be used
  • Processing of internal paperwork. Can paperwork be eliminated|?
  • Detailed checking of prior work. If it is picked accurately in the first place then checking becomes superfluous
  • Moving stock to access other stock
  • Two person picking

One should be asking “Can the number of movements be reduced?” Cross-docking would be ideal whether automated or manually controlled. Failing that, then can movements be reduced to one on the way in to the warehouse and one on the way out.

3. Minimise Access Times

This prompts questions such as:

  • Is the picking method optimal for the required throughput? For example, it is well known that vertical storage carousels, or paternosters, are slow to access if the travel distance between shelves is large. So whilst they may be suitable for slower moving lines, they are less likely to be suitable for fast moving lines. One solution to this is for one picker to pick from multiple carousels – 5 in the case of Scottish Hydro Electric in Perth.
  • Is the picking route as short as it reasonably can be? This implies zoning of the warehouse. Typically fast moving stock will be stored closest to the despatch areas and stock arranged across the warehouse in stacking sequence i.e. heavy to light. Another solution is for there to be a consolidated pick of slow movers and for these to be placed in a temporary location within the fast pick area. If there are insufficient low level locations for the product range then further locations should be concentrated at the next most accessible level rather than randomly at all levels. Or a dynamic pick face could be introduced. This would enable all items requiring picking to be concentrated at low level.

The goal is for pickers to spend the maximum amount of time picking and the minimum amount of time travelling or doing administration tasks.

4. Aid the Next Process

This implies storing reserve pallets or other stock close to its final destination in the pick face. Even if not above the picking location they should be in the same aisle. In turn this will minimise the time required for the subsequent replenishment movement.

Similarly, on Receiving can we allocate Goods In docks that are the closest to the zone of the warehouse where the goods will be stored.

Is the fast picking area, where most of the picking activity will take place, close to the despatch marshalling area?

5. Minimise Interruptions

By this I mean simplifying the activities and procedures that divert an operative from his primary role. Examples include:

  1. Minimising the time taken for an operative to acquire his next job. Traditionally it has been necessary to fetch picking lists and other instructions from the warehouse office – an unproductive activity. With correct use of radio or voice terminals these can be relayed directly to the picker as soon as he has finished his previous task.
  2. Problem solving is another activity which can be time consuming. In some warehouses supervisor assistance has to be sought for a decision on what to do, or a truck operator found in order to have a replenishment pallet lowered. This is unnecessary as all common problems, such as occupied locations, stock outs and damages can be recorded and controlled through the use of the mobile terminals and appropriate supporting software.

6. Retain Flexibility

When designing a warehouse process it is inevitable that the design will relate to today’s requirement. It is also likely to take in the requirements of the next 1-2 years. However, if recent history has taught us anything it is that the pace of change is getting faster. And this affects warehousing as well as other processes.

The warehouses may still be there in 5 or 10 years time but what business processes will it be required to support. Will the percentage of case picking be the same as now? Or will cross-docking have been introduced?

This implies 2 things:

  • Equipment and storage facilities must be capable of supporting different styles of operation.
  • Warehouse management systems must also be capable of being adapted to new methods. In particular one must be aware that not all software packages have the same level of flexibility and many may be inadequate to meet the needs of the future.


The importance of people cannot be over estimated. However good the warehouse management system, it cannot stop people making mistakes, it can only help them identify and correct the mistakes quickly.

If the improvement process is to be successful, the people who will be operating the new system must be involved if you are going to get the most out of it. This means early exposure to the new methods of working and the opportunity to contribute their own ideas.

For example, at SmithKline Beecham a specification document was drawn up by the project team over a period of months and was thought to be perfect. On reviewing it with warehouse staff 2 observations rapidly emerged that had not been allowed for previously:

“What happens if I want to skip a pick location and come back to it later?”

“What if 2 picked pallets are of unequal height and I want to level them off?”

There is no substitute for getting people that will actually do the work involved and motivated and this is born out by experience at a number of sites.

As Chris Sherriff, Distribution Development Manager at SmithKline Beecham put it:

“If you wish to reduce errors and improve accuracy, it can only be accomplished by the involvement of people carrying out the day-to-day activities. Even if the system is perfect, unmotivated people are not.”



A good example of what is achievable is provided by the installation at Johnson Brothers in Dublin.

Johnson Brothers are the sole agents in Ireland for around 30 big name UK manufacturing companies including United Biscuits, Proctor and Gamble, Gillette, and Campbell’s Soups. For some clients Johnson brothers provide a total service package including marketing, selling and distribution. For others, such as Bristol Myers, Johnson Brothers provide a more limited range of services covering import warehousing, delivery and cash collection.

Their customers cover the complete spectrum of retail outlets, from multiples, through wholesalers to individual chemists and newsagents.

Coincident with the installation of a WMS, the warehouse was expanded to its current 96,000 ft2. The majority of products are stored in 6,000 locations of wide aisle pallet storage, zoned by client and product type to take account of weight and product compatibility. Picking of these products is from floor level locations onto powered pallet trucks.

Slower moving stock is picked from bays of carton live storage incorporated into the zone for each client’s products. High value attractive items, such as cosmetics are held separately in secure areas.

Throughput is 500 orders/day totalling 10,000 lines and is picked by a team of 21 order pickers.

Why WMS?

Operating on a manual system and expanding quickly as they took on new clients, Johnson Brothers looked to address a number of quality issues. As warehouse transactions grew so did the operational and financial implications. Due to the considerable success and growth, the warehouse became slow to react and they were marking product out of stock when in fact it was in stock and consequently were losing sales. They were also suffering from a high “paper” stock loss, promotional pack product cross-overs, and difficulties with accounting, reporting and documentation.

To control the warehouse efficiently a specification was therefore drawn up for a warehouse management system, and to meet the requirements of the range of products handled this had to be highly flexible. Typical of the requirements were the inclusion of:

  • Multi-client
  • Multi-warehouse
  • RDT or paper based control
  • Highly flexible picking
    • bulk
    • cases
    • singles
    • by load
    • by drop
    • by single order
    • by multiple order
    • consolidated with secondary sortation
    • onto pallets
    • into despatch cartons
    • in system controlled sequence
    • by starting at any point in the order
    • by specifying the aisle the picker wishes to pick from
    • by skipping locations and returning to them later
    • paper pick with confirmation by supermarket style flat bed scanner
  • Task management module with 3 methods of workings:

o   All moves are directed by the system, ie operator logs onto “work”.

o   Operator selects an activity, e.g. retrievals, and is then directed by the system

o   Operator selects an activity, e.g. putaway, and the pallet he wishes to move

When using the task management module, all tasks are allocated an initial priority. If the task is not started within a pre-determined time then its priority level will be increased by a specified amount.

  • Demand sensitive replenishments module:
    • Replenishments are normally generated automatically as a consequence of picking and may be for full pallets or for a quantity of cases sufficient to top-up the pick face
    • The replenishment level is supervisor adjustable
    • If the stock remaining in the pick face is insufficient to meet outstanding orders, the replenishment becomes “urgent” and is given a higher priority

Other features were:

  • Stock checking as an integral part of picking
  • Route planning
  • Proof of delivery procedures
  • Comprehensive returns and collections module
  • EDI interfaces to customers

Armed with a shortlist of potential suppliers they spent time with each, assessing their ability to do the job. Whereas most of the companies wanted to change the operation to match their software, there was one that had provided most of the functions previously and were able to combine them to provide a perfect fit to Johnson Brothers’ requirements. This matches my general philosophy that operational software should be tailored to match the business rather than the other way round.

An added benefit was that the supplier had their own range of rugged RDT’s which were an integral part of the solution.


The warehouse management system was introduced in 2 stages. First to go live was Goods In and control of bulk stocks, and this was followed by Goods Out which was implemented on a phased basis, client-by-client over a 4 month period.

Critical to the process was getting all the core data correct, bar codes, pack sizes, weights etc, which were essential to enable the system to be able to plan correctly.

Management concerns about possible staff difficulties in adapting to the new computerised system proved largely unfounded and as the Distribution Manager describes, “they took to it like ducks to water, and within 48 hours the majority of them were operating very confidently.”

Johnson Brothers are more than happy with the system, “We have a product that meets or exceeds our expectations. It was implemented successfully without disruption to our business, and that speaks highly of both our IT team here and the supplier. Not only that but, it went totally to budget.”


“Astronomical” is how Donald Alford describes the benefits Johnson Brothers have accrued from their WMS system.

Specifically these include:

  • 98% reduction in stock loss to below 0.1% of turnover
  • Reduction in the number of deliveries with errors from 17% to under 3%, within 3 months of completing installation. 3% may sound high but it should be remembered that this is for an average order size of 200 lines so the incidence of errors in order lines is actually less than 1 in 6,000
  • Fewer credit notes
  • Less administration
  • Faster payments
  • Increased effectiveness in purchasing and sales through accurate information
  • Increased sales resulting from improved service level

To quote Donald Alford, Distribution Manager at Johnson Brothers:

“The spin-offs are dramatic, but much more important than that, we are now the leading supply company in Ireland, in terms of the systems and facilities we have and the service we can provide to our customers on behalf of our principals. We are back at the leading edge of the distribution industry.”

For further information on Johnson Brothers visit www.xxxxxxxxxxxxxxxxxxxxxx


Such levels of benefit are not unusual. SmithKline Beecham went though a similar process some time earlier.

Their warehouse is in Welwyn Garden City and was established to receive products from the two production sites in the UK (Worthing and Crawley) as well as from continental suppliers in Belgium, France and Germany.

Three primary ranges were to be handled, namely pharmaceuticals, including vaccines, animal health products and promotional items.

Despatches are to a variety of distributors, wholesalers, retail outlets and individual GP’s, a wide range across which shipment volumes vary from a single vaccine to several pallets of goods.

Storage is split between a bulk store comprising 3,200 pallet locations of high bay racking accessed by narrow aisle trucks, and pick and pack stores for:

  • UK despatch
  • Export
  • Promotional

Making a total of 6,000 pallet locations including 100 in cold storage with a total of 3.000 SKU’s.

Throughputs are currently about 100,000 orders/year, totalling 325,000 order lines.

Prior to the introduction of the warehouse management system, operations had been controlled by a paper based system with complicated computer control. Problems included:

  • Poor service, frequently due to failed overnight batch processes
  • Difficulties in achieving accurate batch control
  • Inaccurate stocks
  • The need to close at 11.30 ever Friday for a stock check

To meet the distribution needs of a wider project known as COMPASS (Customer Order Management Process and System Support), the OPUS warehouse management system was selected, primarily because it was designed to be tailored to the physical processes of the customer’s site, and this rapidly enabled a number of benefits to be achieved.

  • Reduced stock adjustments
  • Accuracy of inventory
  • Correct stock rotation
  • Streamlined administration
  • Optimal use of personnel, equipment and space


Specific quantifiable benefits included:

  • Ability to enter urgent orders without loss of control
  • Reduction in order lead time
  • 35% reduction in stock holding
  • Returns room virtually emptied
  • 41% reduction in pick errors, 18 in UK Pharmaceuticals in 12 months
  • Resources reduced by 40%
  • 20% increase in quantity of orders
  • Ability to measure individual performance
  • No need to stop other work for stock checking


Warehouse processes can be improved by focusing on:

  • Visibility
  • Accuracy
  • Productivity

And the drivers and results of improvement processes are illustrated by the experiences of Johnson Brothers and SmithKline Beecham.

To quote Donald Alford, Distribution Manager at Johnson Brothers:

“The spin-offs are dramatic, but much more important than that, we are now the leading supply company in Ireland, in terms of the systems and facilities we have and the service we can provide to our customers on behalf of our principals. We are back at the leading edge of the distribution industry.”

Originally presented at an ICM conference on Pan European Logistics

Leave a Reply




You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>