Flowstore Logos and Order fulfilment systems for order picking | Flowstore Systems EU

order picking • fulfilment • distribution • warehouse • live storage


Order picking single items is nothing new – mail order companies have been doing it for decades.  Automation techniques have also been available for 20 years, such as pick-to-light.  What is new is the emphasis on shorter time frames and demand for fast payback, low investment and low cost of pick, especially in the area of e-commerce.  Customers who are used to spending a few minutes on the web to select their purchase, are not prepared to wait a long time for their delivery.

This is also true of the traditional retail environment, where EPOS (electronic point-of-sale) technology in the store generates a replenishment pick in the warehouse and new stock delivery to the retail store must be very rapid to maximise the sales trend for the product.

There are various order fulfilment systems.  Firstly, it’s a mistake to say that the warehouse operation adds no value and is merely a cost.  The DVD in the warehouse earns no money.  The DVD correctly picked, produces the invoice and delivered into the customer’s hand will generate the revenue to turn the item’s cost into profitable income.  The picking operation needs maximum efficiency to maximise the profit potential, without incurring wasteful costs.  A warehouse is a machine, which can be simple or complex, for assembly of a customer order.

Many companies are unaware of their ‘normal’ order fulfilment costs as a percentage of total operating costs and this is a basic question which should be investigated.  How many people keep such analysis records?

Let’s briefly outline some of the most common order fulfilment systems.

Orders can be picked as individual customer orders from shelving, racking or live storage with assistance from hand-held terminals or pick-by-light equipment, and involve conveyors, automatic picking machines and carousels.  Orders can also be grouped together, so that each SKU (stock keeping unit) is picked in a multiple quantity and then re-sorted to individual customer orders.  This is known as batch picking and sorting and can be done manually or automatically using conveyor sortation or put-to-light systems.

Picking can be carried out by an operator picking all lines through the Warehouse and this is usually referred to as Linear Picking.  Alternatively, picking can be segmented into zones of activity ie Zone Picking, partially completed orders being passed from one pick area to the next, where another team of operators do the next picks.  In practise, it is highly unlikely that one method will suit the whole of a warehouse pick profile.  Fastest moving goods usually require a different method from the slowest lines.  The first mistake is to imagine a ‘one solution fits all’ scenario.


Data Analysis

Very few companies keep adequate analyses of their order picking requirements.  We are usually presented with a product-by-product total volume movement, with no indication of the varying amounts handled on an order-by-order basis.  Nor is there adequate information of seasonal variations.

If product X moves 100,000 units per year, was that 2,000 picks of 50 pieces, or 50,000 picks of 2 pieces or 500 picks of 200 pieces?  Or a mixture of all of these?  The solutions may be radically different.

An ideal approach is one that gives a picture of an actual day or week based on actual orders processed, product by product, orderline by orderline.

Done at different times of the year this will be far more indicative than a record of total yearly volumes, since it shows what you actually had to do on any given day!  Immediately, you will probably see that only 20%-50% of your part numbers will be active on any one day.  Graphical analysis can then show the true pattern of activity.  Visually you can immediately break down the picking activity to say 3 bands – fast, medium and slow.



When is automation appropriate?  It’s partly a question of higher investment and long commissioning versus lower investment and faster pay back.  Is there a trade off between the ultimate in automation at high cost, high efficiency, versus a lower cost faster pay- back less automated solution?  Sometimes the lack of an available pool of suitable labour, or the need to build warehouses taller due to high land costs, seem to drive some decisions to automate but the question of operational flexibility in markets which change more rapidly than in yesteryear, appears to be sometimesoverlooked.  Perhaps there are solutions to be investigated, which will provide maybe 80% of the benefits of automation for only 50% of the investment.  This is not to decry the genre, but to question whether real alternatives have been studied.  Perhaps a mix of manual and automation can be the most cost-effective.


Pick Speeds

Consider the human hand which possibly took millions of years to evolve and yet still is the most versatile machine available.  What is the ultimate pick rate by manual means?  One step takes 1 second. 1 pick takes 1 second.  The ‘sound barrier’ is thus 1800 picks per hour per person.  Some pick systems are achieving 1200 picks per hour with zone picking and pick-by-light.  If you can save only a quarter of a second per pick on a pick of 100,000 items per day, you save the cost of one picker.  So has anyone yet exhausted the scope of hand picking?  And how do you head towards the ‘sound barrier’?.  There are some basic steps:-

Actual pick rates achieved depend on Pick Density ie the number of picks required in any one zone without the picker moving a significant distance.  Parts to be picked should generally be arranged to put fastest (most location accessions) together, then medium movers, then slow movers.  Probably slowest movers should be picked first.  Pick density is a combination of the travel time to walk through the zone, the time taken to do the physical pick, plus the administration time, eg finding the right location and ticking-off the picked item from the picking list.


Pick Accuracy and Picking Sequence

In computer jargon, rubbish in = rubbish out.  Is your stock correctly put away?  Correct stock replenishment can be done using hand-held scanners.  Pick-by-light and check weighing techniques can also improve pick accuracy but have their own limitations.

The sequence in which orders are picked can be important for picking speeds.  This is especially true when picking multi-orders eg mini batch picking and sorting to intelligent trolley.  If you can sort the order sequence to put picks for similar items together (ie adjacent locations), you can improve the pick density and reduce travelling time.  It’s far easier to sort bits and bytes than to sort physical packages!

A typical mixed solution could involve the following:-

Fastest Movers


Top 10-20 lines pick from pallet live storage, with pick-by-light

Fast Movers


5%-10% lines in carton flow with pick-by-light, zone picking.  Possibly pick-to-conveyor.

Medium/Slow Movers


Carton flow and shelving, linear picking.  Multi-order picking to intelligent trolleys (mini batch pick and sort to trolley).

Very Slow Movers


Shelving, hand-hand terminals pick-to-trolley.

It’s not possible to do more than pose a few questions on this topic.  Order picking is at best an inexact science, relying on many other factors which determine the chosen solution.  However, to sum up, don’t ignore the basics!
Simon Dennis, Managing Director, Flowstore Systems Ltd.


The Author

As a Director of Flowstore Systems, Simon Dennis has over 20 years experience of order picking applications, covering both distribution and manufacturing activities.

Flowstore is a UK market leader in live storage picking techniques, which are often combined with pick-to-light, conveyors, shelving, racking and automated picking systems.

Recently he has been involved in e-commerce solutions for grocery home delivery operations, computer games and software suppliers, books and DVD distributors as well as clothing and home furnishing products.