Warehouse Automation


For some, warehouse automation is bleeding edge technology, to be distrusted, too expensive and too inflexible.  For others, it is an essential part of their fulfilment strategy and the only way to control costs in a market demanding ever higher levels of customer service.

In reality both views have some validity (apart from the one about it being ‘new’ – warehouse automation has been around for more than fifty years).  So how can that be the case and how do you make sure your experience of warehouse automation is a positive one?  In this article we will be exploring a number of key areas: –

  • What can go wrong?
  • How is the world of automation changing?
  • How to ensure success?

What can go wrong?

Clearly the answer is lots of things can go wrong and industry is littered with systems that have not delivered what was expected of them.  As with most things in life though we often only get to hear about the problem systems, never the success stories.

Warehouse automation is a long term investment. You have to do your homework and build as full an understanding of what the future is likely to bring as you can.  But of course the future sometimes doesn’t turn out quite as expected and so you also need to understand how and where flexibility is needed. Failure to do this might result in a solution optimised for the past rather than the future.

If you don’t have automation expertise in your business it can be tempting to let the suppliers do the work for you and come up with the solutions.  But suppliers don’t know your business as well as you do and the result will be their solution not yours. If you don’t want to recruit that expertise (and it can be hard to find) you can use a consultant experienced in this type of work.

When undertaking a project as important as this you want to make sure that your key team members have done it before and have already made the mistakes that they need to learn from.

There is an old Russian proverb that translates as “trust but verify”.  In the context of building an automated warehouse this means it’s fine to trust your supplier but does not mean you shouldn’t fully test the system before you take it over.

Systems that are fully tested work better and achieve their full potential much earlier than ones that are not.  Particularly important is a full volume test.  Suppliers will often tell you it is not possible to do volume tests for more than a few minutes but with some ingenuity and thought it almost always is and always pays dividends.

How is the world of automation changing?

  • Offsetting increased fulfilment costs in e-commerce;
  • Much less about bulk storage density, more about pick and pack efficiency;
  • More complex control strategies to optimise performance.

In the early days of warehouses automation the focus of activity was on high bay pallet stores.  Companies were using automation to reduce the building footprint by building high (in some cases up to 50 metres) and to reduce the labour cost of moving stock in and out.

Two things then happened – companies started to get better control of their supply chains and found ways to offer greater customer service with much lower stock levels as well as realising that the biggest cost component is often order picking.  These two factors are related but unfortunately not always in a complimentary way.  A consequence of each link in a supply chain reducing its stock holding is that each link will have to serve its customers with more frequent, smaller orders, i.e. they will have to do more picking.

And the ultimate example of this is in the form of home shopping where stock that was once shipped in pallet or case quantities to stores is now having to be picked for individual end customers.  This has resulted in a dramatic increase in fulfilment costs which retailers are now having to address by investing in automation.

Finally, and probably the biggest change we have seen in warehouse automation has been in the optimisation and control functionality.  It’s systems are now much more feature rich, suppliers have far more experience of how they can optimise their systems and many more functions are parameter driven rather than hard coded.  It is vital to spend time discovering and learning about these functions.

How to ensure success?

The starting point is usually a set of historical product, stock and order data for up to 12 months of trading.  This almost always needs some cleaning, for example, to remove transactions that did not result in physical stock movements.  The data then needs to be projected forward based on growth and other business forecasts.

This however can be a dangerous phase in many projects.  Once numbers appear on screen or in reports people start to believe them, especially suppliers who understandably want a fixed, clear base on which to design a system.  And yet the only thing we can be certain about with this projection is that it could be wrong.  Warehouse designers are no more able to predict the future than fortune tellers and in both cases it can be fortunes that are at stake!  So solutions have to be flexible and the data should help describe where the flexibility is most needed.

With a base of data to work from solutions can be developed.  In doing so it is tempting to automate everything, but different parts of an automated system will provide more benefit than others. So it is vital to carry out incremental cost benefit analysis to make sure each part is paying its way and if it isn’t, use a more conventional solution.  There is nothing wrong with mixing manual and automated processes if that is what makes most sense.

Equally important is designing in flexibility from the start.  Try to come up with a range of future scenarios for the business (different range and sales growth rates, perhaps the possibility of taking on new ranges or brands, new services, etc.) and test the solutions on paper to make sure they will still work should the forecasts turn out to be wrong as they almost certainly will.

Make sure you own the solution.  If you don’t you will not understand it well enough to get the best out of it.  At The Logistics Business we talk about projects having an implementation phase and then an exploitation phase.  Both are equally important but because the exploitation phase usually takes place after the supplier has left the site you must take ownership of the process and embed it in your normal operating procedures.

Testing is an art in itself.  You can never do too much and although the law of diminishing returns applies, few ever come close to that being a problem.  The biggest danger with regard to testing is that no matter how sensible the original plan these installations often suffer delays and so with penalties for late delivery looming there is almost always pressure to shorten the testing programme.  This is almost always a mistake.

The best way to avoid or minimise the risk is to make an early start on the development of test plans and scripts.  Ideally this should start as soon as the functional specification is complete; whilst the functionality is still fresh in everyone’s minds.  Test every detail and above all make sure you fully test the performance before handover.  And not just for ten or fifteen minutes but for several hours.

If done well warehouse automation can be key to enabling you deliver great service at low cost.  Done badly it can be a millstone around the corporate neck.  If you are not confident about having the right skills to ensure success please give The Logistics Business a call.  We have the experience and would be delighted to help.