Calculating the Risks and Benefits of Automation

LinkedIn +

The latest ‘Supply Chain Standard’ Roundtable, supported by automation specialists Swisslog, focused on making the business case for warehouse automation. The panel (see below) represented a range of retail industries with experience ranging from those who have implemented extensive automation, with varying results, to others who are still trying to make the business case for anything above the most elementary mechanisation.

For starters, Nick Allen asked David Armstrong, logistics director at HJ Heinz in the UK and Ireland, to explain the thought processes that informed Heinz’ move into automated warehousing, a project which came on-stream in 2003.

Armstrong emphasised the importance of lookingfirst at ‘how the business goes to market’. ‘Ten years ago Heinz was a fairly standard beans and ketchup operation. We grew through acquisition, brands like Farley’s and John West, and so we moved from having essentially a single factory to being a multiorigin business, drawing product from all over the country and indeed Europe.

‘We looked at our new density distribution, realised that we could do something different. If we could reduce what were now 14 satellite shipping points down to one, we could fill wagons. But we also had to make the justification to a lot of business units, each with there own concerns, not just about return on investment but about customer relationships. 80 per cent of our product goes to around six customers, and they have to remain happy and enthused that we are not just an efficient supplier, but one they can go forward with. We convinced the company that centralising and automating was the right thing to do and in fact it’s been massively successful’.

It is sometimes said that automation reduces flexibility – Armstrong suggests that often it is the people, rather than the systems, that aren’t flexible. ‘You can design the system to do pretty well anything from full truckloads to small packages if you have the money and can show the payback.

But high regular volumes show an easy payback on automation – for the “tail” you are generally better going into partnership with a company that has that as it’s core business, keeping your own “core” clean: or you have to build in a lot of flexibility’. (For Heinz, Wincanton has been a very suitable partner).

Phill Brooks, warehouse & distribution manager at Arquest (who are in nappies, pull-ups and similar products, mostly for retailer ‘own-brands’), but formerly with Exel, said ‘I know from experience that automation works well, but I need to convince my people that it works well’. Peer pressure would be significant, but Armstrong conceded that not many in the UK are building automated warehouses at the moment. ‘It’s dead expensive – 80 per cent of the costs are fixed so you have to sweat the assets in a stable business. It is a risk, if for example volumes drop, and you have to ask whether it would be better to be a shared user in someone else’s automated warehouse?’

But, pointed out Ian Nicholls, business development manager at Swisslog, there is a problem even in partnering with an logistics service provider (LSP) if the customer isn’t looking for a long term partnership. Contracts now are typically three years or less, and you often can’t show a payback in less than five.

Mark Barnett, COO at The Consortium (a public sector procurement and distribution organisation) agreed that automated solutions not only have to fit the current business strategy but also have to have flexibility for the future. ‘A business of our size has to be able to take advantage of new opportunities – we can’t have a supply chain strategy that depends on things staying the same’, but Guy Meisl, logistics head at Virgin Megastore, suggested an incremental approach. ‘You can reduce the scale of what you are looking at. Look at particular elements or operations – what is that costing me in labour? What elements aren’t productive? I’ll automate that bit. Just starting with a few conveyors might be just as valuable as automating the whole thing. There is less risk (because less cost), the payback is faster’. But, he emphasised, ‘that an incremental approach is fine provided it is part of the overall plan of where you want to get to ultimately’.

Meisl shared a, perhaps, common problem. ‘Our WMS doesn’t tell us a lot about productivity. It’s difficult to show potential savings, although logic says they must be there. To produce the numbers,we have to change the WMS. As it happens, that tends to make our procedures better anyway, so we can move forward’.

Each situation is unique
As Phil Greening, a consultant with Swisslog, noted, each situation is unique, but justification would be easier if you knew how to benchmark, and against whom. But there lies another risk. Armstrong claimed, ‘As a project in itself, [warehouse automation]is doomed. It has to be part of a bigger business change programme.

That may start back in the factories – they have to have manufacturing standards, palletisation, that will suit. How you deal with labels and coding – the way the factory codes, and uses the codes, in the production process doesn’t necessarily fit the warehouse need for, for example, automated layer picking. The automated warehouse needs other information so that, for example, we don’t put a layer of heavy baked bean cans on top of a lighter product. You have to develop a whole load of rules and parameters, right down to how strong the boxes are.

‘So this has to be a complete business change programme, bought into by the packaging buyer, the factories, the customers, your commercial guy who has to sell the benefits of a different way of working: full pallet deliveries, or creating more space in the customer’s back room. If the argument is just about the warehouse, it will fail.

‘And the case needs to be one that makes you more, not less, future-proof – for example, we may only be using part of the capability today, but we can sell to headquarters the idea that this means we will be able to continue to enhance our offer to customers for a good number of years’.

Greening agreed that ‘Automation forces supply chain integration. You have to talk to suppliers and customers. It can’t be a short-term strategy, you need to look far ahead’.

That implies that a significant warehouse automation project will need not just ‘sign-off’, but real understanding, across functions and right up the chain of command, which produces, as Armstrong said ‘a lot of daft questions – but also a lot of relevant questions; and you need to be able to answer them all’. Barnett agreed that ‘one of the biggest challenges is to get sign-off and understanding, that “they” can then percolate to their own aspects of the organisation’.

How do you eat an elephant?
But there is a dichotomy. Paul Petts, logistics director at apparel retailer Reiss, acknowledged the primacy of the business strategy but ‘How do you eat an elephant? Bite-size chunks. The only way I can start is with [local]productivity and labour’.

There was, it appeared, a clear difference in the room, only partly explained by scale of operation, as Barnett pointed out. ‘Some of you, for example Petts and Meisl, are trying to put a case together based on cost savings. Armstrong has been able to do the case based on opportunities to theorganisation, but it isn’t easy in many organisations to get a handle on those opportunities’.

Early engagement with other functions (and external stakeholders) is key. We need to understand all the drivers – labour, transport costs, service levels, and how these may change – would road use pricing change the strategy, for example? As Meisl insisted ‘You need to be able to see and work to the real strategic direction of the business. There are organisations where the supply chain goes off and does its own thing and then the Board turns round and says ‘great, but this doesn’t fit our strategy’. You want to take the concept to the Board before you have invested a massive amount of work in it’.

Nicholls pointed out that LSPs have similar problems – the customer presents them with a box (ie warehouse) and says ‘make it work’, but there may be little consideration for the impact on the rest of the supply chain.

A complex area – is there a case for consultants? Yes, but with caution, seemed to be the consensus. There are many cases of consultancies so tightly specifying that there is no scope for the system supplier to use their own experience, suggested Meisl; while Greening perceived a risk that consultants distance suppliers from the decisionmaking subtleties. At the same time, Armstrong pointed out, many LSPs now develop their businesses through their own consultancy strengths (as opposed to taking the client to the football!). ‘A good consultant’, he said, ‘can tell me how close the proposal is to what I want, and how much truth I am being told’.

But when do you need to hear this? Greening pointed out that ‘the perceived risk changes through the project. The closer you are to the end-game, the huger the level of risk. So everything needs to be upfront, out on the table, really early’.

There are more practical barriers to warehouse automation as well. Petts noted that his business is changing. He has suppliers in Turkey, for example, where ‘I’m sure they use the supplier manual to wedge the doors open’ – so how do you get the consistency that automation generally requires. And Petts has it easy – he is only serving his company’s own stores, so any automation he might install doesn’t have to be configured around varying customer requirements.

What also came out clearly is how very different market environments can throw up different arguments against investment in automation. For Petts, Reiss is a relatively small company with fairly decent margins (‘the owner is only interested in service’). High margins mean every single sale is really important, but there are new products almost every week, and an element of exclusivity (‘when they are gone, they’re gone’). And would imposing rigid rules on suppliers scare them off? How much automation can be viable on one million, vary variable, units a year?

Operational savings are key
Meisl at Virgin Megastore can, by contrast, ‘only dream of margins’. He is looking at up to 20 per cent price deflation year on year, so any operational saving is absolutely key. And although you would imagine that products such as CDs are pretty standard in terms of warehouse automation, Meisl explained that retailers like Virgin ‘have no control on how suppliers deliver to us. The box quantity can be changed without notice, which rather limits our abilities to program things’. And the level of automation required to deal with, for example, the Christmas peak, leaves massive overcapacity at other times of year.

That suggests possible collaboration between mutually compatible supply chains. Armstrong mentioned ‘Aunt Bessie’s’, the Hull-based maker of frozen Yorkshire puddings. 60 per cent of volume is in December, apparently, so they have a shared user agreement with an extensively automated warehouse. But Armstrong cautioned that you need partners with well-understood seasonality or volatility. (Breweries, for example, are not – weather changes, sporting success and so on can alter demand patterns without notice).

The factors for and against warehouse automation are, it became apparent, even more complex. There is labour availability, for example: at least one major supermarket is known to have justified heavy automation on a DC in North London purely because of the absence of labour in suitable quantity and quality.

The Round Table was convened primarily to discuss the justification process for warehouse automation. There is an increasing awareness that in this process, ‘reconnaissance is seldom wasted’, although several reported a continuing resistance to simulation exercises. In fairness, they may not be totally wrong. As David Armstrong said ‘Once you’ve commissioned the warehouse, then you have to learn how to run it’.

Meet the Panelists

[asset_ref id=”15″]Nick Allen – Editor Supply Chain Standard
‘What were the thought processes that lead Heinz to undertake an investment into warehouse automation?’



[asset_ref id=”16″]Guy Meisl – Head of Logistics Virgin Megastore
‘You can reduce the scale of what you are looking at. Look at paticular elements or operations – what is that costing me in labour?’



[asset_ref id=”17″]David Armstrong – Logistics Director UK & ireland HJ Heinz
‘This has to be a complet business change programme, brought into by packaging buyers, factories and customers’



[asset_ref id=”18″]Phill Brooks – Warehouse & Distribution Manager Arquest
‘I know from experience that automation works well, but I need to convince my people that it works well’



[asset_ref id=”19″]Mark Barnett – COO the Consortium
‘A business of our size has to be able to take advantage of new oppertunities – we can’t have a supply chain strategy that depends on things that staying the same’



[asset_ref id=”20″]Paul Petts – Logistics Director Reiss
‘How do you eat an elephant? Bite-size chunks. The only way I can start is with [local]productivity and labour’



[asset_ref id=”21″]Phil Greening – Consultant Swisslog
‘Each situation is unique but justification would be easier if you knew how to benchmark, and against whom.’



[asset_ref id=”22″]Ian Nichols – Business Development Manager Swisslog
‘[LSP] contracts now are typically three years or less, and you often can’t show a playback in less than five years’


Share this story: