Wouldn't you just like to churn them out?

'Mass customisation’ – another of those dreadful market-speak phrases perhaps; poorly defined certainly; but for once meaningful insofar as it conveys the idea of using mass production processes but tuned, at the extreme, to make items each different and configured for an individual customer. That’s the apparent contradiction of today’s business drive for delivering better, faster customer service while simultaneously cutting costs and lead times in multi-product manufacturing operations. It’s all part of that rich tapestry of metaphorical hoops we have to jump through, and the operative word here is ‘configured’. Because for the vast majority of us, mass customisation translates (where it makes sense at all) to moving away from, on the one hand make-to-forecast/stock and pull through to order, and on the other make/engineer-to-order, towards much more manageable ‘configure to order’. Why bother? Ultimately, many of us have little choice if we are to survive and prosper in this competitive world, and we either do it willingly, or we find ourselves being driven. But there is another side to this: the business benefits – in terms of significantly reduced lead times, reduced operational costs (particularly on all the inventory and double handling measures), better efficiencies and utilisation, and improved all round agility and responsiveness – can be very compelling. At the very least, getting on top of this with a plan makes a whole lot of sense. Is it relevant to you? Almost certainly if you fit any of the descriptions above and are manufacturing reasonably complex products and/or a lot of variants. Current practice throughout the spread of industry sectors and company sizes of essentially pushing the problems of inventory cost and flexibility further down the supply chain simply can’t go much further. And a substantial spread of manufacturing SMEs throughout the industry sectors could realise some or all of the benefits – along with their suppliers. Is it easy to achieve? A cautious ‘no’ – although you can make significant strides without spending a fortune. As with most things, there’s a lot more to it than first meets the eye, and – depending on your industry, market, position in the ‘food chain’ and current IT and manufacturing infrastructure – there’s more than one way to do it. Whatever you do, you’re going to have to treat this as a serious set of linked prospective projects having short and longer term goals – with everything that implies in terms of board commitment, investment, training, education and change management. The good news is that since most of us have an ERP system of sorts, the fundamental integration, data and visualisation tools may well be in place to make the first steps at least feasible. The bad news is that the scale of change you need to consider includes everything from: factory floor layouts, processes, machines and tools; to staff skill sets; product, component, assembly, SKU (stock keeping unit) and supplier rationalisation; stock holding and replenishment practices; and potentially also decisions about subcontracting. Nightmare scenario Then there’s all the associated business processes (both within and between departments) to review: sales order processing (SOP) and the links into engineering, estimating, procurement and scheduling are key targets. Also, supply chain interaction will need smartening up to match the requirement for agility. And all of this while reducing, rather than increasing, costs! This is not trivial, but if it’s already sounding like the stuff of which nightmares are made, hold on. First, for those that have done it the rewards have been absolutely worth it, and second, thankfully, there are initiatives (lean manufacturing, Five S, Theory of Constraints (TOC)) as well as some relatively inexpensive IT tools that provide real structured assistance and ongoing flexible support. Top of the list here are, without doubt, modern rules-based product configurators (www.mcsolutions.co.uk for a list). With their focus on building product from ‘standards’, and their governance of both sales (quickly guiding the selection of validated, manufacturable configurations) and manufacturing (automatically creating item number, BoMs and routings data for ERP without the need for production engineer intervention), they can transform lead times, costs and customer service. What’s more, for some they even help with the up-front analytics for starting on product and component rationalisation, rethinking replenishment cycles and methods and so forth. One company that went this way is office dividers manufacturer Screen Solutions. This is an entirely make-to-order firm in which virtually nothing can be made, even ordered, from suppliers in advance. There are no standards – just 10 basic product ranges, and thousands of permutations, with variables in terms of dimensions, fabrics, colours and so on – 3,000 fabrics alone. Operations director Phil Saunders says: “We can’t even forecast much, so we just have to respond as fast and accurately as we can.” The key issue here is maximising what little lead time is available for procurement, manufacturing and distribution, and for him that meant automating the manual process of having production engineers assess and convert order specs for manufacturing. Screen chose a Geac Streamline NT-based ERP system but, importantly, with an Eden Origin product configurator package on top. After six months of development, on the sales side it’s hierarchical and does the job of narrowing down the choices directly to fully parameterised, manufacturable selections without the production engineering stage. Then, as per theory, from the sales order it automatically creates a temporary BoM and costing and passes item number, BoM and routing into ERP for materials procurement, production planning and the rest. “The time saving, the duplication of effort and the waste it’s removed are fantastic,” says Saunders. “It gives us a huge competitive advantage: 95% of orders are now being automatically configured.” Another that’s chosen to embark on a similar path is food and drink packaging film manufacturer FFP, which recently signed a £235,000 contract with ERP vendor XKO due for ‘go live’ in six months. It’s bespoke manufacturing again, with every film specific from design and printing, to laminating and slitting for the likes of Dairy Crest and Northern Foods. Financial director David Perkins says the configurator will be central: it too will generate temporary BoMs and routings straight from confirmed sales order, all on the fly. It will also hold templates for quoting against similar jobs with rolled up costs, and have links into capacity planning for lead time promising over the phone, creating back-to-back works orders automatically. “That’s vital to us,” he says. “I can’t imagine doing without this; it will be the engine that takes the hard work out of the whole process.” But there’s more to this: automatic valve actuator manufacturer Rotork, making millions of variants to order, is another ardent user. Configurators have been at the heart of its business since 1992, then based on a heavily bespoked MTMS ERP system, now Fourth Shift with Eden Origin’s system again. Engineering manager Ivan Burnell says that it not only does the guided order entry bit, BoM configuration and the rest, but also uses engineering rules to generate customer information in the form of installation drawings and wiring diagrams from his CAD system, and for the shop floor. “It’s the hub of the system,” he says. So they’re not new, but they are easier to implement and use, wider in scope, far cheaper – and they work. George Longworth, IT manager for voice recorders maker Thales Contact Solutions, says: “We used to use an Access database with someone keying output into the production system.” Now, the firm has an Access Commerce configurator integrated with its Minerva QAD Mfg/Pro ERP system and, via EDI, also with its outsourced manufacturer’s ERP. The system builds assisted sales orders and BoMs on the fly, transfers data as it’s entered into ERP, which creates back-to-back purchase orders for the subcontractor, which in turn responds automatically with sales orders, advance shipping notes and so on. There’s slick. The point is clear: even without SKU ‘standards’ rationalised on the 80/20 rule (with base products and subassemblies defined such that 80% are common, so moving to late configuration), a configurator can transform sales, manufacturing and supply chain management. It gets you away from the complexity and cost of engineer-to-order, almost to assemble-to-order, and for those with unpredictable repeat batches, stores and retrieves templates for re-ordering without filling your database with redundant BoMs. And don’t underestimate the complexity it removes on the sales side. If you have multiple sites, selling multiple lines with thousands of variants and engineering on top, there are often comprehension issues, errors and long cycle times involved with configuring, estimating and offering price and delivery. Hence the pressures, the duplication and the fact of customers playing off sales people against one another. As Soren Brogaard Jensen, CTO with $1.4 billion power protection system maker American Power Conversion, says: “Engineer-to-order is very time consuming and expensive. You have highly educated people sitting and putting all the pieces together. It increases cost and delivery cycle time. The turnaround time is much longer than if you have everything automated, and you know the variants and what kind of options you want to offer.” His company went with Cincom ERP to do the job, and transformed itself. How many could benefit from configurators alone? It’s hard to say, but Saunders claims his system cost around £50,000. “We’ll recover that within one year without any doubt,” he says. Indeed, he confirms that for his kind of operation there’s little point in ERP/MRP without it. Which is another important point: standard MRP/ERP is unlikely to be able to support the flexibility you need – look instead to the ‘lean’ systems, like those from Fourth Shift and SSA GT. So far so good, but as Adrian Edwards, partner in charge of supply chain processes with consultancy PwC, puts it: “There’s a lot of stages even to configuring an order and you need to move all that nearer to real time. You need on-line configuration, but you also need content management and on-line pricing, all linked into SCEM (supply chain event management) and ERP. You need a shared vision of resources, inventory, capacity. This is not a small undertaking.” Just so, and indeed the next obvious step for Screen would be to involve its key suppliers – with the system going the extra mile and placing demand as it comes direct on a web portal, with alerts etc, using an SCEM add-on. Not only would this further cut costs and time, but also provide exception-based visibility of problems throughout remote production, packing, despatch, goods in, inspection and so on. There are also possibilities for more dominant manufacturers with deeper pockets to consider latest generation rules-based ‘dynamic resolvers’ (simulators) either to opportunistically re-optimise their supply chains as supplier events dictate, or to continuously allocate production order-by-order, site-by-site against optimum cost and delivery. This is not an aside: early and continued involvement of suppliers is important to the success of a mass customisation strategy. It helps with spikes in demand; it helps with troughs. Giving suppliers early visibility of what’s required enables them too to tune their production and act more as part of your own factory – assuming they implement lean processes. Lean supply chain thinking can not only shave hundreds of millions of pounds worth of costs from operating expenses, inventory holding and obsolescence, but cut cycle time – along the lines taken in the automotive and food and beverage sectors. Indeed, particularly if you don’t have design authority, and are subject to weekly engineering change orders, this could be a first port of call. Sorting the factory Rotork’s Burnell makes the point that 10 years ago suppliers were mostly local or at least in Europe; now many are in the Far East, so there are issues around efficiency. Engineering change management is one: ensuring that suppliers are aware, and managing effectivity dates so the right components and assemblies come through without undue stock write off, for example. SCEM and portals, reports from MRP pushed to extranets – both provide workable options. Actually though it’s less the technology that’s the problem here: it’s the culture. Manufacturers are still wedded to holding suppliers at arm’s length: ‘collaboration’ remains a difficult word for many to spit out. My bet is that most will only change when their companies come under more pressure from customers in the usual cycle of price cutting and/or when middle management metrics start to encourage more radical thinking. Then there’s the factory floor. As Saunders says, with make- or engineer-to-order, the more complex and messy the manufacturing management the further away from ordered flow, and the closer to a jobbing shop, you become. The power of early, accurate data is in part the fact that it allows you to plan better and have more time to manufacture efficiently. But there’s more. FFP’s Perkins, for example, says he’s only expecting to see real gains from his system when he’s implemented Phase Two with shop floor data collection and touch screen terminals at the presses, laminators and slitters, and bar coding for all jobs to record material use, operation times and so on. Why? Because it closes the loop at the sharp end: you can’t expect value out of supply chain e-commerce, in terms of two way visibility of stock, planning and so on, if you don’t have the basics. And Rotork’s Burnell confirms the importance of the factory floor. His firm underwent a major investment in production engineering in 1998/99, moving from batch to flow manufacturing to improve efficiencies and flexibility, with internal and external Kanbans for the most frequently requested parts and assemblies, as well as JIT (Just in Time) and supply chain management where it could. It meant reviewing the configurator BoMs to direct specific operations to particular workstations via the works orders, as well as allowing staging for line balancing and the rest and finding ways to accommodate engineering change. But it contributed directly to further savings in terms of stock and reduced lead times, with today well over double production of more complex products for more global markets being handled by the same number of people. Plainly, if your production types and quantities will allow, you need to move towards more cell production with multi-skilling and a matrix layout to accommodate strategically located specialist CNC machines and the like – and schedule only the latter at the micro level, ensuring material release as late as possible. Rationalised SKUs (through database or configurator analysis), will also certainly help there. If that sounds like lean, it is – and if you can do it you’ll start achieving more with less, as accurate simplified orders get routed through flexible, efficient, uncluttered production with the economies that flow from that. Clearly there are limits. Kaizen, two bin systems and TOC, for example, underpinned with advanced planning and scheduling (APS) or at least finite capacity scheduling, can make substantial differences in addressing material release, replenishment cycles and bottleneck management, but some of us are too much in the hands of our customers. Nevertheless, there is a place for simulation systems and computer-aided process planning (CAPP) to make not just the fundamental layout of people and machines more responsive, but also the location of materials and component bins, tools, recording points and the rest. Other initiatives focus ‘lean’ more on the sales order processing side – and not just with product configurators. I’ve alluded to it already, but IT can help you cut lead time from SOP by, for example, linking sales, planning, engineering, purchasing and progress chasing into one department. It can provide the business process re-mapping and simulation to get there, and the workflow and configurator technology to support it. There are plenty of examples where harnessing ‘demand flow technology’ or similar IT with lean methodologies and variable Kanbans have helped slash both the order administration cycle time and the manufacturing cycle time literally by several weeks. And there are many that would argue we should all do all of this before we even consider a product configurator. But you be the judge: there are also instances where configurator IT has contributed handsomely, acting as a catalyst to improve lean factory achievements. No one is pretending SMEs aren’t already doing some of this: if they’re still trading many have had to make changes. But the issue now is to consider it in the round and that’s where some are going to struggle. How many can do it? As Mark Brewer, sales support manager for mid-market enterprise software vendor QAD notes, “Most manufacturers are limited by their infrastructure… It’s going to take them a step change – different tools, different processes. You can’t get that from ERP.” And for as long as there is management autonomy they are likely to be forced into adopting piecemeal approaches and point solutions. All true. But remembering that key points are to improve demand forecasting where you can, and obtain early visibility of detailed configured requirements where you can’t, you may make a good start. Stan Stewart, who heads up lean business management for ERP vendor SSA GT, puts it thus: “No one wants to see one-off purchase orders being raised for everything – that’s far too wasteful.” If even email can make a difference, think what a configurator could do for you. And get away from the silo/departmental mentality.