WIP cut by30%, stocks by 70%

You can’t afford to carry on in business without this… We were gobsmacked.” So says Phil Hawkins, supply chain manager at Telford-based multi-million pound photocopier manufacturer Ricoh Products, of his advanced planning and scheduling (APS) system. His firm looked at systems from Frontstep (Symix at the time) through Ricoh’s ERP vendor Geac, STG (OPT, now owned by Manugistics) and Preactor before selecting Profax’s Asprova. And he says it’s been even better than they could possibly have anticipated. “It’s fantastic.” Ricoh turned to APS because, as Hawkins says, scheduling 22 photocopier consumables production lines with works-orderless repetitive scheduling and mostly four or five level BOMs (bills of materials) “was taking four or five planners two days every week” for weekly planning. Although the firm is still in the final throes of installation, first in the plastic moulding shop, as we go to press, he says that Asprova will cut this to one planner and just two hours – allowing the firm to move easily up to daily scheduling. And that, and the visibility it brings, is set to bring serious hard benefits. He expects WIP (work in progress) to be cut by 30% and finished goods to be slashed from “10 or 12 days stocks to about four. We’re talking about very big money: you can imagine.” Alright, that’s impressive. Now backtrack a minute. Advanced planning and scheduling (APS) used to mean just that: the technology to support doing production planning and scheduling better – mostly by getting away from the limitations of MRPII. And you’re into theory of constraints (TOC) and the range of advanced algorithms that consider multiple parameters (like materials, capacity and factory constraints), complex sequence dependencies and key performance criteria (like customer service, cost reduction and inventory minimisation) concurrently – and do it frequently – to get there and keep getting there. This is the territory that Ricoh is so successfully now exploring. But times, technology and terms have moved on. Today, APS and SCM (supply chain management) tend to be treated as blanket acronyms, by vendors and analysts alike, for much wider suites of applications – covering everything from supply chain design and optimisation, to forecasting, distribution and ultimately transportation – as well as manufacturing. Does it matter? Well yes it does. Partly because there’s the small issue of comparing apples with apples. Partly because at each level you should be considering different technologies. And partly because, with the advent of Internet technology, we need to recognise the implications for smarter, more effective collaborative working at every level. This is what’s significant: beyond the sheer sophistication and multi-parameter compute power (simulation and optimisation) of these systems, is the fact that real time information can be shared relatively easily among as many people, departments and applications as you like. And that’s a very big deal. Clearly, it’s high time to start thinking differently about planning and scheduling. Because the technology now readily available means we can, and the prizes of greater efficiency and competitiveness, and hugely reduced costs – through much slicker manufacturing and business methods – can mean significant and indeed rapid return on investment if you get it right. We’re all affected There’s a temptation to think that much of this won’t directly impact most mid manufacturers: it’s one thing to be a major corporate with production plants and suppliers around the world; it’s quite another to be one of those suppliers. But the fact is we need to be very aware of the scope of collaborative SCM in formulating our business and IT investment plans. Because as the bigger boys start rolling out web-based collaborative planning and scheduling initiatives (believing that this will make a big difference to their profitability), we in turn will have to be able to collaborate. So lets examine the power of the key APS/SCM levels (see panel, page 28). Simon Bragg, senior enterprise software consultant in the UK at analyst ARC, sees demand planning, distribution planning and production planning and scheduling – in that order – as most likely to deliver most benefit to most manufacturers. First distribution planning (safety stock levels, where they’re kept, the re-order points, etc). Says Bragg: “There is a tendency to keep product as close to customers as possible, but if it’s centralised, although transportation costs go up, variability of demand reduces, which improves production efficiency, reduces lead times and cuts costs.” It’s classic TOC thinking and some of this will be achievable without huge investment. Bragg notes that if manufacturers’ ERP systems have been implemented properly with links to warehouse systems, planning knows roughly where stock is in the distribution chain. “With integrated data you can then look at local forecasts, lead times, forecast and lead time variability – and plan better for defined customer service levels versus inventory costs.” Adding formal distribution planning software lets you take this to the next level with both simulation and full continuous optimisation. Then there’s demand planning, and what’s new here is ‘collaborative forecasting’. According to Bragg: “Thanks to IT, people are beginning to forecast at a European level instead of a country level.” He believes that whereas local sales departments have traditionally produced forecasts of sorts that eventually get to production and are promptly ignored in favour of expediters’ key customers and the sense of big batch manufacturing, “data is now being centralised. It’s ‘one number planning’ for everything – finance, sales, marketing, production, everyone.” Whereas consultancy Oliver Wight’s sales and operations planning (S&OP) methodologies have been all about changing company culture to do this – getting forecasting, production and the rest all singing from the same hymn sheet – this is about making the IT work for it. And it’s powerful stuff. David Tudor, senior principal for industry at enterprise and e-business software giant Oracle, says scheduling at the various levels in isolation is technically history. “We introduced our own APS with demand planning, long term production planning, short range scheduling and global ATP (available to promise) a couple of years ago. Manufacturers need to be able to see their finite capability, and optimise resources, materials and so on with the business rules, and take account of changing demand and supply. With our one database you can generate a holistic view and plan for all your time horizons.” Similarly, SAP’s APO (Advanced Planner and Optimiser) covers multiple levels – indeed the company mantra is collaborative SCM, everything from network planning, to supply chain execution and distribution. And with its web-based mySAP.com effectively making all of this available incrementally to the wider manufacturing market, on the face of it, it’s a profound opportunity. Whereas you would, for example, have plugged forecast and supplier stock data into MRP/MPS for long term planning and then exported the data into a separate tool for finite scheduling, now everything can be seamless. Tudor adds that the fact that you have one data set covering everything internally and externally means you can keep harnessing refined, sanitised information (data plus judgement). You don’t have to rely on internal departments’ or partners’ ambitious or pessimistic forecasts: the system can reflect whatever makes most sense at every level. “We see this as a key requirement,” he says. “You can go on for ever trying to get more accuracy into your demand schedule, but it’s far better to have something that’s 80% right all the time so that all other scheduling flows from it.” Production scheduling Which leads us on to production planning and scheduling – getting it right inside the four walls. They’re considered separately only where complexity dictates. As Bragg says: “BASF and any of the major process manufacturers would get excited about the difference, but a Midlands metal basher won’t.” For most of us this is the big one, and key APS software suppliers here include Profax, Inorda, Manugistics (with OPT), Mapics, Frontstep, Intentia, Lilly, Supply Chain Consultants (SCC), TXT e-Solutions and Preactor. These are useful systems. Andy Ferrar, Profax’s managing director, says his APS is can replace all the manufacturing planning modules of ERP – and he includes MPS/MRP, capacity planning and detailed scheduling. Citing the Institute of Operations Management (IOM), he says there’s a growing belief that ERP is “no longer the solution of choice”. ERP, he says, should only handle administrative transactions, “such as sales order processing, pricing, inventory control and so on.” And he adds: “The next five years will see a radical upgrading of the information systems used by manufacturers to embrace APS – perhaps as many as 40%.” ARC’s Bragg refers, somewhat paradoxically, to “the post-MRPII era”. Paradoxically because so few have done it, but he insists that more will move this way as understanding dawns and competitive pressures force firms beyond their legacy systems. “It’s a matter of getting away from planning in fixed cycles with fixed time buckets, batch sizes, lead times, sequences, queue times and set-up times, stuffing as much material as possible into the plant... Integration of shopfloor data, re-planning when things go wrong... Going beyond MRPII’s single feasible plan to choosing the best plan from the many feasible ones that best meets the chosen key criteria all the time.” And he’s right. No-one, other than the APS vendors, is saying don’t use ERP – many mainline ERP vendors have APS as an optional part of their systems anyway, or offer them through partnerships. But they are saying, if you need to deliver manufacturing plans faster, better, with time granularity down to minutes or even seconds and produce a useful optimised plan every time – dealing with the realities of variable lead times, process and routing choices, different priorities and so on concurrently – APS is your man. They’re also saying you’ll get greater visibility and more responsive and flexible manufacturing because plan regeneration can be as frequent as you like. And that means you can promise and deliver on time, implement and support real Just In Time (JIT) manufacturing and move to profitable make-to-order with the system helping you rather than getting in the way. You’ll also have a decision support tool – and the faster moving and more complex your environment (large numbers of items, orders, machines, set-up times, alternatives and people), the more useful that is. Ricoh’s experience certainly confirms this. Says Hawkins: “We have blow moulding plant, for example, so there are preferred sequences for colours and tools. Tools can weigh eight tonnes and we only have one big crane, so tool setters are one of our constraints. Before APS, if the schedule for optimal batches meant moving two tools at once, we couldn’t, so one of the machines would have to carry on making to stock or stand idle. We’d never thought about it because there was nothing we could do – we couldn’t get the information into the system. But Asprova knows this: the visibility for the planners is fantastic.” He continues: “And there’s manpower and holidays: previously we fully loaded our schedules because there was nothing else we could do – so they never got achieved.” He also refers to chemical mixes, yield factors and critical raw material storage tank levels, all of which couldn’t be adequately handled before. “It’s all within the capability of the APS. We didn’t go into APS with this much in mind. It seems like the sky’s the limit.” Hawkins’ enthusiasm is palpable. “We’ll roll it out to the whole production plant,” he says. “We also want to see it used to optimise warehouse space use: if you keep bashing out product you’re going to run out, and APS can highlight that.” And he has similar plans for managing WIP, stock parts and raw materials inventory better. “At the moment we don’t have visibility of these operations so we don’t know what’s there, but with APS the system will say ‘no, you’re out of space’.” So much is becoming possible, he says. Immediate plans are to use the APS to help move the business to daily planning. Ricoh’s distributed sales operations have already gone to weekly firm orders based on stock movements, but the plan is to take this to daily re-ordering. “So we need to be able to plan production on a daily basis,” he says. “We’re going to drive WIP and finished goods stocks down to the ground.” So why hasn’t this taken off throughout manufacturing? Bragg believes that although managements are becoming more aware, there remain several hurdles. One is suspicion following the perceived failure of the IT industry to deliver on users’ expectations. Robert Gordon, country manager for Italian APS vendor TXT e-Solutions (which launched over here in July this year selling best of breed APS against competitors like IBS, JD Edwards and Geac), agrees. He says, worth it or not, APS isn’t going to take the market by storm: “People talk about new ways of working – Internet collaboration and so on – but very few are actually doing it. The computer industry over the last 25 years has kept on bringing out new things, and people are suspicious, they’re sceptical.” Pathetic take-up? He also believes that given the economic situation “they’re retrenching”. Most important, he observes that for as long as people in this country continue to be rewarded for making small improvements to the status quo, it’s unlikely to change. “People are too frightened of making a bad decision,” he says. Ed Stubbs, pre-sales consultant with JD Edwards, says much the same: when it comes to sophisticated supply chain stuff, relatively little is happening. Even vendor managed inventory (VMI) isn’t exactly going like a train. “Take up is really pathetic, really poor,” he says. Nevertheless, most of the vendors claim to see improvement coming. “We’re starting to see more interest,” says Stubbs. “In the last six months since we launched the collaborative production and distribution planning tools, we’re starting to see take up in the middle ground.” And Gordon says: “We’re talking to several people now. There’s an automotive components manufacturer turning over £200 million, an industrial company with a single product line turning over £250 million and we’re talking to a very large food conglomerate.” Likewise, SAP’s solutions marketing manager Nigel Ford claims considerable successes in the US, with big names like the Hylfa steel company and Goodyear, and “a few (albeit unnamed) implementations in the UK”. He says there’s “significant interest” in supply chain applications, including in the mid-market. Indeed 20% of SAP’s UK APO projects originated from its VAR channel – and “the ‘pipeline’ for APO is looking very healthy at the moment.” Oracle’s Tudor says: “I liken it to EDI in the late ‘80s: the technology could support it but there was reluctance from the users, and early adopters used it but rekeyed information. Until their customers insisted. That’s what’s starting to happen now. Suspicion is disappearing very quickly: Ford is mandating that suppliers link electronically with Covisint, and it’s happening less publicly everywhere.” What remains is cultural change – the human issues, change management, training and the rest (which we cover so regularly). The often voiced concern is that for APS at this level to work, there’s a seriously big training programme to be undertaken, taking people on the factory floor and across the business away from the received wisdom of personal efficiency, individual machine efficiency and so on, towards collective efficiency. Add to that the prospect of dynamically rescheduled plans potentially changing production jobs on the fly, and the image of APS does little to inspire confidence in a workforce used to conventional, locally controlled manufacturing operations. Bragg notes: “People don’t like interrupting what they’re doing to enter data on machines, and if they’re bonused on how much they’re produced they won’t do it.” Ricoh certainly noticed the cuture issues. Says Hawkins: “You have to take a leap of faith. It’s so fast and so complex you can’t see your way through it, and that is difficult. You think ‘I can’t see it working, so it isn’t working’... But we’ve decided if nothing’s going to be late we’ll just go with it.” He adds, however, that for APS to work well, production control has to be fully in charge. International chain manufacturer Ewart Chain, which went live in January with £180,000 worth of extended ERP and APS software from Frontstep, also found culture a hurdle. According to Chris Heath, IT manager at the firm, under MRP jobs were typically released onto the factory floor “for the next two or three weeks,” which meant machine operators could choose the jobs they wanted. “But APS releases jobs as late as possible,” he says, “which is extremely difficult to get used to.” Costs/benefits But Heath insists it’s worth it: doing things the old way meant customer orders involving several components weren’t necessarily synchronised – so jobs could sit on the factory floor as WIP for days or even weeks. “Now we’ve got much better throughput on the factory floor; WIP is turning around better and it’s synchronised with customer orders, which is saving us a huge amount of money. And customer service is improved.” Further, he too alludes to the new visibility APS has brought to production planners, and to sales. “One of the features we like is CTP (capable to promise): sales people can just click a button and get delivery dates straight off the APS. They don’t have to keep on coming down and disturbing our planners to get dates. We reckon 50% of our planners’ time is now being better used.” Beyond this kind of anecdotal evidence it’s hard to come by good cost/benefit analysis. Bragg says it’s difficult. “We’ve been trying to talk to a host of companies and they’re not giving anything away. There are several reasons. One is that some pay by results so they’re not telling us in case they’re forced to pay more. Another has to do with organisations spending surplus budget before year end, and then no-one knows if they saved or made money!” Then again he says: “So few organisations have a clear idea of metrics before and after. One very large high tech firm I researched spent $1 million – 300 man weeks – on its new forecasting system across Europe – moving from individual country and departmental spreadsheets. They seemed to be doing everything right … they’ve got the right business processes [but] they think forecast accuracy was 15% before and it’s 15% now!” However, in the last nine months, with the economic turndown, forecasting has become much more difficult. Either way, he’s certain that at the very least companies with APS will now be able to take and deliver on orders where they wouldn’t before, because of their ATP and CTP facilities.