Nobody ever wanted to install an enterprise resource planning (ERP) system. What they wanted was to solve their manufacturing business problems. And for the last few years ERP has been the tool that’s promised to do that for them. This simple salient observation puts a healthy context on the following, which is aimed at helping you to select the IT system(s), service(s) and supplier(s) that will deliver best business benefits. So first, lets get a handle on what manufacturers think are the real problems. Last year’s Computers in Manufacturing (CIM) annual audit of UK industry’s IT investment by Bechmark Research, which showed a staggering 1999 figure of £3.366 billion, revealed some important clues. Beyond the usual concerns, like strength of the pound, IT managers told Benchmark’s researchers that their most intractable issues were quite specific: customers changing order details and/or delivery dates (20%); lack of accurate forecasting (18%); late and/or unreliable suppliers (10%); skills shortage (9%); and production machine breakdowns (8%). Here’s a list of others that will ring a bell. Poor visibility of production progress, making for schedule and accounting nightmares and impacting on customer service. Unreliability of staff adding to the supplier and machinery problems – and resulting in difficulty with revising schedules to respond to externally-imposed changes. Poor inventory accuracy – meaning that MRP is constantly undermined, with the two extremes of either too much expensive stock or none available – and machines and people idle, and customers waiting. Then there’s the elusive bottlenecks that seem to roam around so many factories, depending on the demand mix at the time, characterised by uncontrolled build-ups of work in progress in front of machines and again delayed orders. On top of all this, the pressure is on to cut costs and slash lead times. And then there’s e-business! It’s tough in manufacturing right now. Before we launch into the IT that can help here – and it certainly can – it’s worth pondering seriously the bigger picture. Ralph Seeley, senior consultant at industry watcher Cambashi, makes the very valid point that these problems ultimately comes down to “the lack of a production configuration that’s in line with demand in a statistical sense”. In other words, the symptoms they complain of reflect the fact that there isn’t enough of the right resource and/or that it isn’t managed adequately. Seeley: “That’s not just the number of presses, turning equipment, milling machines: but flexibly skilled people, flexible machine tools, multi-product production lines.” They also need business processes (the organisation and working practices) that are slick, efficient and working – properly glued together and supported by integrated IT systems that span finance, HR, sales, engineering, manufacturing (not just planners, but the shop floor too) and marketing – and deliver accurate information in the right format at the right time to the people that need it. Seeley opines that companies would do very well to start with some simulation and business modelling. It’s all about “finding the minimum investments to give the maximum flexibility to respond to your problems of demand, demand mix, staff and suppliers.” And he notes that the greater the willingness of a company to take on incompatible production, the greater the problems to resolve first. Certainly, there are plenty of suppliers in the following pages that can oblige with this first analysis, offering everything from graphical business and manufacturing modelling tools, to visual ‘digital factory’ simulators and full consultancy services. Why should you do it? Because you’re not going to solve these kinds of problems with IT alone. Trying to cut production cycle times from, say, six weeks to two takes rather more than implementing a business system – or even something more specific: an advanced planning and scheduling (APS) engine or a finite capacity scheduler (FCS). But that’s exactly what so many companies do! It may be the right technology, but employed alone the most you can do is tackle the symptoms. If you’re lucky you might get cycle times down to five and a half weeks! “IT is seen as a panacea – partly because it’s not understood,” says Seeley. “That’s why there’s so much ritual discontent. It’s a mixture of self delusion on the part of the purchasers who really want the problems to go away without too much change, money and effort – and over zealous IT vendors.” All this accepted, lets now attempt to map the perceived problems to the available IT. First then, consider customers changing order details and/or delivery dates. For this you need to be looking at product configurator software and on-line ‘availability to promise’ (ATP) – the IT that lets you see when you can reschedule delivery of the new order. And that means a fast planning tool, with modern MRP II as a minimum. Analyst ARC’s chief enterprise systems consultant Simon Bragg says: “ATP is usually a component of an advanced planning system, but it generally means you need fast MRP and stock data held on a separate server so you can provide new dates and schedules quickly.” IT here ranges from the constraints- and rules-based heuristics software of companies such as Scheduling Technologies, to the more sophisticated genetic algorithms-based software from firms like Synquest – which offer supply chain optimisation software that can constantly reschedule for any eventuality. Incidentally, there are also the intelligent caching systems from, for example, Acta, which can help beef up existing ERP systems (mostly SAP) to handle high query levels. Second up is lack of accurate forecasting, and here there are choices. Bragg says you could go for the “clever statistical packages from firms like Eventus Logistics and Mercia in the hope that history will repeat itself”. Or you could try the worthwhile, but ultimately limited path – develop better relations with your customers (and suppliers) with the help of datamart and CRM (customer relationship management) technology. Bragg also suggests a change in manufacturing approach, for example, to assemble to order (with to-order ERP to support that) so that forecasting needn’t be so long range. Like Seeley, he also notes that doing some simulation could improve the manufacturing and organisational basics. Others suggest taking a long hard look at your forecasting: doing it right involves getting senior production, marketing, sales and purchasing people together. “It’s what Oliver Wight calls sales and operations planning – treating forecasting intelligently and changing your processes to work more effectively,” says Bragg. Next, there’s unreliable suppliers, and the obvious remedies include paying more attention to multi-sourcing and improving on your procurement practices. But very soon now, you’ll be able to use the industry-specific Internet trading exchanges with their catalogues of approved suppliers’ products, reverse auctions and the rest, all over the Web. Going to companies like 3S, Ariba, CommerceOne and Ironside Technologies – as well as many of the existing leading ERP software firms – will be your route to getting into e-procurement. Beyond this, consultants suggest checking out why aren’t they reliable. Are you paying them on time? Do you communicate good forecasts? Are your lead time expectations reasonable? Beyond the obvious links back to other problems, there are ways to improve over phoning and faxing suppliers after your MRP re-run: like the new wave of ‘e-synchronisation’ and supply chain management software – business to business Web software and services – from Softlab, Extricity, Oracle, Synquest, Numetrix and others. Fourth on the list is skills shortages, and the consultants I speak to are of one voice: it’s hard in manufacturing right now, but thorough on-going training is essential. You might also want to ask yourself whether you’re offering large enough salaries and good enough working conditions – including modern graphical IT – to attract and keep the people you apparently need. On to machine breakdowns, and IT can do quite a lot to improve this – or at least to ameliorate its consequences. There are several packages providing for asset management, from relatively simple PC-based asset register and scheduling software to full CMMS (computerised maintenance management systems) integrated with machine monitoring and delivering predictive maintenance. The latter are particularly useful for critical production assets, allowing planners to schedule dynamically in the knowledge of impending shutdowns or reduced capacity. As Bragg says: “This is something that is within your control.” And he points also to the new Web portals providing for e-procurement, diagnostics, on-line documentation, ordering and the rest (sites like, run by PSDI). Users might also want to consider vendor-managed inventory services, some again available over the Web. Moving on to poor visibility of production on the shop floor, there’s no substitute for integration – of the business system with shop floor data collection (SFDC). Non-integrated or paper-based processes just don’t work. Seeley: “Without integrated systems you put huge demands on data entry – which isn’t going to be accurate or up to date. So it fails at the first hurdle.” Choices here range from linking in time and attendance systems to CMMS, workcentre HMIs (human-machine interfaces), automated counting systems and bar coding. It’s worth noting incidentally that many of the older MRP systems provided no pegging of production orders to customer orders. So there are fundamental limits with these in terms of providing, for example, customer updates that might well mean moving up to something newer and better. As for staff unreliability, although the real problem has more to do with motivation, company culture, environment and management, IT can provide some support. Providing slick, sensible, functional information to all helps. It can be as simple as email: whatever it is, it needs to be accessible. Now the old classic, poor inventory accuracy, which impacts MRP, and thus production, work in progress and profitability – all negatively. There are several methods for improvement, all involving cycle counting, but the emphasis here is on the sophistication of your warehouse management systems, and their links into the rest of ERP. Moving on, if bottlenecks are your problem, there are several avenues to follow. First is factory simulation – and we’re back to Seeley’s point about using IT to understand and change working practices, business processes, machinery, factory layouts, staff, etc. Second is the range of advanced planning engines: there are plenty of case studies showing huge bottleneck busting improvements once a constraints/rules-based approach is taken to scheduling. Finally, there’s the almost universal drive to cut costs and reduce lead times. And again, as Seeley’s says, a thorough manufacturing and business review, followed by a move to total integration, is essential. And this is it. Although it’s certainly useful to map IT solutions to manufacturers’ problems, ultimately most manufacturers are going to need the whole nine yards. Everyone needs an integrated system that includes at least MRP II (and most of ERP, although they might get away with something else for financials and HR), some sort of fast planning and scheduling engine, shop floor data collection and business intelligence. As Seeley says: “It’s a bit all or nothing.” Accepting that pockets are only so deep and resources not unlimited, which should you do first? Beyond analysis and simulation, Bragg maintains that the key is integration so ERP has to come first. “ERP is still the most cost effective way of getting a completely integrated manufacturing business system,” he insists. Generalisations notwithstanding, with the foundation ERP laid and bedded in, the next greatest value, he says, will come from APS or FCS. Then you should look at SFDC “because that tends to take you the last 20—30% as you tighten the accuracy and validate data”. Business intelligence and associated systems (datamarts etc) can come in with the ERP system depending on the scale of the business and its information analysis and reporting needs.
  • What are all these acronyms? In IT, the late ‘90s will be remembered as the era of the acronym. Was it all about an emerging plethora of different IT products? Or was it marketers desperate to up the ante by appearing to launch something new and better? Cynics beware, but if it was the latter it served the manufacturing market ill. Speaking after last year’s launch of the annual Computers in Manufacturing (CIM) survey, Benchmark Research managing director Guy Washer, for example, observed: “Many [users] don’t even know what APS [advanced planning and scheduling] and CRM [customer relationship management] mean.” Benchmark’s research showed that only 53% had even heard of APS, while just 15% were familiar with it. A mere 2% reckoned they had the software – and even then there was doubt as to definition: many simply meant finite capacity scheduling (FCS) add-ons to their corporate ERP (enterprise resource planning) system. See MCS, May 2000, page 20. As for CRM, it was a similar picture. Benchmark found 42% of IT managers had heard of it, 13% said they were familiar with it and 7% had implemented elements. Of those who claimed to be in the know, Benchmark’s survey suggested that sales force management was the most sought after element of CRM (56% of respondents). Data warehousing of customer data came second (at 50%); followed by telephone call centre systems (37%); using EIS (executive information systems) and OLAP (on-line analytical processing) for customer analyses (31%); and designing one-to-one marketing programmes (24%). Paul Watts, Benchmark’s research manager, put these findings down to UK manufacturing’s different understandings of the term CRM. Moving on to e-business, Benchmark found huge growth, albeit from small beginnings. But here’s another term embracing multiple meanings. Most users think only of marketing, buying and selling on the Web. In fact, e-business is just business harnessing the medium of the Internet – so that includes collaboration, supply chain management and all the rest moving up a gear. Benchmark’s analysis, for example, revealed that only 18% of IT managers in the UK had even heard of the term ‘e-synchronisation’ [simultaneously alerting all nodes in supply chains of the arrival of new orders now being scheduled for production – over the Web]. Less than 1% were actually using it! And truthfully, even this figure sounds high. How many UK manufacturers are set up to run their supply chain operations that tightly? The vast majority simply haven’t got the accuracy of data or forecasting to take advantage of the resulting inventory and lead time reductions. And what about much maligned ERP? Washer insists that manufacturers don’t for a moment believe that ERP, or at least the functionality we understand by the term, has gone as far as it can. Nor are they are rushing headlong into APS – whether that’s construed as an MRPII replacement, FCS, or full scale supply chain management and optimisation software – as some would suggest. “Most [manufacturers] are still focusing on the basics,” insists Washer. “It’s still all about getting MRP II and integrated ERP in place – getting the business and manufacturing IT foundation right.”
  • Manufacturing software: market trends Benchmark’s figures for 1999 revealed hardware investment at £1.284 billion (up 5.3% on 1998), total software at £1.038bn (growth of 3.5%) and services at £1.043bn (15.3%). Last year the firm forecast negative growth for hardware (-1.7%), falling to £1.263bn, while software would rise by 17.5% to £1.22bn and services ramp up 6.3% to £1.11bn. For enterprise resource planning (ERP) and associated applications like advanced planning and scheduling (APS), supply chain management (SCM) and finite capacity scheduling (software, hardware and services), Benchmark estimated user investment in the UK at £1.318 billion for 2000 (one third software), with growth at 6.3%. Specifically on ERP, predictions showed investment climbing by 5.7% to reach £1.392bn this year, although much of this would be in service. So far, however, managing director of Bechmark Guy Washer expects that estimate to be on the high side. Meanwhile, e-business in 1999 grew to £308 million – up 52% on the previous year. And Benchmark says it expects 33.5% growth to £411m this year. In 1999 it found 13% using Web-based purchasing tools (rising to 27% in 2000) and 10% had direct electronic links to suppliers’ ERP systems (26%). Beyond this, 8% were giving customers Web-based access to their own ERP (27% this year), 8% were selling direct to the public over the Web (rising to 15%) and 7% allowed customers to check order status over the Web (26% in 2000). Against these apparently glowing figures, however, Paul Watts, Benchmark’s research manager, comments that 31% of manufacturers state that they won’t be doing any of this more sophisticated stuff. For them, the Web remains a marketing tool. Incidentally, 23% state that their ERP systems couldn’t cope with Internet trading. Back to basics, and the CIM survey found applications like basic BoMs and stock control pretty well catered for across industry type and size: SMEs 79%, mid market 91% and corporate 97%. As for more integrated manufacturing management software, Benchmark found the figures down at 50%, 64% and 73% respectively – and companies with full MRP II were down at just 33%, 46% and 53%. Watts: “There’s no doubt from our research that packaged manufacturing management, especially in the small to mid company market, can expect the strongest growth.”