IT Strategy

Given today's global markets, post recessionary jitters and the relentless pressure on costs and margins, it's no surprise that many manufacturers are looking to sweat their assets harder than ever. Efficiency gains, streamlined business processes, radical improvements in business and market intelligence, enhanced competitiveness, and greater business and production responsiveness are among primary goals and all eyes are on information technology to deliver. All well and good, but where exactly is the investment? In straitened times, companies are understandably focusing on war chests and, as a result, boards of directors are increasingly demanding concrete proof of unreasonably fast ROI well before any cash is released. And that has several serious implications for those responsible for IT. How, for example, does an ERP upgrade stack up, in terms of ROI? Try explaining the risks and potential barriers of not doing it, when all around you are raising eyebrows in memory of the last self-imposed turmoil. Lindsay Watson, managing director of automotive styling and engineering specialist Ultramotive, believes the way forward is to demonstrate the value of real-time information access and analysis, interoperability between existing systems, monitoring performance against the plan, getting hold of problems before they become crises etc. But he also urges IT folk to do some analysis of their own "to make sure they're doing the right things and doing them well". Watson recalls working with one major automotive firm that had around 25 imminent projects on its wish list. "They ran an analysis of their projects' strategic value, compared against risk, and cut the list to 12 projects, with no reduction in strategic value," he advises. Clearly, such a scientific approach isn't always feasible, but the thinking is – and it's bound to demonstrate the positive value versus risk standing of, for example, spending on a centralised information repository, improving access to business intelligence, implementing CRM, improving the links between design and production, particularly in configure/engineer-to-order shops, and so on. That said, Sarah Cobb, group IT director at Moss Plastics, warns wannabe IT spenders that many companies may rightly want to focus on "crystallising the advantages they've achieved in the last couple of years". That means anything from acquiring failing but strategically useful businesses to picking up worthy IT projects that had hitherto been on hold. "So today it's about spending stuff wisely and investing in a slick project management and delivery capability, without throwing people at it." For her, the primary issue is bringing on good people with IT and business analysis skills. Nevertheless, from a functional perspective, Cobb nods to her own group's ongoing and significant investment in extended but highly standardised ERP, in her case from Infor. "That's the starting point and it's not just ERP, but what you put around it – such as e-commerce. For a lot of businesses, that pays back very quickly, in terms of economies and making it easier for customers to do business with you," she explains. And she also puts business intelligence and appropriate CRM in the 'must have' list. "We invested in BI and CRM some time ago, and I can't see how we could do without those in some shape or form," she says. However, Cobb also believes that MESs (manufacturing execution systems) should be on more companies' radar, because of their ability to make production more efficient, to cement in improvements and to make shopfloors more responsive to change. "When more capital becomes available, MES will come on, because most of us need to take the next steps in production," she explains. For the next two or three years, though – given Moss Plastics' strength in ERP – Cobb says her focus will be on "better web, better BI, improving efficiency and extending all of that into new territories". Intruder alarms manufacturer Texecom's operations director, Russell Trotter, has an only slightly different take on her thinking. He also comes at IT investment very much from the perspective of adding value and/or stripping out waste – meaning that upgrading ERP simply to keep up to date is unlikely to be justification enough. However, for him, best advice is to identify the non value-add physical activities and process flows that amount to lean's classic waste. "Then what matters is how IT can make sure product velocity, for example, stays high," he suggests. For an electronics industry manufacturer, that thinking suggests electronic kanbans, with automatic management of materials and works orders to production cells, directly in response to sales orders – all supported by barcode material scanning and lean supermarkets that minimise human intervention. Extending the approach also points to web portals for direct customer order entry, which, in turn, cuts out internal clerical processes. But others beyond the electronics industry might profit from this kind of IT, too. It's going to appeal to anyone thinking of the joys of schedule-less factory operations, backed only by rolling monthly forecasts on major suppliers and minimum stockholding triggers – which signals a natural limit. More generally, though, part of making this concept work efficiently (and waste-free) is also implementing web EDI to suppliers, with automatic updates into your ERP system that, in turn, reflect the logical and physical movements of materials into assembled products. Either way, Trotter believes many more manufacturers could make an easy case for investing in another of Texecom's big wins: getting CRM and ERP data available instantly on managers' and customer-facing employees' Blackberries. "We made information such as order status, stocks, purchase authorisation requests etc available, and we're now deploying that into all departments where we think faster information flow will make a difference," explains Trotter. It's all about stopping roving managers from remaining bottlenecks, and he indicates that it needn't be expensive or difficult. "We created some scripted ASPs [active server pages] that link into our Epicor Vantage [ERP] server via SQL tables, and generate XML files. Obviously, we already had the Blackberry Enterprise server, so it was only a matter of linking through Service Connect into ERP to provide the remote connectivity. We're now talking about also giving the board access to iPads for their greater screen size, but the ethos remains the same." That's a good, practical and relatively simple suggestion. Slightly more complex, but even more valuable for manufacturers that develop their own or others' products, is improving the links and processes between ERP and PLM (product lifecycle management) systems. Nathan Bailey, currently operations manager at UAV Engines (which manufactures, er, unmanned aerial vehicle engines) makes the very valid point that while most companies will have automated data transfer between some systems, they struggle when it comes to connecting development/production engineering and their ERP/shopfloor systems. "They seem to over-complicate it, with a lot of manual steps. Rather than asking, 'What do we need to do to move from development to production with nice singular flow?' they break the process into separate activities," observes Bailey. He suggests getting away from the 'this bit of the process belongs to me' approach and instead putting some effort into understanding that departments and functions are "all just nodes on a flow chart" – certainly before they invest in any further IT. And if the cornerstone of a modern IT strategy is reducing waste, eliminating paperwork and duplication, and easing information flows to make processes more efficient, you can't argue with that. For him, breaking through the impasse from a systems point of view need not involve anything more sophisticated than joined-up thinking. "For example, instead of emailing whoever to say 'this drawing is ready for approval or review', and making the recipient hunt it down, the system needs to show that drawing – and then the recipient should answer on the same task," he says. How do you do that? Well, you probably need to start by reducing the number of system interfaces and improving on integration. From an IT investment perspective, that's not necessarily such a bitter pill to swallow. There are standard platforms providing for transformation scripts, so it's no longer about starting with a blank sheet of paper and a lot of consultancy. Then you're into drawing logical flow diagrams to move data around and maintaining the control of that within the company. However, at the same time, you'll need to look at streamlining the information flows, while recognising that different formats and types of data (engineering versus production, versus marketing) may necessarily go through different flow paths. "You may have workflows built in – in our case, they're defined in our Efacs ERP system – so you want to make sure that, when they're triggered off, the system can retrieve the information and present it appropriately to the decision maker," explains Bailey. "You also need a standard receptor mechanism – say a right click to approve or not, or to comment, and then the system sends it back," he adds. And it's much the same for the movement of information between design and production. "Your workflow needs to pick up approved designs and then, when they're marked as ready, automatically transfer them into ERP for the BoM explosion, purchasing and manufacture," says Bailey. So much for the functional and process-centric IT investments; Bailey also suggests that a rethink of your network infrastructure might be in order. "We have recently invested time and money into virtualising our software environment, because the systems available now are really good," he advises. "A lot of what you might think is hype is not any more. We looked at virtualising three years ago and, at that time, it wasn't mature enough in the live environment – only for software developers or help desks. Now it's a different story." Bailey makes the point that the more hardware there is on a network – and most of us bought shiny new servers to run each of our shiny new applications, to avoid the problem of common cause failure – the more administration and opportunities for failure you have. "By virtualising your software and providing the capacity for failover between systems you can minimise your admin and system overheads while still maintaining independent application servers," he explains. Was it difficult? "Not at all; we used an off-the-shelf Citrix XenServer [virtualisation engine], got a couple of pretty large servers, created the virtualised applications and transferred them. We went from 24 physical servers down to just three on an internal 'mini-cloud'." His only caveat: "You need to do it gradually and make sure you can handle a mixed platform for a few months. That way, if the virtualised machine doesn't work, for whatever reason, you can turn the original hardware on and you're back where you were." Others suggest going for a big bang approach, with the belt and braces of an additional server for failover. "Either way, you need to do the capacity analysis," comments Bailey. "Some applications, such as ERP and PLM, are used heavily, whereas others, such as time and attendance, are low-load or used infrequently or are low load, and others again run across multiple databases, such as business intelligence, which just fires up when it's invoked. So, do the analysis of CPU usage, hard drive space, memory usage and networks usage. Some of our servers were only using 5—6% of capacity and hence the huge saving… The only other thing is that you want common storage, so eventually you'll need a SAN [storage area network]." Panel 1 What the IT vendors say For a full analysis of the opportunities as the IT vendor community sees them turn to the following pages. However, here is a taster. According to Epicor's senior vice president of world wide product marketing, James Norwood, while the last two years have been about "cost cutting and customer responsiveness", now the focus is on "more efficient business process execution and new product introduction". Wishful thinking, geared to spurring IT spending? Norwood says not, insisting that IT vendors have recognised the need for lower costs and rapid ROI, and citing Epicor's introduction of yet more software packages preconfigured for specific industry requirements as proof. He also advises IT directors and managers to reassess cloud computing, arguing that, if the requirement is to rapidly switch on computing resource, it's all there, just waiting to be tapped into, without any of the usual provisioning hassles. Beyond those, Norwood suggests that many manufacturers' e-commerce storefronts are going to need a refresh and that 'Enterprise 2.0' is not vapourware, but very real and ready to facilitate collaboration anywhere. Infor's director of solutions marketing, Andrew Kinder, takes a more function-led approach, suggesting first that it's time to revisit S&OP (sales and operations planning). "Any practitioner will tell you that S&OP is a process, but the technology is now ready to provide intelligent insight into demand patterns, for example, so that manufacturers can predict requirements more accurately," he explains. And he says that, whereas S&OP used to be the purview of large organisations, SMEs can now step up to the plate. Beyond that, Kinder urges manufacturers to consider tightening up their operations at both the supply chain and factory levels – the former with event management and alerting technologies, and the latter with APS (advanced planning and scheduling). "If your strategy is about squeezing more capacity out of what you've already got, then this is the way to do it," he asserts.