Chilling out on advanced planning

Just a few short years ago, the food industry was deemed to be 'catching up' on factory automation, according mostly to instrumentation and control system developers' marketeers. It was a huge generalisation then, and more hype than substance. Although there has been considerable movement in some areas of some plants – notably on extruders, fryers, chillers, multi-head feeders, packing machines and the like – for many it remains more a hope than an objective observation. Most food companies I visit (as opposed to beverage) are labour-intensive and likely to remain so. Why? Partly because margins are so slim that capex paybacks are a problem, partly pressure on resources makes serious change difficult to feed through, and partly customers (mostly the big retailers) have such a grip on their suppliers they can demand flexibility that's difficult to fulfil with realistically-priced automation systems. It's another generalisation, but this is particularly the case with the more fashion-orientated and seasonal food producers, like those of snacks, confectionery and packaged meals, deli components, salads, prepared sweets and so forth. Interestingly though, although money is still tight, this is an industry that is alive to the power of packaged ERP systems for business and production management, so that in 2004 the picture is one of cautious but sustained investment. In the last year alone I've witnessed several SME food producers – the big boys have already done this – moving away from managing with spreadsheets, and securing growth, cost reduction and efficiency improvement through integrated systems. They've been doing so to manage their labour content, and scheduling around that, at least as much as their product inventories, machine and line sequencing, warehouse movements, purchasing, supply chains, sales order processing, demand management, financials and so on – and also capex planning. Now we come across SK Chilled Foods, based in Middlesbrough, which absolutely makes the point – but more importantly, is defining its IT very much in line with its business requirements, even to the extent of flouting the 'rules' around what comes first. Indeed, SK is currently rolling out an advanced planning and scheduling system (APS) before implementing ERP from a different vendor. Its Workplace APS is now working with the firm's departmental Excel spreadsheets primarily to plan for its people and shifts in production. Why APS? Mike Hepworth, SK's planning and logistics director: "We had to be able to schedule our factories better to improve efficiency, take out cost and underpin current growth. In terms of production efficiency I'm expecting 15–20% improvement from that: we'll get ROI certainly within 18 months. But I believe we'll also see longer-range savings. For example, we'll be able to make better capex decisions around stock holdings, building warehouses, buying machines... In a year that could save us around £500,000." SK has three factories, two in Middlesbrough and another in Skelton, turns over around £30m and employs 700. It produces 90% chilled and the rest frozen foods, and Hepworth describes its products as "mostly private label ethnic snacks". We're talking bhajis, samosas, wontons, spring rolls, potato wedges and burritos – some 90,000 cases a week, both pre-packaged and loose in boxes for retailers' deli counters. The plants are a mix of extremes: relatively highly automated lines side by side with very labour intensive manufacturing. Bhaji production, for example, is well automated, involving essentially a couple of people in preparation, while machines handle deposition of the filling, frying and chilling, followed by two more people for deli box packing and a couple for pallet handling. But samosas and spring rolls are entirely hand-filled and hand-folded after extruding the pastry leaves onto conveyor belts. And the packing lines are also highly manual: they have to be for the sheer numbers of ranges, mixes of product, preferred trays and the rest. "We're virtually contract producers," says Hepworth. "You've got to be very, very flexible in this kind of business. The supermarkets are constantly revisiting their ranges and you need to be able to respond to requirements very quickly." Prior to its APS, SK had an Opera system for financials only: everything else was on Excel spreadsheets. Orders would come into Opera mostly on EDI; from there demand would be entered manually into the planning spreadsheet, purchase orders would be raised manually, entered back into Opera – and so forth. Getting sophisticated Says Hepworth: "As we were growing, the spreadsheets were just getting too big, too many links… We couldn't do anything sophisticated, we had no visibility." And the focus of attention was on the big problem: planning and scheduling. "We needed a system sophisticated and flexible enough to take us into the future – something more advanced than MRPII, an optimising machine. But it also had to be compatible with as many systems as possible since we didn't know, at the time, which business system we would go with, or when." And hence the Workplace APS. "Labour is about 20% of our cost, and materials about 30%, but labour is catching up, and we needed to be more efficient." Currently, APS has replaced the main planning spreadsheet and is providing work-to lists for two of the factories, as well as feedback for purchasing. "We're not getting everything we can out of it yet," admits Hepworth. "We've still got to get changeover times, yields and run rates buttoned down." But he adds that there's been great value in implementing APS over and above the efficiency gains. "The journey is very important: you find yourself asking dumb questions, but then getting odd answers; then you dig further and find out how things are really done!" It's a good point worth factoring into your justification thinking. Beyond that, it's also worth noting that APS here works at two levels. On the one hand, it manages daily scheduling; on the other, it looks after the strategic picture 18 months out – planning capacity, capex, appropriate product mixes and so on. Returning to the former, EDI demand comes in and, although planners sanity check that, it's downloaded straight into the APS. At the moment, all sites remotely generate a daily snapshot of stock at the component level, and APS sucks in the data so it knows how many mixes, components, which varieties – and generates the work-to list. "We will soon use the APS to do our production sequence optimisation as well," he says. Subsequently, SK has selected Operations II ERP to replace and augment Opera with everything it hadn't got – so stock control, purchase order processing, MRP – thus replacing most of its spreadsheets, and integrating and automating purchase demand for the buying department, works orders for production and so on. Once Operations II is in, while scheduling will remain on the APS, its output will be improved and admin reduced. Factory stock positions will be generated automatically and integrated directly into APS so that it can plan against up-to-date data without human intervention. In fact, Operations II will then take over works order generation, driven by the APS engine, also exploding out from the BoMs into the ingredients and packaging for purchase order recommendations. From then on we're back to the standard reasons for ERP: visibility of inventory and the rest, shared real-time information. Says Hepworth: "We're in the 21st century. You need the visibility and sophistication to be sure you never run out of raw materials or packaging, you don't over-stock, you don't miss something… We're a growing company, and we want to make sure we use the right technology to sustain that – so ... IT handles the extra work."