Failing to plan?

Supply chain executives face an unprecedented challenge: to balance the need for cost efficiencies with the growing clamour for agility. So intones analyst Gartner, suggesting that for most in industry this requires "a fundamental shift in strategy, planning and implementation", as well as "a new ability to provide guidance and leadership to every member of their global supply chain teams". Grand stuff and, on the face of it, not a little intimidating, given the point that manufacturers are already under considerable pressure to do more with less. In fact, however, Gartner's challenge is not all that unprecedented: manufacturers have been coming to terms with cutting costs while simultaneously improving responsiveness and customer service – largely using measures from OEE to OTIF as drivers – for decades. And, arguably, their problems have been simplified in recent years, given the increasing sophistication, scope and ease of use of computer assistance – particularly when it comes to planning and scheduling software. Much the same is also true of associated systems: for example, order management, warehouse management, demand planning, B2B/EDI and tracking software. In the right hands, these systems bring even the most difficult supply chain problems to heel and deliver competitive advantage to boot. Take high-performance crankshaft engineering and manufacturing specialist Pankl Racing Systems. On the one hand, the firm lives and dies by deadlines imposed upon it by immoveable motorsport race dates. But, on the other, it also has to respond to relatively unpredictable changes on what are invariably bespoke projects. It's not just about getting theright products delivered on time, every time: if ever there was a requirement for accurate, yet ultra-flexible production planning and scheduling capable of driving an entire business, and ultimately its supply chain, this is surely it. Looking at the detail of Pankl's US-based operations, order sizes range from two to 12 units and can be either manufactured or designed, developed and manufactured to customer specifications. Also, depending on the complexity of the product and the degree of engineering, lead times vary between 10 and 15 weeks. Production itself involves in excess of 50 process steps, some undertaken in Pankl's own works, others sub-contracted. And, given the exotic nature of its raw materials, these are often supplied, yet Pankl still needs to retain working buffer stocks to guarantee those all-important delivery dates. But that's not all. Like most manufacturers, Pankl has to manage multiple concurrent projects running different – sometimes conflicting – routes, with shared resources and finite capacity. So production balancing is an ongoing issue. That, in turn, means that managers and supervisors need good visibility across production, right down to job status and location, as well as individual resource capacity constraints, and set-up and changeover times. Further, from a day-to-day operational perspective, Pankl runs with a production planning horizon of 16 weeks, yet also needs to react quickly to changes within that window – whether due to planned or unplanned maintenance or customer updates. So it's also about being able to re-prioritise work while revealing the impact on all orders up to and beyond the planning horizon. And if a change happens in the design phase, rescheduling also needs to include other projects in engineering as well as the knock-on effects in production. Yet until last year Pankl – along with much of UK manufacturing – relied on spreadsheets for its planning and scheduling. A Pankl spokesperson describes working with Microsoft Excel as a matter of "manually moving coloured boxes around", adding that it used to tie up significant skilled human resource, even if just one change was required. He also says that it simply wasn't possible to create 'what if' scenarios. So in 2011, shortly after the acquisition of former Superior Crankshaft by Pankl, the company started searching the web for a scheduling system that wasn't too difficult to use, yet had the power to handle its environment. That led it to Production Modelling and its Orchestrate simulation system. And, after product demonstrations and a free download for 14 days, with training supplied and further online training available, the company bought the software. Eureka: according to our Pankl man, what the firm had thought might be daunting was anything but. The system proved easy to pick up and indeed the local team largely taught itself. Less than two months later, planners had forgotten spreadsheets, and were managing plant optimisation and change management entirely on Orchestrate APS. The result? Although Pankl's spokesperson cautions that it's still early days, the benefits, he says, have been considerable, and not only in terms of easily identifying optimal job sequence, no matter what the change demanded. He cites a newfound ability to visualise the bigger picture and, as a result, to see any potential for improvements. Additionally, scenario planning that was hitherto impossible is now provided in real time, meaning that decision making is far faster and more reliable. Indeed, Pankl has subsequently increased its planning horizon by a further 25%, and our spokesperson states that, in terms of time alone, this software is saving at least 43 days per year of expensive admin. And it's not just about planning. For the first time, engineering and production have been linked into a coherent plan with do-able deadlines – meaning that when designers are given a completion date, everyone knows it's feasible and production can plan to start on time. Just as important, the impact of changes in one department can be seen in the other. Interestingly, that has led to better decision making throughout, despite the inevitably conflicting agendas between sales and production. It appears that, with representatives from each department seeing the same 'what if' graphics, consensus decisions that are best for the customer and the business have now become the rule, not the exception. Indeed, the system is also being used to determine the benefits of capital investments by examining faithful simulations of different sales and capacity scenarios. As for the future, Pankl says it is only scratching the surface. Areas next slated for improvement using Orchestrate include harnessing its reports to improve forecasting for suppliers, while also smoothing workflow and helping to bear down further on bottlenecks by using its WIP analysis. "We've got loads of things to try but we're taking it one step at a time and getting each step right," comments the spokesperson. "We have a saying and we are proving it at each stage: do what Orchestrate says and everything will be good." But look at virtually any APS implementation, almost irrespective of the industry involved, and you'll hear similar stories. On the other side of the production spectrum, for example, confectionery group Mars is now reporting success with an installation of Preactor's GMPS (graphical master production scheduler) and APS software. This one is at its Basingstoke factory, which is responsible for producing Klix and Flavia drinks as part of another complex, inter-dependent, but very different supply chain. We're talking about a plant that manufactures for Mars business units in the US, Europe and Japan – so very large volumes but also a wide range of variants, in terms of formulations and permutations, pallet sizes and quantities. Also, there are the issues of fixed machine numbers running continuously over three shifts, meaning a challenge for sequencing to maximise capacity utilisation. And, just to add spice, product groupings and allergen controls that need to be factored in to minimise changeover and clean-down times while maintaining smooth supply. Oh, and certain products have specific storage requirements. Additionally, as supply logistics manager Paul Hazelwood explains, different demand mechanism pertain in different markets. "When dealing with the UK and Europe we receive a forecast breakdown to an individual product level in a given period, typically a rolling 20 weeks. For the US and Japan, however, we make to order, again over 20 weeks but with a four-week locked window." Also, whereas orders for the UK, US and Japan are scheduled weekly, Europe works on a less regular basis. Hazelwood explains that Mars had been using a system called Schedule X. However, the combination of Mars IS withdrawing system support and Mars itself wanting to improve its lean credentials, meant a requirement for more accurate data and reduced stock levels – and, in turn, a better scheduling solution. Hence GMPS and Preactor APS, which went live in February of this year in record time, which Hazelwood puts down to intuitive software, "a lot of hands-on training and a large amount of scenario testing". This project, too, is still in its relative infancy, but Hazelwood says there have already been marked improvements. He explains that all demand forecasts and orders are now received and imported into Preactor GMPS, along with sales and stock data, which generates the rolling 20-week, high-level schedule. "GMPS looks at high level scheduling at a weekly level and ensures that we remain within target levels across all our SKUs, taking into consideration our high-level constraints. From here, we export the GMPS data into Preactor APS, which then handles detailed scheduling down to what needs to be done, where and when, taking into account our detailed constraints. Once the schedule has been exported to Preactor, we generate an MRP report, which we use to order materials." Beautifully integrated. However, the most immediate hard benefit has been stock levels reduced across the entire pipeline. Hazelwood believes it's still too early to quantify the full impact, but cites one SKU where stock has now been cut from 10,000 pallets to just 4,000. Also, because Preactor GMPS and APS are interlinked, the impacts of every decision can be seen on everything else, so the company is now working hand in glove at a global level. "Anyone involved has visibility of the entire pipeline and this increases confidence and trust at every level," he asserts. BE Aerospace transforms multi-site production on APS BE Aerospace's super first class division, which manufactures interior products for the commercial and military aircraft markets, is reporting increased visibility and accuracy of information since implementing Orchestrate planning and scheduling software, from Production Modelling. "Now we can see where everything is across the entire business unit and at which stage," states a spokesperson, explaining that the system was implemented to manage operations across three sites and two countries. And he adds that improvements have also been enabled at the firm's Tucson, Arizona site by the installation of flatscreen displays on the assembly floor, which allow everyone to see not just what is supposed to be happening, but also what's coming down the line. As for the scale of time and cost savings, he says that for one schedule change alone, Orchestrate is one to three hours quicker than updating Excel files. "Also Production Modelling developed a presentation plug-in that turns our master production schedule into a presentation, which refreshes and runs continuously on the flatscreen monitors on the assembly floor. This keeps our production schedule current and encourages employee buy-in, as they see the demands of their individual production lines," adds the spokesperson. That matters, because of the complexity and inter-dependence of BE Aerospace's manufacturing plants – with seating systems for Emirates, Qantas and British Airways, for example, requiring components from all three production facilities. "While final assembly takes place in Tucson, this is dependent on operations being completed on time at our Miami, Florida and Nogales, Mexico sites," he explains. Comparing the past to the present, he also states that three of the biggest challenges used to be in managing workflow, capacity and new business. "The last three years have seen us grow from a 300-employee operation to now having over 500 in our business unit across multiple sites. We had a separate spreadsheet for Miami, Nogales and Tucson, all of which had to be updated manually, and all of which were on top of the data stored in our ERP system. We had multiple ways of communication to get the same data, but often different spreadsheets had different data, because nothing was consolidated. "Now, with Orchestrate, all data across each site is consolidated into one source from which clear, concise and accurate information can be quickly extracted. This has positively impacted our on-time delivery capabilities... Not only are we now able to schedule more efficiently across all sites, but we can also take on new business with increased confidence – which is essential to our ongoing growth."