When you're zooming down the road in a Ferrari, the last thing you want to think about is dirt or debris entering the sports car's bearings or other vital engine parts. RFT S.p.A. does the worrying for you.
One of the world's largest manufacturers of oil, bearing and shock absorber seals, RFT S.p.A. is a division of Swedish-based SKF, the leading global provider of rolling bearings and seals with 80 manufacturing sites worldwide, sales in 70 countries and 15,000 worldwide distributors and dealers. SKF tests bearings, seals and other products under the harshest automotive conditions amidst the heated competition and intense action of NASCAR and Formula One racing. The current Ferrari model includes approximately 150 SKF bearings and RFT seals.
Automating the manufacturing environment RFT manufactures over 3000 individual seal designs, in over 25,000 variations of sizes and materials. As a second and third tier manufacturer to the automotive industry supply chain, RFT experiences highly volatile demand yet must maintain the high service levels that automotive customers demand. We have to be able to tell our customers whether or not we have the capacity to manufacture the products that they need, and it was becoming increasingly difficult to fulfil those orders, said Flavio Rapezzano, plant controller and project manager for RFT.
Our three main manufacturing steps are preparing, moulding, and finishing, but the moulding step in the middle of the process, is the real bottleneck in our capacity constraint, said Rapezzano. A further problem is that RFT has less than one-week visibility into orders, and the manufacturing throughput time is longer than the customer order horizon. To compound this difficulty, demand is extremely volatile, even within the fixed horizon. We have two types of customers, those who place orders for twice as many seals as they need, and those who order only half as many as they need. Moving forward, we know that we need to add intelligence to every order that we receive based upon history, Rapezzano said. We knew that we needed to forecast manufacturing from a plan, rather than from customer orders.
RFT was using manually intensive spreadsheets and time-consuming planning processes to schedule and respond to continual changes in demand, and to re-plan manufacturing activities.
Master planners at each facility typically spent three days per month planning production and inventory valued at EURO 5 billion (US $6.5 billion), and any variation in demand caused complex, lengthy recalculations. Given the demand volatility, and long re-planning times, RFT was forced to invest significantly in high levels of component inventory to meet customer demand.
RFT had several objectives when selecting a best-of-breed manufacturing planning system: reduce time-to-market, improve customer service, reduce overall inventory investment, enable plant schedulers to quickly respond to demand changes and integrate seamlessly with the company's MAPICS ERP system.
Moulding the future
RFT selected Logility Voyager Manufacturing Planning, and began the rollout at two plants in Northern ItalyVarese and Villanova d'Asti. At Varese, SKF began with master planning, and at Villanova d'Asti, shopfloor scheduling, the idea being that each plant would cross train the other plant on its specialty. Both locations have minimum quantity constraints, but different products and moulds.
Varese has a fixed order horizon of three months, but SKF plans to extend the timeframe as the company continues to use Voyager Manufacturing Planning. Our goal is to have an order horizon of a year or better in the Varese plant so that we may better satisfy future demand from our customers, said Rapezzano. Varese uses Voyager Manufacturing Planning to derive a plan to a precision of 20 minutes for nearly 3000 product/machine route options. A plan is created every 15 days and the result is passed to the MAPICS ERP system.
The Villanova d'Asti plant, where shop floor scheduling was first deployed, has an order horizon of 84 days. Moving forward, the goal is to restrict the order horizon to 56 days or less.
Released production orders are imported daily from MAPICS into Voyager Manufacturing Planning, which is then used to resolve constraints and create a feasible shop floor schedule.
At both plants, the results from Voyager Manufacturing Planning are used to derive an accurate procurement plan, determine the timely supply of semi-finished materials from the preparation stage prior to moulding, and determine accurate available-to-promise dates.
Business benefits
Voyager Manufacturing Planning has helped RFT shorten lead times and increase throughput by enabling us to create constrained plans for each plant, said Rapezzano. We can now take better care of our customers and meet their service level requirements, as well as create a manufacturing plan to quickly adapt to fluctuations in demand. RFT has already reduced inventory levels in one plant by 24 per cent, and we anticipate even greater reductions as more moulding machines are included across both plants. With only 25 per cent of the volume at the Villanova d'Asti plant alone, SKF has calculated that the 24 per cent savings currently equates to nearly EURO 150,000 (US $195,000).
In addition to the significant inventory reduction, RFT has reduced its master planning cycle by 87 per cent from three days to just three hours, of which the Voyager Manufacturing Planning portion is only ten minutes. RFT has also reduced the number of changeovers by five per cent, reduced setup time by seven per cent, increased machine utilization and has observed an immediate increase in customer service and greater accuracy in terms of plan versus actual performance. Voyager Manufacturing Planning is fully integrated with MAPICS, and there is automated synchronization between the two systems.
Moving forward, RFT plans to cross-train between the Varese and Villanova d'Asti plants for master planning and scheduling. The final master plan will be part of a unique and integrated system to manage customer order availability and reliability.
Add a Comment
No messages on this article yet