Smart Software’s (SmartForecast®) forecasting solutions increase parts availability and reduce inventory costs
Prevost Car, Inc. is a leading Canadian manufacturer of intercity buses and coach shells for high-end motor home and specialty conversion. Prevost Parts, a division of Prevost Car, serves the North American auto aftermarket distributing original coach and urban bus parts for Prevost Car and Nova Bus, as well as replacement parts for other models. To serve its market, the company maintains two distribution centers, one in Canada and one in the US; five service centers; and over 25,000 active parts. 70 percent of those parts have an intermittent, irregular demand profile, which makes them very hard to forecast and manage.
At the beginning, Prevost Parts’ management started a comprehensive program to improve the company’s parts distribution system. Prevost needed inventories at its warehouses to more accurately reflect demand at those locations, and required more accurate forecasting inputs to its production planning processes. 25 percent of its orders were not being shipped from the distribution point closest to the customer or were backlogged. This situation caused increased transportation costs, delivery delays, and unacceptable customer service levels. In addition, because Prevost’s SAP Min/Max system didn’t compensate for seasonality or trends, the company had too much stock in the off-season and too little when the need was greatest.
To respond to these issues, Prevost’s management conducted a formal evaluation of six demand planning and forecasting systems, including SAP’s demand planning module, to find a solution that would overcome the limitations of the Min/Max system and enable it to use the full potential of SAP’s distribution resources planning (DRP) module. Because the DRP system wasn’t in sync with the demand forecast, product orders weren’t being issued in a timely and accurate manner.
In addition, Prevost Parts was lacking effective demand planning within its multiple site/multiple level distribution network. To accomplish this, the company needed a forecasting solution that could accurately forecast demand for all of its products and deliver accurate safety stock estimates at the lowest level in the network. It also required a system that could provide the flexibility of cumulative lead time measurements, isolate and identify extreme values, recognize seasonal patterns, and provide top-down and bottom-up product group forecasts.
Of the six applications considered, Prevost Parts selected SmartForecasts Enterprise from Smart Software, Inc. SmartForecasts scored highest in the evaluation of more than eight functional criteria. The key deciding factor, however, was SmartForecasts’ ability to generate accurate sales forecasts and safety stock requirements for intermittently demand products. Products of this type are very difficult to forecast because of the slow-moving, irregular nature of their demand patterns.
As Dave Gilbert, logistics director at Prevost Parts, noted, “With most of our parts having intermittent demand, the ability to solve that problem was very important to us, and SmartForecasts’ unique solution in that area was a major advantage.”
In addition, the test results showed that SmartForecasts’ inventory stocking recommendations could reduce required safety stock levels as much as eighteen percent more than the next best application. There were other factors that also weighed in the decision, including SmartForecasts’ ability to interface easily with SAP’s DRP system and simplify the process of managing safety stocks.
The first year, SmartForecasts was installed at corporate headquarters in Ste. Claire, Quebec, where the software is directly linked to an Oracle database used by Prevost Part’s SAP system to store data and forecasting results. Since implementation, the company’s demand planners have used the system to make a number of planning changes, and those changes have begun to pay off. Backorders of the company’s most frequently demanded parts have been reduced 65%, lost sales are down 59%, and fill rates increased from 93% to 96% in just 3 months.
But the implementation of these changes required Prevost Parts to creatively adjust the way it operates. The company’s strategy for improving its distribution activities required a major shift in the way it forecasted demand. Rather than simply forecasting overall demand, the company needed to forecast demand for each product item at each of its distribution and service centers. This would give a more accurate picture of local demand and enable the company to better align its inventory where it was needed most. Local forecasts could then be rolled up into a company-wide forecast for planning purposes.
At the beginning of every month, Prevost Parts transfers the past month’s consumption data directly from SAP tables in the Oracle database into a Demand History table maintained by SmartForecasts in the same Oracle database. Using SmartForecasts, Prevost automatically produces a forecast of demand and safety stock requirements for all products at each of its distribution and service centers using thirty-six months of history. These forecasts are validated by a demand planner and branch parts managers using audit reports and graphical adjustment facilities in SmartForecasts and then passed to the SAP tables for direct use by the SAP DRP program.
Based on these results, Prevost Parts started stocking its warehouses on a monthly basis, but management quickly found that monthly shipments weren’t meeting its replenishment needs. For this reason, demand is now broken down into weekly buckets, resulting in a Just-in-Time ordering process that has lowered stocking levels at its warehouses.
Prevost plans to continue to measure its progress and savings, and fully expects that it will continue to see major improvements in reducing costs, increasing customer service, and effectively deploying its production and distribution assets. According to Gilbert, “We need to have the right parts in the right place to support our customers. SmartForecasts helps us to not only improve our inventory allocation but also significantly reduce transportation and inventory costs.”