Smart Software for Service-level Driven Forecasting




The issue of forecasting intermittent demand has been one that has plagued companies forever. Many very smart people have tried to tackle it; many vendors like to say they focus on it. Few vendors make it as singular a focus as Massachusetts-based Smart Software. Principals Charles SmartNelson Hartunian (Ph.D.), and Thomas Willemain(Ph.D.) have been studying the issue for almost 30 years, have perfected their own proprietary forecasting method (the Smart-Willemain method), and won numerous grants to study the issue of intermittent demand.

Forecasting demand when the demand history has been a consistent level of X units per period of time is not particularly difficult; neither is forecasting demand when there has been a consistent and steady growth in the level of demand. This is not to say that forecasting is easy, but forecasting steady demand, or steadily growing demand, is relatively straightforward.

Volatile, intermittent demand is another matter altogether. Demand that has numerous, unpredictable spikes in certain periods, and other periods where there is no demand, can make for a particularly challenging forecasting scenario. Working with a few items, one might be able to apply some human judgment to the unveven demand pattern and generate a forecast of reasonable accuracy. Service parts, however, often represent a large investment for many companies, sometimes one of the top assets on a company’s balance sheet, comprising thousands of parts, where applying human judgment across the portfolio is often impractical.

The most common sources for this type of demand include spare parts, aftermarket supplies, and capital equipment—common needs for enterprises such as mass transit operations, maintenance facilities, service operations, distributors, and the military.  Smart Software’s customers are in, chiefly, the automotive, aerospace, transit, consumer products, distribution, durable goods, industrial equipment, food and beverage, healthcare, hospitality, high tech and electronics, pharmaceuticals and chemicals, and utilities and communications industries.

Smart SoftwareSmart Software, founded in 1984, focuses its competencies around forecasting, demand planning, and inventory optimization (no multi-echelon inventory optimization). Its core intellectual capital is, however, clearly its forecasting expertise, which is the foundation for all that Smart Software does. Smart Software has applied its deep forecasting expertise to a service-level driven approach designed to help find the inventory sweet spot for those items with intermittent demand.

Smart Software approaches these problems with just two products, SmartForecasts and SmartCollaborator. SmartForecasts is an engine that selects the right forecasting methodology, whether that be one of the established forecasting techniques, or Smart Software’s patented methodology, to manipulate service level, lead-time, and safety stock in the generation of a forecast. SmartCollaborator is a Web-based companion to SmartForecasts that can be used by others inside the company, such as regional or divisional sales managers, to interact with the forecasting process.

Ever the innovator, Smart Software now seems to be working on taking the forecasting of intermittent demand a step or two further, as suggested in their November 2012 announcement of another grant from the National Science Foundation. Although Smart Software has been loathe to talk in much detail about this new research effort, the effort of the research will be to consider the correlations between the demand for one item with the simultaneous demand for other items, analyzing these relationships across clusters of items, to produce a better understanding of service part relationships.

This software has been built to complement ERP and other systems. Smart Software has in fact been an Epicor partner for the past ten years, and has had earlier alliances with other supply chain solutions. These alliances, together with Smart Software’s deep forecasting expertise, focus, and straight-forward approach and interface, means that Smart Software finds itself on the shortlist for numerous forecasting opportunities. There is nothing intermittent about Smart Software’s focus on forecasting for intermittent demand.
 
comments powered by Disqus