A Materials Planning Case Study, Part 1

Long replenishment cycles and purchasing restrictions are a constant challenge in demand planning. Add volatility to the mix, and companies find they are either overstocked or not ready for delivery. “Gremlin Electronics Inc.” (name changed for confidentiality reasons) sought an automated solution to manage both planning and delivery.

UK-based Gremlin Electronics is one of Europe's leading electronics companies. The components used in its products are designed by Gremlin but manufactured in China. Component delivery schedules range from 60 to 150 days as shipping containers from Asia to Britain takes about six weeks. To satisfy the customers’ needs with earliest possible delivery times while keeping stocking costs low, quick reaction and flexibility are essential.

In order to achieve this balance, sufficient stock is required to allow for long lead times and to maintain delivery readiness. Gremlin's clients expect delivery within 24 hours after placing a regular order. Purchasing, however, has to order components much further in advance. If the demand ceases or changes, items become surplus and overstock — a common challenge for many companies.

If, on the other hand, demand rises, purchasing can't acquire components quickly enough. To reduce just the shipping time, higher air freight charges must be paid. Even lower-cost sea cargo is a bargain only if cargo containers are filled completely. This usually means regular purchasing of surplus amounts. Combining shipments– that is, aggregating demand from a number of suppliers — doesn't always solve the problem. Once a container is loaded, orders can't be revised.

Gremlin Electronics’ range of products includes about 3,000 active items annually and about 600 items in either a phase-out or a phase-in process. Planning these phase-ins and -outs is an additional challenge which, if automated, requires appropriate software support.

Intense monitoring of the phase-in process, including immediate correction of planning, is an established part of the routine. Phase-out, which is more difficult, is measured by the residual stocks of phased-out items and in the residual demands that cannot be covered anymore.

Gremlin implemented the Diskover SCO software system to simplify the phase-out task. The system surveys the demand development and proposes items for phase-out. During the process, it calculates remaining stocks and then under-stocks all the components associated with a phased-out item.

How to handle exceptions while 20,000 orders pour in each day
Gremlin's customers are specialized trade and wholesale companies. Demand and material planning with these B2B clients requires a high level of flexibility. Failure to meet expectations can result in penalties. Gremlin processes about 20,000 standard orders a day, most of them for overnight shipment.

A small number are categorized as “exceptional.” These exceptions are sporadic, usually very large numbers of items to be delivered within a defined period. Exception-planning is an important but difficult task for materials planners. Exceptions include marketing campaigns or store openings for which the customer needs large batches of certain items just once. If the customer alters the order after it's placed, stocks inevitably increase. Returns and allocations contribute their fair share to a rising stock-level as well.

Gremlin's main target was to increase availability for delivery while significantly and sustainably reducing costs — including stock and storage, air freight, penalties, and inevitable depreciation and scrapping.

Gremlin was able to use Diskover SCO's advanced demand- and materials-planning functions in support of its existing ERP system. A specific feature of Diskover SCO is its simulation function. The system simulates the entire supply chain's performance on a daily basis over a certain period of time (usually the previous year) and helps select the best alternative for each situation. Gremlin defines the parameter settings of the ERP/Diskover SCO system by running a variety of “what if” scenarios based on the question: “What logistical performance would have been reached if we had applied the strategies in question over the past year?”

In the second part of this article, we'll look at how Gremlin uses these parameters to optimize materials planning.

6 comments on “A Materials Planning Case Study, Part 1

  1. elctrnx_lyf
    June 20, 2012

    I believe many of the big manufacturing houses employ certain kind of software to analyse the demands for the products and also provide the results to the planning team to create the right manufacturing plan. But most of the times these software could be commercially available with lot of customisation requirements.

  2. Barbara Jorgensen
    June 20, 2012

    I'm out of my league when it comes to software. I'm sure there are out of the box solutions, but in manufacturing and supply chain, most of them seem to be customized. I think SaaS is supposed to replace that to some extent.

  3. Taimoor Zubar
    June 21, 2012

    I believe many of the big manufacturing houses employ certain kind of software to analyse the demands for the products”

    @electrnx_lyf: I've worked on the development of these software, but there's a certain limitation to them. They can take into account past trends and other inputs like external factors but the output rarely matches the actual demand. This is because there's so much volatility out of the scope of the software and you cannot take into account all of it.

  4. Taimoor Zubar
    June 21, 2012

    I think SaaS is supposed to replace that to some extent.”

    @Barbara: SaaS is a very useful way for manufacturers to use these demand planning software. Through SaaS, they do not have to purchase the software themselves and deploy it. Hence, they save on the hassle of managing it. Besides this, they don't pay for it once and can pay as they use the service. This makes SaaS as a cost-effective option.

  5. Himanshugupta
    June 21, 2012

    @TaimoorZ, i wonder what kind of engine is deployed in these softwares so that the software is useful to some extent. I do not know whether my comparison is applicable or not but financial industry also faces a lot of uncertainity on a daily basis and there are risk management departments which lower the risk exposures. Can similar algorithms be useful in managing the inventories, i wonder.

  6. Taimoor Zubar
    June 21, 2012

    @Himanshugupta: It varies from industry to industry but in order to predict demand, the software looks at historical data for the last few years and for the relevant time period (month, week, day etc) and the growth factor. It also considers other factors that influence demand and are not reflected by the trends. This could include extra demand for a product because of a promotion or weak demand because of the launch of another competing product etc. All these together help in giving an estimate of the demand.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.