A-B Forecast / Order Classification and Blocked Stock Monitoring Screens Integration in Arcelik

System Analysis

Arcelik A.S. operates in the durable consumer goods industry with production, marketing and after sales services. Arcelik A.S. manufactures its products in 18 different production facilities in seven countries and offers its products more than 145 countries. Products manufactured in Turkey facilities supply the majority of the customer orders. There are five main facilities in Turkey: Cayirova washing machine plant (CMI), Ankara dishwasher plant (BMI), Eskisehir refrigerator plant (EBI), Cerkezkoy dryer machine plant (KMI) and Bolu oven and cooking appliances plant (PCI).

In the current system, customers send their monthly orders to OM team that combines orders in a single file, as they are a single order. Then, OM team sends total demand to PP teams of facilities. PP teams prepare and declare production plans of relevant month with respect to production requirement of each SKU by considering total order and stock-on-hand amounts. However, in the scheduling process, PP teams make production plans by only considering their own constraints because OM team does not send any due date constraint for urgent orders. Therefore, resulting production plans cannot match delivery plans of the OM team. At that point, OM manually manages the prioritization of orders by directly communicating with PP teams. This method cannot solve the problem completely, but it also causes additional problems.

Secondly, all finished-products go through quality control process. If any product cannot pass quality control test, it returns to manufacturing processes for reworking. RRP are re-planned before delivery. However, SC cannot monitor rework plans.

Observed structure of the system causes:

  • Some unurgent products are manufactured early part of the planning horizon so that the stock level increases. On the other hand, some urgent products are produced lately that the OTIF level decreases. Customers become unsatisfactory for late orders and worry about orders of next months so that they cause bullwhip effect.
  • OM and PP try to manage prioritization problem manually which causes extra workload and planning mistakes. Moreover, manual changes in production plans decreases productivity of facilities.
  • Manual changes cause chaotic problems in the material planning process.
  • Effect of PP teams on OTIF cannot be measured.
  • OM team cannot take action across rework plans of RRP so that OTIF level decreases.

Proposed Solutions

To increase OTIF level and decrease stock levels, two projects are put into practice:

  • A-B Forecast / Order Classification
  • Blocked Stock Monitoring Screens Integration
>A-B Forecast / Order Classification project

A-B Order / Forecast Classification project divides each month into weeks and distributes production requirements to weeks according to their priority class. Demands are collected and classified within the SOP tables of SAP system. To avoid calculation mistakes, production requirements are calculated by inner macros of SAP. These macros are coded by IT team before start-up of the project. On the other hand, algorithms of macros are prepared by Cem Kazan (project leader). After that, users are trained by Cem Kazan. In the last step, all KPIs and related reports are prepared and published by Cem Kazan.

Blocked Stock Monitoring Screens are Web based screens to makes it possible for OM and PP teams to collaboratively plan deliveries and reworks of RRP. These user-friendly screens are prepared together with IT team by project leader Cem Kazan. Project leader also gives trainings to users.


These two projects enables to increase OTIF level from 87% to 95% while helping to decrease stock levels of facilities below 10 days of stock.