Optilon uses service optimization, a more recent adaption of traditional inventory optimization.
Service optimization use inventory modeling techniques developed to correct the limitation of traditional inventory planning. The techniques can accurately describe the expected inventory behavior of each SKU/location, even for slow moving items. Service optimization is based on the probability distributions which can present a wide variety of dispersion and skewness patterns, and considers all relevant parameters. The result of this fully automated process is the stock-to-service (STS) curve. It gives a precise relationship between the desired service level (order line fill-rate) and the average stock on hand required, while providing the necessary control levels (safety stock, reorder level, order-up-to, etc.).
The forecasting approach in service optimization can describe a wide variety of stochastic demand behaviors, from fast moving to intermittent demand patterns, with a self-adaptive probability distribution model. This is essential to provide reliable information to the inventory modeling step.
Two key features in service optimization are:
Automatically determines the optimal target service level which must be assigned to each SKU/location to achieve the service level defined by management as an aggregate global business objective, while optimizing, at the same time, an objective function. This stock optimization process can also consider various constraints that can be defined both at an aggregate level and/or by individual SKU/location.
Optimizing the “vertical” allocation of inventory across the various echelons of the supply chain which can be distribution network tiers (often referred to as staging or risk pooling) or bill of material levels (often referred to as Postponement). In contrast to similar approaches, service optimization translates “imperfect” service levels in the upstream echelon into stochastic delays to the downstream echelon. Deterministic values are not good enough.