DOSE
is an abbreviation meaning Determine Optimal Stocks Efficiently.
This is a common name for the two software products DOSE-TRADE
and DOSE-SERVICE solving optimal inventory control problems
for trading enterprises and tech service enterprises.
You have no idea of mathematical inventory control theory
but would like to know what it is about.
You will probably agree that among conditions for success
of a trading enterprise the ability to avoid overstocks
and lengthy stock-outs is not the least. These ailments
occur when the timing and the volume of purchases are badly
coordinated with sales.
Demand, determining the volume of sales, is usually formed
by random factors and to influence it using such methods
as advertising and sales-out is rather expensive. On the
other hand you can usually order replenishments when and
as many as you wish, so purchases are easily controlled.
The question is how to find optimal timing and volume of
purchases on the basis of random demand characteristics
and replenishment pattern. Intuitive (or rule of thumb)
solutions in this case are not very efficient.
Efficient solutions of such problems are due to mathematical
inventory control theory developed by many economists and
mathematicians from various countries over the last half
century. DOSE-TRADE implements the latest achievements of
this theory.
The main premise for its successful implementation is relative
stability of the trading enterprise (constant suppliers,
stable demand, technological discipline) and there are no
other stumbling stones.
DOSE-SERVICE can calculate optimal full stocks for any
number of stored spare parts on the basis of minimal input
data and taking into account the most important quantitative
indices of service level.
You have met with similar programs and would like to know
how DOSE differs from these.
Beginning from 70's there appeared on Western markets numerous
program implementations of mathematical inventory control
models often built in trading control systems as modules
producing recommended orders. Mostly they were continuous-review
models with demand described either by Poisson or normal
distribution. DOSE-TRADE can solve problems with compound
Poisson demand for continuous-review case as well as for
periodic and random opportunities for replenishments . This
approach is more adequate and gives (as simulation experiments
with real-life data show) much better solutions. In addition
to usual two-level control policies DOSE-TRADE can find
optimal policies with fixed case size. It always uses exact
methods and finds global optimum in acceptable time. It
has a built-in simulation block which allows quantitative
efficiency comparison of your existing inventory control
system with DOSE.
DOSE-SERVICE gives you a choice of three replenishment patterns
and three recovery patterns for each item in store. It provides
you with graphic representation of optimal balance between
investment in stocks ("price of warehouse") and
service level. It lets you choose interactively the final
solution out of a range of solutions . It uses very quick
algorithms which are capable of solving problems with many
thousands items in acceptable time. It may take into account
the current budget constrains on purchases you make.
