Page 330 - Orlicky's Material Requirements Planning

CHAPTER 18 Process Industry Application 309

APS engines include the business application of simulation, heuristics (best-of-busi-
ness rules), linear programming (LP), and constraint-controlled intuitive modeling such
as fuzzy logic. These sophisticated mathematical modeling tools embrace all aspects of
the business that can be affected by a supply-chain management implementation.
Integrating demand information from the customer’s customer to the supplier’s suppli-
er is enhanced by also incorporating additional dimensions of the business such as the
revenue chain. This allows more comprehensive “what if” scenario planning in evaluat-
ing new market potentials, corporate takeover return-on-investment analysis, price-sen-
sitivity analysis, and logistics configuration analysis.
The first process flow scheduling tools were implemented in the late 1980s and
focused on the consumer goods industry. The calculation models were rather simplistic
and attempted to provide the single “best” answer based on the model input. As comput-
ers have become more sophisticated, the models representing the business have become
more sophisticated as well. It is now possible to create almost “virtual reality” for the busi-
ness to evaluate alternate plans. As in running a real business, a single, large model is
insufficient to represent all the integrated functions within an enterprise. Effective PFS sys-
tems have the ability to define multiple models to best describe the business.
The three main approaches for solving the process industry scheduling problem
include simulation, heuristics, and optimization. Simulation attempts to represent in a
computer the interrelationships of a business. An effective simulator directly relates to the
business and allows the iteration of many different decisions to determine the impact on
the enterprise. The simulator can be as simple as a spreadsheet that provides “what if”
capability for different production schedules. The openness of current technology sup-
porting MRP systems provides the opportunity to download information from the main
system into spreadsheets for seamless manipulation in “what if” analysis. The simulator
can handle important tradeoffs analytically and clearly identify the impact of certain deci-
sions. This quantifiable analysis allows decisions to be made on the basis of fact, and fewer
decisions are based on intuition. At times, the amount of data becomes overwhelming,
causing difficulty in building a mental picture of a particular problem or solution.
The simulator can be used to develop a more sophisticated model to reflect a par-
ticular enterprise. These simulators can provide an excellent teaching tool or can help in
managing a real business by allowing management to see the consequences of decisions
before implementation. With the continuing growth of computer processing power and
the decline in computer processing costs, simulations are sure to become more widely
used in day-to-day operations.
Heuristics are simplifying rules or rules of thumb that are used to develop a feasi-
ble schedule. These rules are based on intuition or experience instead of mathematical
optimization. These rules may be required because the simulation and optimization may
not be able to provide a feasible solution without them. Another use for heuristic rules is
to develop an initial solution from which improvements can be made. An example may
be that production cannot be increased or decreased more than 10,000 units for each
major schedule change. Another rule may be that major schedule changes can be incor-