| Understanding optimization |
|
How Optimization Works
There's no mystery to optimization. It's a straightforward process that achieves measurable results:
- An optimization model defines and structures your problem
- An optimization engine applies your model to data and searches for a solution
- The output is the best plan or schedule
It all starts with an optimization model
An optimization model is a set of equations that define all of the components in a planning or scheduling problem, such as:
- Resources available
- Demand to be filled or services to be performed
- Operating and capital costs
- Yield and throughput assumptions
- Global operating constraints
- Individual operating constraints and preferences
- Goals, either individual or in weighted combination
- Key performance indicators (information about the plan)
- Decisions to be made

Optimization engines produce plans or schedules
What's especially significant is the sheer amount of information that an optimization model can process. A well-built optimization model is capable of evaluating millions of possibilities, and recommending thousands of individual decisions.
An optimization model's output can take any number of forms. Examples include:
- General six month production plan
- Detailed one week production schedule
- One month workforce schedule
- Truck loading plan
- Set of routes to deliver a day's worth of goods or services
- Number of trades to bring a stock index fund back into compliance
- Marketing-offer assignments for a marketing campaign
- Best loan package at the best price
- Bids to accept in a procurement management system
- When to release seats or hotel nights at a lower price
Plans and schedules deliver astonishing results
Nothing can prepare you for the effect optimization can have on operational efficiency. Plans or schedules you once labored over for days appear in just minutes -- without errors. After you've integrated optimization into your planning and scheduling, savings can be huge:
- A car manufacturer increased productivity by 30%
- Chile's two largest forest-products companies reduced
their truck fleets by 30%
- A semiconductor manufacturer cut wafer-processing cycle time in half, to just 30 days
- A major airline responded to unexpected delays with efficient crew rescheduling, saving $40 million in one year
- A package-delivery company cut costs by $87 million
- A television network increased annual advertising revenue by $50 million
- An investment firm cut transaction costs by $100 million
- A major consumer packaged goods (CPG) manufacturer dramatically increased the direct loading of trucks off its packaging lines
Often, optimization uncovers decisions you might never have considered. No one can analyze so many options, so fast. Optimization's speed gives you time to experiment with different assumptions. You're free to study a range of scenarios, applying your judgement to all your options.
|