The term ‘optimization’ can and does have different meanings to different groups. For the folks who build and develop scheduling algorithms, creating the best schedule is defined in terms of cost criteria – perfectly logical. For business settings (from manufacturing to hospitals) optimization refers to the entire process. Let’s look at the scheduler’s world to identify key steps to optimize the process.
Step 1: Determining and Collecting the Data Needed
The scheduler has a difficult job because they have to deal with many people who have conflicting objectives. The job is further complicated by the scheduler having to gather a bunch of information from different sources. The scheduler spends most of their time gathering the right information, like the current status of the shop floor, the demand in the next few weeks, the shutdown schedule, and the projected receipts of raw materials and so on. The scheduler will make the job easier by automating this data collection and providing the data in a single window. Therefore, step 1 in the scheduling application is to determine the data needed by the scheduler to do scheduling and gather it in one place.
Step 2: Data Consistency and Gathering Outside Data
The next problem that the scheduler has is the data from many sources is not necessarily consistent. Also, some of the data the scheduler needs is not kept in any system. Therefore, we need to provide the tools to analyze this data and point out inconsistencies. We need to identify problems for example, the schedule has an operation for 200 for a product, but you are swimming in inventory for that product. The scheduler has something scheduled, but there is no inventory of a component required for producing it, we are missing orders for product xyz etc. Therefore, step 2 in the scheduling application is to relate the data and determine internal consistency.
Step 3: Assessing, Projecting, or Making Visible the Consequences of a Decision
The other problem you have is even if the scheduler knows the problems; they can only react to the NEXT problem. It is difficult for you to see the consequences of the changes an scheduler makes without a tool to quickly show the consequences of changes. The scheduler may well be creating problems in the future, but you don’t know because there is no time to evaluate things well for a reasonable length of time. The best that you can do is worry about the next major event like a shutdown of demand spike. Therefore, in step 3, the scheduling application provides an interactive environment that captures the scheduler’s decision(s), “simulates” the impact of the decision(s) to create a projected sequence of activities, and creates a visual of the projected schedule. The direct benefit of this is that the scheduler will have a longer time horizon, the ability to see projected schedule, and an environment to try what ifs.
Step 4: Is the Decision Consistent with the Business Objective?
If the consequences of the changes are going to be visible to the scheduler, we need to make sure that the changes are consistent with the business objectives as represented by the S&OP. Step 4 is to represent the S&OP plan within the scheduling environment and provide the tools to easily compare it with the current schedule.
Step 5: Dealing with Conflicting Objectives
The scheduler has to deal with conflicting objectives from people in the organization. The marketing people always want to react to their needs, and meanwhile the manufacturing guys are complaining about the schedule changes. To help the scheduler get some of these conflicts resolved, we’ll need to create a quantitative framework to measure how good a schedule is. Therefore, step 5 is to create quantitative criteria for measuring the schedule. The objective is to quantify only those elements that can be supported with data. Of course we can’t quantify everything, and it is not the intent to, but at least we can try to list the quantitative things like number of orders missed, setup costs, etc. to provide a way of costing out the schedule. The advantage of this is that the scheduler will be able to compare the quantitative attributes of two schedules. So now, if marketing wants to make a change, the scheduler is able to calculate if it is going to cost $50,000 or $2,000. This will let the scheduler channel his or her energy to fight the important battles rather than fighting all of them.
Step 6: Automate the Routine Aspects of Scheduling
If we do all the things above, we’ve gotten to the point where we have some control over the scheduling. Once we are comfortable with this, the next thing is to reduce the repetitive work we do in maintaining the schedule. Therefore, step 6 is to automate the routine activities in scheduling.
Step 7: Optimize an Ongoing Exercise Contributes to Sustainability of Solution
The steps up to now haven’t really optimized anything and there may be much better ways of handling some of the tasks like lot sizing, sequencing, and resource usage. Together with the scheduler, we can identify those opportunities, so step 7 is to identify opportunities to optimize and implement these opportunities. This last step is an ongoing exercise and contributes to the sustainability of the solution.