Evaluating an APS

Is an APS System Right for You?

This is the third in a series of three articles on scheduling systems.  
See the first article here. 
As discussed in part two of this series on scheduling systems, evaluation and planning is crucial before determining what type of system is right for your company.  The previous article listed eight questions your company should answer before committing to an APS system.  Once these questions are answered and you have a designated scheduling person in place, it is now time to evaluate the right APS to match your requirements.  
In addition to the importance of having a designated person to handle the scheduling, it’s imperative that you know how you will update the scheduling system. If that is unclear, take the time to work this out before starting an APS. You will need to have a good way to send and receive frequent updates to/from the factory floor.  Otherwise your wonderful new powerful scheduling tool most likely will not reflect reality to the people on the floor or the person doing the scheduling.
Next do a little investigation into your data files.  Specifically, is the information in your routings accurate with regard to designated resources?  Often, changes on the factory floor are not maintained in the master routing files since production management are usually the last department of the organization on the list when it comes to having useful information from the ERP system.  The Bills of Material (BOM) are, in most cases, much more current because the buyer/planner and engineering teams tend to be part/item focused.  So print a list and do a quick check.  Make sure you have the right resources attached to each operation step.  These resources can be consolidated at the work center level, because a good APS will allow you to associate a specific set or a group of resources (machine, lines, and cells) to a work center.  
Next have a meeting to discuss your “rules” for scheduling.  That is, how do you decide on what to run today on first shift (and if appropriate second and third shifts) today and tomorrow?  Is it simply trying to make due dates, or do you use grouping logic, or theory of constraints for example? Do the supervisors on the floor have the authority to override the scheduling order, and if so, what is their logic in rearranging the order (picking the jobs with the smallest quantities to help flush the work center for example).  Write down what you have learned so you can share it with the scheduling software partner organization.
Next, visualize having a scheduling software system that has the ability for frequency updating with accurate routing data and is up-to-the-minute with the floor activity.  In addition, you should have (or you must create) rules by which to guide the logic of the APS.  After rules are established and agreed upon, you will have the basis for evaluating the results of each scheduling run. 
Now that you can schedule at the touch of a button, it is time to evaluate your schedule outputs by a measurement that matches your company’s goals.  There are many such KPI’s (Key Performance Indicators).  For example:
1. Queue time 
Queue time is the measure of how affectively you are scheduling your plant.  Queue time, from an accounting perspective, is a measure of a potential unnecessary accumulation of WIP inventory build-up and possibly of an inefficient production process.  Queue time is often created during MRP planning. MRP uses infinite capacity logic.  In laymen’s terms, whatever demand that you have from customer orders and/or forecasts, MRP assumes you can fulfill regardless of the factory capacity. Worse yet, some MRP systems calculate a “queue” which is never a fixed number to depend upon. APS uses finite capacity logic and will only schedule work for the actual capacity that you have, therefore reducing queue (which is dynamic and changes as demand/supply changes) and increasing velocity of parts through the plant.  It is therefore essential to create an easy-to-use method of tracking actual queue time once you start using an APS so you can measure your improvements over time.
2. Late Jobs
Another key KPI is a factory total percentage of a job that is going to be late.  This “macro” look at the total lateness of all the jobs can be useful when combined with a Scenario Analysis Feature found in some APS products.  Ultimately how you satisfy your customers due dates (or not) is true testament of how effective your scheduling systems is.
3. Set-up Time 
A third key KPI to keep in mind, is overall plant setup time.  An effective APS implementation should see significant reduction in setups by allowing “rules” to group like items, materials, processes, etc. together to minimize set-up time (and therefore cost) without hurting on-time delivery.  In addition, tracking that trend over time will show you how affective your APS systems is.
Additional KPIs:
1. Daily Setup Hours
2. Average Lateness 
3. Number of Late Jobs
4. Percent of Late Jobs
Scenarios can be a useful tool to help analyze various scheduling methods with different rules or to simulate overtime conditions that will directly affect capacity.  Also, a scheduling person should be able to run multiple simulations while keeping the “live” or actual current scheduling database intact.  Examples of simulations might include adding in downtime for preventative maintenance to one or more of the floor machines.  Through simulation in a scenario, the scheduling person can see when he/she can schedule it without hurting on-time delivery.  Simulations may also help you to understand how adding machines and “dummy” jobs will affect the floor based on using some of the above KPI’s.
Two final thoughts: first, some APS’ have a built-in automation that will help use the KPIs to drive the best scheduling scenarios.  These “co-pilot” facilities combine the power of KPIs and scenarios in logic that can help improve scheduling productivity while resulting in the optimum schedule for the factory.
And last, make sure your factory has a “systems” approach to analyze the effectiveness of any APS system.  Beyond the requirement to have up-to-the-minute visibility of factory floor production is the need to validate that the schedule is working as implemented.
Next in the series is the “what to avoid in factory floor data collection systems.”