Smart production scheduling can have a substantial impact on your delivery performance, productivity and bottom line results. ERP systems are of little value because they will not support real time rescheduling when disruptions occur nor are most ERP systems equipped to capture specific constraints or business rules. So production scheduling is an important addition to your ERP system. But selecting the right production scheduling technology and using the capabilities of a technology in the best possible way are of the essence to really get a good return out of your investment. So what are the key topics to address before entering into a selection and implementation process?
1. KPI’s
First of all you need to define what KPI’s you would like to improve. Lower inventory levels? Higher productivity? Improved delivery performance for your most important customers? Take into account that often improving one KPI may be at the expense of another KPI. E.g. improved delivery performance may lead to higher inventories. Therefore only defining what KPI’s should be improved is not sufficient, defining what KPI’s should not deteriorate is equally important. E.g. Improving delivery performance without impacting stock levels. Once you have defined a set of KPI’s, you can check if a tool is actually capable of handling these KPI’s or is limited to a predefined set of KPI’s.
2. Constraints and business rules
If you read the webpages of Production Scheduling tools it may look as if the sky is the limit. But in reality a perfect fit is essential. A perfect fit means that all relevant constraints and business rules should be captured by the tool and that the optimizer is able to take these constraints and business rules into account while finding the best production schedule. And here is point, if some of your specific constraints are not taken into account, the production schedule generated by the tool is useless. Planners will manually have to reschedule which is costly and time consuming.
So what should you do? Well, before entering in the selection process, try to describe your specific constraints and rules. E.g. sequencing rules or specific rules about production quantities, raw material assignment rules, maintenance rules, machine constraints and so on and so forth. Later in the selection process you could ask the providers the simple question: “does it fit?”, and if they say Yes, “show me it fits”.
3. Horizon
It seems almost obvious but it is essential to know for what horizon you are planning the production. Production scheduling is normally referred to as short term planning, real time up to days or maybe weeks ahead and real time rescheduling in case of disturbances. Production scheduling takes care of exactly determining what to manufacture at which machine with which materials in which sequence. Production planning or capacity planning is normally used weeks or months ahead and is typically based on a mix of real orders and forecasted orders. The goal of capacity planning is to distribute orders in the most efficient way over time buckets, e.g. weeks, without exactly defining the sequence of the production within such a time bucket. The objective of capacity planning is to use the available capacity in the most efficient way and to support the order acceptance and delivery time quoting processes.
Analyse profoundly what will help you more, production scheduling or capacity planning? Starting in the wrong order will cost you valuable time.
4. Optimization power
Here is one of the toughest points in the selection process, what optimization power is needed? The optimization power refers to the capability of the production scheduling tool to really find the best, or almost best, production schedule while taking all constraints into account. But how to compare or how to define this? You will find that most providers promise you they have “great and powerful” optimizers and solvers in their tools. So what to do? Don’t worry, you don’t have to be a mathematician to get some idea of this (important) capability, there are ways to check out. First of all you should require references with similar complexity compared to your situation and check if they really use the optimized production schedules generated by the tool. The second way to manage this is to set targets, e.g. 3% productivity increase without sacrificing other KPI’s. Of course these targets should be realistic, but making this a contractual part of the deal, may help to differentiate between tools that have mediocre optimizing capabilities and real optimization capabilities.
5. Propagation + optimization to avoid the black box disaster
Are production scheduling tools black boxes? Just getting orders from the ERP and sending a production schedule to the MES system? The answer is almost always a clear No. Production scheduling is a dynamic process. Some companies generate a production schedule on a daily basis and adjust it throughout the day, others per week and others have a completely dynamic schedule that is continuously optimized. The key point is that planners will always be needed to resolve problems that cannot be resolved by a tool. So the tool should continuously calculate all the consequences of any disturbance and warn the planners for any problem that will occur and provide the full possibility to planners to manually change the schedule. Here is the important point: to do so not just an optimizer is needed but also an engine that continuously, in split seconds, calculates all the consequences of any change. This is engine is called a propagator because it propagates the impact of a change or disruption throughout the schedule. Therefore an important requirement is if such an engine is “in the box”, if not, you may be caught into the black box disaster, not having the possibility to control and adjust the schedule if needed.
decide4AI supports companies in the selection and implementation of production scheduling and production planning systems.
If you want to know more about decide4AI, follow us on social networks (Linkedin, Twitter, Youtube).
