If you do not know your supply constraints, then it is likely your constraint is the MRP system.
An operation professional needs to make decisions about three types of actions: sell, make and buy. Supply constraints are centred around the resources you use to make and material that you buy. Sometimes it is obvious which resources or materials act as the constraint. In a complex business, this is not clear to the naked eye.
Operations planning is about managing constraints. Operational excellence means delivering customer excellence whilst operating under limitations of materials and capacity. You need daily visibility on where the constraints are and how they act. Without this visibility you are flying blind, unable to choose the most effective action to maximise customer performance.
Broadly speaking, this is the difference between MRP and Advanced Planning and Scheduling. To understand why MRP fails to deliver constraints-based planning, we have to look back a little into history. In fact, the majority of MRP systems today still suffer from a restriction that hails from the 1960’s. This was when the first MRP systems were being developed for complex manufacturing.
MRP Heuristics- The Rule of Thumb
This is the IBM 305 RAMAC — the first computer to use a moving hard head disk drive and less powerful than a pocket calculator is today. When Joe Orlicky and the people at IBM ran the first MRP systems on this machine, they needed to make some serious computation compromises.
The most relevant compromise is concerning the most important unit of measure in any business: time. Material Requirements Planning did not run in real time, of course, so systems were run overnight. But, more crucially, MRP also made an assumption on measuring time, using discrete days rather than continuous minutes and seconds.
As MRP evolved through Manufacturing Resources to ERP, the overnight runs were preserved to relieve pressure on networked mainframes and servers. Few people recognised the need to use extra computing power to run capacity requirements through minutes and seconds.
The problem is, without a continuous measure of time it is impossible to model constraints. Most resources produce a mix of products that have different run-times. Capacity can not be measured in units of product, it needs units of time that are smaller than a day. JIT aside, materials can largely be controlled to the day, but you need availability to be tied to capacity and updated more frequently than the typical MRP run.
Finite Capacity Planning
So, if you rely on MRP for planning, it is safe to say that you use an infinite capacity model. And you probably have trouble in pegging material directly to a single work order and cannot connect it to a precise release into production. Constraints-based production planning requires a finite capacity model. To many people reading this, and suffering daily from the limitations of MRP, this insight may be blindingly obvious. However, we have to ask why software vendors continue to develop and support a planning logic that is largely unchanged since the 1960’s.
Fast forward to the computing of today. We have computers that are more than capable of doing these calculations across millions of records. And you do not need to buy a super-computer either: A modern PC is sufficient. This is the age of distributed computing where we do not need to rely on a central monolith to do our business calculations.
Constraints-Based Planning for Everyone
Constraints-based planning needs a new logic. This logic can come from expensive and complicated Advanced Planning and Scheduling systems. Or it can be done to simple rules using computers and software that everybody can access. Finite scheduling and constraints planning are capabilities not restricted to those with million-dollar budgets. The logic can be built from the ground up , starting with the process definition and linking it to the flow of demand and supply. Think of this as constraints-based business modeling for everyone.
The Fast Excel Development Method enables operations professionals to apply this logic to demand data and construct a constraints-based model of the business. There is a free finite scheduling course here that will teach you everything you need to know in order to model production constraints over continuous time.
Process improvement techniques like Demand Flow and Theory of Constraints are important to balance the flow across processes, take action to solve constraint problems and quicken the throughput time. A constraints-based model built in Excel is a powerful complement to these ideas. It will also give you a constraints planning tool that can be used to deploy Kanban replenishment and demand-driven order fulfillment.
Knowing your business constraints is becoming a necessity. Businesses of any size can rise up and meet this challenge by leveraging the wonders of modern computing. And, perhaps, a little bit of learning and experimentation.