A Crystal Ball for the Supply Chain: Bringing Current and Future Capacity into the Delivery Equation
In companies all over the world there is a trend that seems to be common. Competition is increasing both from the company’s traditional competitors and from new entrants to the market place, often from overseas. Their customers are looking for a faster turn-round and on-time delivery and these, rather than price, are becoming the key differentiator.
This pressure to be more agile is often seen in make-to-order companies, but even where it is a make to stock environment there is pressure to reduce stocks to a minimum while maintaining a full range of products. This inevitably leads to smaller lot sizes going through the plant and any issues that reduce the efficiency of the plant, lot changeovers, etc, become important.
Many companies concentrate on the engineering issues such as commonality of parts and assemblies, changeover issues using techniques like SMED (Single Minute Exchange of Dies), design issues such as design for ease of assembly, and reducing ‘floor to floor’ processing time by using bigger, better and faster machines.
All of these may or may not be important to your company, but it may miss the point. What we are interested in is ‘door to door’ time rather than ‘floor to floor’ time. In effect we want to concentrate on issues that maximize throughput and minimize lead times and this may not be the same as maximizing resource usage (as traditional accounting methods such as using overhead recovery to cost production would tend to encourage).
Thus sequencing of lots to maximize throughput and minimize lead times is the key to increased service levels while reducing the cost of inventory.
So why don’t ERP systems do this for us? Most ERP/MRP systems have a Capacity Requirements Planning (CRP) module that uses a coarse ‘buckets of capacity’ model in an attempt to indicate capacity overloads. This does not take account of the sequencing of the work within the capacity buckets and provides no ability to carry out ‘what if’ scenarios on potential solutions to problems created by late updates on existing orders or new, high priority orders arriving. In addition the constraints are modelled at work center level and cannot include additional constraints such as operators, setters, tools, space and etc. This is why many ERP suppliers have added some form of graphical and interactive tool, often referred to as Finite Capacity Scheduling (FCS), to replace or augment CRP, either by integrating ‘best of breed’ tools that are proven in the market or developing their own.
Independent studies have shown that there are three critical success factors that contribute to successful computer-based scheduling systems. First, they are fully integrated with other legacy software packages, second, they accurately represent the capacity constraints and process flows, and finally the schedules are generated at least once a day.
There are also many documented case studies that show the benefits of using FCS systems. The wide range of applications and size of company using FCS software these days is a good indicator of these products’ ease-of-use and flexibility to adapt to almost any manufacturing environment.
So now FCS is doing a good job producing better sequencing and achievable schedules that allow companies to be more agile and responsive while maintaining customer service levels. How can we add material constraints? MRP works in particular ways that can prevent many FCS packages from providing an Advanced Planning and Scheduling (APS) solution, where the system takes into account both the availability of materials and resources while generating the schedule. This is because most MRP systems take customer orders and break them down into requirements for individual parts using a Bill of Materials (BOM), then aggregate the requirements for the parts into work and purchase orders at each level of the BOM. The relationship between a work or purchase order for a part and the customer orders is often lost in this process. Sometimes all the relationships at each level of the BOM (not just the customer orders) are lost.
To provide APS functionality the scheduling system must understand these relationships in order to know how to sequence the work/purchase orders to make, say, an assembly. For example, you need to make the frame and wheels, and buy the saddle before you can assemble a bike. In addition, you need the operational sequence for making the frame and wheels (the process route), and the purchase lead-time for the saddle.
APS systems have traditionally been expensive because they often duplicate MRP functionality, re-blowing the BOM to understand the links between the work/purchase orders. They can also duplicate ERP functionality including forecasting systems, distribution software, and other features which introduces unnecessary functionality that users already have and are satisfied with.
Other solutions take a different approach and use the information provided by ERP to peg or allocate materials between ‘producing’ and ‘consuming’ orders and use these links as constraints on the sequencing of work. Schedules produced in this way are more realistic and, importantly, achievable, so that purchased materials are not brought in too early taking up unnecessary space and cash.
The next stage is to spread the use of capacity as a constraint across a supply chain. In today’s environment of ‘lean’ manufacturing and supply chains, you can no longer treat your company as an ‘island’ when it comes to making delivery promises to your clients. You must take into account your suppliers, sub-contractors, logistics contractors, etc., when you answer an Available to Promise (ATP) inquiry.
The more advanced applications of APS solutions allow you to take the actual current stocks and workloads of a supply chain into account when making delivery promises. Therefore, the solution can consider the current and future finite capacity into the delivery calculation.
In the past other Supply Chain Management solutions have used a single high-level model of the entire supply chain that is maintained by you. This model can never be accurate enough to take into account the current and future workloads of your entire supply chain, since much of the work of your suppliers and sub-contractors is not related to you, and therefore can’t be included in your supply chain model. Modern and secure messaging systems and the latest applications of APS solutions have provided a way forward.
Let us imagine that our factory makes Widgets, and our best sales person is currently sitting in front of one of our regular clients discussing delivery dates for a possible new order for 5000 Widgets. Our competitor has already offered a good delivery date, so to secure the order our sales person must beat the competitor’s promised delivery but be certain that the date quoted is achievable. Unfortunately, we do not carry stocks of the materials that Widgets are made from, and to compound the problem we have to sub-contract the Widget plating operation. To give accurate delivery promises we must consider the stocks and resource availability of both our supplier and sub-contractor as well as the capacity of our own manufacturing plant.
To obtain the delivery promise our sales person emails an inquiry to our Widget Maker’s APS. This processes the inquiry and determines that to make the Widgets we must first buy the materials, so it automatically sends its own inquiry to the supplier’s live schedule.
The supplier may be a manufacturer or a logistics/warehousing company. In either case the supplier’s APS system determines if the materials are available from stock, and if not it schedules the operations necessary to manufacture the materials, and then responds to the Widget Makers APS system with a confirmed delivery date.
The Widget Maker’s APS now knows when the materials will be available and is able to schedule the manufacturing operations up to the point of the plating sub-contract process. When it reaches this point it sends an inquiry to the plating sub-contractor’s live schedule. This schedules the necessary operations and again returns the plating completion time to the Widget Makers’s APS. This is then able to complete the scheduling of the last assembly operation on the Widgets, and then respond to the sales person with the required promise date. Although we have not seen the schedules and workloads of our suppliers and sub-contractors during this process, we do know that they have been considered in generating our promise date.
A basic requirement of this ‘Supply Chain Scheduling’ process is a simple and robust messaging system which can pass the inquiries and responses between the users. Both systems will store the messages when an APS is not available, so no communications are lost.
In summary, a system that provides visibility at each stage of the supply chain supported by a peer to peer communication system on a PC network shows what can be achieved with modern APS systems.
A Crystal Ball for the Supply Chain.
About the Author
Dave Snyder is a MOM Channel Sales Executive at Siemens Digital Industries Software. Dave has 30 years’ experience in the Advanced Scheduling space.
