Making more money with Finite Capacity Scheduling
The ever increasing pressures from the market place has caused companies to re-assess how the production planning function is carried out and to look for tools that will support the decision making process and generate achievable schedules in a timely manner. This paper describes the use of a graphical, finite capacity scheduling package that can be linked with existing systems such as MRP software to provide the user with a rapid decision support tool that balances demand with capacity.
Introduction
This paper describes the increasing use of decision support tools such graphical schedulers or 'electronic planning boards' to help to understand problems in the production environment and to generate achievable schedules on a day-to-day basis. The text firstly covers the term 'finite capacity planning' and explains how this compares to traditional planning systems. This is followed by a case study of the installation of a FCS system, developed from an European collaborative project under the EUREKA initiative, in an engineering company. Finally the benefits of using finite capacity scheduling and how this is translated into savings in variable costs is discussed.
Current Status
Production planning and scheduling have always been an important element in maximizing the efficiency of production plant. Inefficient planning and unclear objectives can lead to significant increases in variable costs. However the emphasis in terms of the way that schedules are produced, is changing. Currently, production planners see their goal as making best use of resources since standard methods of costing invariably makes 'maximum utilization of plant' the key cost driver. The pressure in the market place is now shifting to providing 100% service levels at shorter and shorter notice. This has shifted the production planner's emphasis from 'maximum resource utilization' to the elimination of stock-outs and on-time deliveries by making the right products, the right orders, at the right time. This may require a re-evaluation of traditional techniques such as batch consolidation, sequencing for minimum set-up times etc.. Now producers will be forced to examine the impact of running smaller and smaller production batches in order to react to variable demand. Attention is thus turning to decision support tools that can help production planners make decisions on a daily, if not hourly, basis.
In many industrial sectors, particularly the small to medium make-to-order business, companies have undergone significant changes over the last 10 years. Many of these were controlled by the original owners of the business who had an intimate knowledge of the products and processes. As they retired or their businesses were acquired by larger organizations, they were replaced by managers with a more financial background, and in trying to deal with the day-to-day control of production, they tried the tools with which they were most familiar, invariably spreadsheets. These managers then realized that a spreadsheet could not deal with the complexities of their business.
Other systems on the market have proved to be expensive, require extensive consulting activity to install and maintain, and often do not provide an interactive interface that the planner can use to make fast changes in the event of unexpected changes in demand or plant performance.
Finite Capacity Planning/Scheduling?
Finite capacity planning and scheduling is a process whereby a production plan consisting of a sequence of operations to fulfill orders is generated based on the real capacity of resources. Resources can be machines, tanks, piping, labour, storage areas, fork-lift trucks, delivery lorries, tooling or anything that could constrain production processes.
Most planning systems assume sufficient resources are available to produce batches when required, that is, resources have infinite capacity. Traditional planning systems take the orders, break them down into component parts or ingredients and then calculate when to launch the batch based on the individual lead times (perhaps adding adjustments for waiting time etc.) and when raw materials are required taking into account current stocks. No account is taken of the real capacity of resources to produce batches in a timely manner so that it does not matter if resources have a high utilization or not, the same lead time is used to calculate the launch time for any batch.
At the same time that the launch time for batches are calculated, the materials required are also ordered to arrive in time for work to start. If there is a delay in production upstream of a particular operation then the material will be ordered too early. With no concept of bottlenecks available to the planning system, resources become overloaded, queues of work get longer and work in progress increases. Because now, batches must join queues for resources at each process step, orders take longer to make progress through the process route, expected lead times are too optimistic and deliveries are late.
Using finite capacity scheduling, operations are only planned when resources are available. Materials are ordered only when they are needed for the operation to be carried out. Case studies have shown that inventory levels fall dramatically, key resources are better utilized and flow of work more controlled. Work in progress remains at a relatively constant level, lead times are more predictable and delivery dates more reliable. In this way production management spends less time progress chasing and fire fighting and can concentrate on the every day job of balancing, often variable demand, with the capacity available.
Keeping a company's production facility running smoothly is no mean feat.
Figure 1 A question of balance
This is the art of the production controller. He/she must constantly try to maintain the balance between demand and capacity. On the demand side the 'flow' of work can vary due to new orders arriving, changes in priority, and stock orders launched against forward forecasts of demand. 'Leakages' take place as deliveries are made, forecasts are translated into real demand and orders cancelled. The options available to the scheduler to increase capacity might include overtime, subcontracting, splitting batches, and avoiding bottlenecks, while leakages in capacity will occur due to machine breakdowns, operator absence, scrap, re-work and excessive changeover times. Any decision support tool must be able to support this balancing task.
Case Study: FFD Ricardo
FFD Ricardo, part of the Ricardo Group of companies, specialises in the design and supply of drive-trains and chassis components for F1 cars, rally cars and off-road vehicles. Based in Wolston, near Coventry in the UK, FFD Ricardo employs around 160 people of whom 60 are involved with design, 60 in manufacturing and 30 in motor sport. A wide variety of products are designed including frames, steering and suspension systems, cooling systems, fuel systems and exhaust systems. The nature of the market means that FFD Ricardo have to respond fast to requests for new components and modifications to existing ones, and the need for accurate and reliable deliveries by their customers is paramount.
A wide variety of process routes and steps are required. Traditional CNC machines for turning, milling and grinding are complemented by sophisticated gear cutting, hobbing, laser etching and grinding machines. In addition FFD use a number of sub-contractors for shot peening and welding, as well as other processes where the in-house facilities cannot provide the capacity. Heat treatment is also carried out on site. Typically up to eight orders would be added every day of which there would be up to 12 components types. The average batch size is 10 and this meant that more than 1000 components would be added to the schedule every day.
Figure 2 An electronic planning board
Prior to installing a FCS system in 1996 FFD's method of scheduling was to use a wall board with T cards. Prioritisation was the result of a morning ritual of progress chasers discussing which operations would be started next on each machine. This method of schedule control could be described as 'He who shouts loudest'. Neil Hackett who is responsible for production at FFD tells a familiar story.
"The situation was becoming serious. The wallboard consisted almost entirely of red (late) cards, work in progress was steadily increasing, and we were losing customers to our competitors. When we asked our customers why, we realised that we had to tackle the unreliable delivery problem fast. We needed to concentrate on overcoming the problems that we have always had, but were hidden away by a sea of inventory. I have for many years been an advocate of the Theory of Constraints (TOC) and had introduced this to a number of my former companies. We looked around for a tool that could be configured to match the techniques of TOC, be interactive, get our work-in-progress down, and get us out of the chaos".
Now, Neil loads the schedule every Monday and identifies which orders are predicted to be completed close to or past their required delivery date. This identifies the priorities and an action plan can be generated to overcome the potential problems. Then, as progress is made, this is reviewed to see if re-scheduling is required.
By introducing finite scheduling FFD Ricardo have achieved the benefits of TOC and the delivery performance has substantially improved. There is no need for expediting and the wallboard has long since been consigned to the scrap bin. They are now able to offer a fast response to 'requests for quotations' and the engineering focus has improved. As Neil reports, "before using FCS our planning board was a sea of red and we were losing business to our competitors. Now we are back in control and beating our competition. Best of all we have been able to focus the company on making money and FCS has been the tool to enable us to do it".
Cost vs. Value
The case studies are good examples of how management is having to re-assess traditional ways of doing things and introduce new methods that may be quite radical for the production planner but increasingly required in the ever more competitive world. No longer is the goal one of maximum utilization of machines. Now the key is giving customers a 100% service level which requires making the right products and the right orders at the right time even if utilization of machinery is lower than in the past.
Control of variable costs is obviously important in meeting the need to maintain good profit margins in any company. The cost of holding inventory is an important contributory element to variable costs and case studies have shown a major reduction in inventory when production is planned using finite resource constraints. This is an obvious tangible benefit.
However how can we measure the advantage of on-time deliveries?
We have seen that pressures in the market place are changing and how some companies are reacting to this. The challenge now, perhaps, is how do you put a value on service. The old cliche that accountants 'know the cost of everything but the value of nothing' springs to mind. What is the value of service? Lost business, perhaps, or the ability to differentiate in the market place, not on price, but on ability to react to your client's changing demand. Some advantages of changing the approach to planning and scheduling of production may well be tangible while others, of equal importance, may be intangible. There will be pressure on managers to put a value on these intangible benefits as well as quantifying the cost of providing an increased quality of delivery.
