Manufacturing Strategy
 - Part 2: The Systems





Part 1 of this series dealt with setting the broad objectives for our Manufacturing System and how we must link this with the overall needs of a business. Without knowing what the business wants it is obviously impossible to determine whether the Manufacturing System is good or bad. Only when we know what the system is required to deliver can we make an assessment as to whether it is doing so or not.

The Manufacturing System delivers the goods, or does not, by means of the production management system and the decisions on production management will largely define the manufacturing strategy or vice versa.

In general terms we are all used to 'production lines'. We take our lead from the original ideas of Henry Ford and the motor industry. The production line cranks along and the work study department balances the line by allocating tasks according to machine rates etc. Even if you do not do it consciously it happens subconsciously.
This whole model of industrial efficiency requires direct labour to be visibly working at all times and the scheduling of the line is largely based on direct labour activity and machine utilisation. This also works on a 'push' type concept i.e. a schedule exists (irrespective of actual sales) and the movement of the line supports the schedule.
Whilst this may have been applicable in the 'any colour as long as it's black' days the world has changed radically since then both for manufacturers and consumers. Automotive manufacturers must now deliver an almost infinite range of vehicles from the same production line and we must also deliver 'bespoke' products of an infinite variety. Production management systems must thus change and adapt and we must recognise that product variety can be a key competitive weapon.

The systems - an overview

There are three contenders for the new production management model and these have varying degrees of applicability to the production of windows. The three basic systems are:

Figure 1: The Systems

In simple terms, for manufacturers of complex items MRP/MRPII can be a powerful tool and it is inherently a 'push' system. A forecast (weekly, monthly or whatever) for sales of standard items is generated by Marketing/Sales and a manufacturing plan is derived (by computer) from this. The plan is exploded into a Bill of Materials for the forecast product requirements and the parts are assembled as per the planning schedule. The forecast demand therefore pushes the production line rather than the actual demand.

On the other hand, JIT is a 'pull' (or demand driven) system and the actual demand for the parts or final articles pulls the system along. The phrase 'sell daily - make daily' encapsulates JIT. This naturally involves a process of reducing inventories and improving system responsiveness at all stages and it can be a very risky game. JIT assumes many other parallel improvements and change to both working practices and culture.

OPT is relatively new in terms of production management strategy and is an overall philosophy for running the business rather than just the management of production. OPT starts by assuming that manufacturing is all about making money, a fairly basic assumption, and then looks at optimising the complete system rather than just individual operations. OPT is a proprietary system and the full version (rather than just the philosophy) is owned by a software and consultancy company. This, however, does not prevent the adoption of some of the excellent ideas it contains and generates.

As philosophies the three basic systems are complementary in some cases and can be mixed and matched. It is important not to regard them as mutually exclusive, you could use MRPII for the rough-cut prediction of purchase requirements, JIT to run the basic production and OPT to cope with bottlenecks and system optimisation.

Materials Requirements Planning / Manufacturing Resource Planning / Enterprise Resource Planning (MRP / MRP II / ERP)

MRP is the most widely used production management system in Europe and the USA and is being adopted in Japan now that they have managed to get right certain other basics such as quality and lead times. A basic concern is that MRP computerises every thing and does not provide a real driving force for change. In the West we have computerised and are now attempting to make things better within the computer system. In the East they attacked the real problems (change, quality, inventory) first and they are now computerising to achieve even better control.

MRPII started out as Materials Requirements Planning and was concerned with ordering and scheduling materials based on inventory. The large-scale introduction of computers led to extensions into a 'closed loop MRPII' system concerned with the whole manufacturing environment. MRPII works on 'push' where a forecast is generated, a manufacturing plan is derived and the plan then drives the manufacturing process via work orders or the purchasing process via purchase orders.
A basic problem for MRPII is that the Master Production Schedule is based on historical data, probably put in when the system was first set up, and this is rarely updated. In fact some companies are not even sure how to update their system data even if they knew who was responsible for doing so. This further ensures that lead times are high even if improvements are made and forces the factory to carry enough stock to produce within the quoted time.

MRPII ensures that you quote a three-month lead-time for an article that may only take 24 hours to make.

The system works like this.

  1. Requirements are calculated and forecast in terms of quantity and time. The Sales/Marketing Department does this.
  2. Requirements are exploded in the Bill of Materials files, which break a product down into its various parts.
  3. The net parts or item requirements are calculated by deducting available inventory from the gross requirements.
  4. A schedule is calculated and work orders issued for the internal production or purchase orders for bought-in items.
  5. The system produces or buys according to the production schedule and the products are available in the forecast volumes at the forecast dates.

Figure 2: The MRP outline

At the heart MRPII is a computer database of parts, components, work-in-progress (WIP), finished inventory and forecast requirements, which are linked by theoretical lead times and production times.

The computer program calculates the best way to meet the forecast requirements and produces a Master Production Schedule to achieve this.

This high dependence on computers means that data in the computer must be highly accurate and kept up to date -the obvious problems can be especially apparent when the computer inventory and physical inventory are different and the system collapses.

The high dependence on computers means the data must be accurate MRPII is best suited to companies that make discrete multi component items, either repetitively, or as one offs. An MRP system accurately tells management what is going on in the factory, enabling levelling of resources. Unlike JIT, however, MRPII does not change or even attempt to change quality control, shop floor planning, relationships with suppliers or levels of stock. Rather, it enables management to make accurate predictions and highlights problems. Inventory levels or lead times may be reduced by improved management control if it is implemented.

As a rule MRPII tends to computerise everything and change little, so it must be operated by a competent and committed management if it is to make a positive impact. As a planning tool, however, it is unsurpassed.


MRPII is not only used to plan and control the manufacturing process but the whole business output. As such the plan must be reviewed if it is to remain valid as customer demands change. The entire plan is generally run for a fixed period e.g. 6 weeks, and is recalculated on a rolling basis every week.

The real problem here is forecast accuracy. If the Sales Department forecasts in Week 1 a demand in Week 6 of 20 off Item A, 10 off Item B, and 30 off Item C then this is what the system will schedule and prepare to produce. If in Week 3 the actual order requirement for delivery in Week 6 changes then it is difficult for the system to produce these (even if the components are physically in stock) without a substantial replanning effort. MRPII is designed primarily for markets which can be accurately predicted and rapidly changing customized products such as windows can lead to continual time consuming replanning.

Simulation and financial planning

The MRPII database and company model can be used to simulate 'what-if' scenarios to allow a response to various possible market permutations. This database can also be linked to various accounting packages to predict inventory levels, cash flow and detailed financial planning information.

For and against MRPII




In terms of applicability to windows there is an obvious concern with MRPII. The system is fine for product such as automotive engines which can have hundreds of parts that all need to come together in predetermined numbers and variants for each engine type. These products are essentially made for 'stock' of a certain type. In windows, at least on the PVC side, we rarely make for stock and make to order is the rule. As such MRP/MRPII has a limited applicability. Some application can be made by allocating each general type of window an Article Number and then exploding this to a general Bill of Materials. This would give a rough estimate of profile usage to enable inventories and purchase orders to be controlled.

"The Manufacturing Strategy" Series

"The Manufacturing Strategy" series is designed to give production managers and their staff some insights into new manufacturing methods and to prompt the industry into considering the benefits of alternative approaches to manufacturing. The series is:

Part 1: Setting the strategy

Part 2: The systems and MRP II (This section)

Part 3: Just in time (1)

Part 4: Just in time (2)

Part 5: Just in time (3)

Part 6: Optimised Production Technology (OPT)

Part 7: A fundamental quality

Part 8: Quality management techniques & tools

Part 9: ‘There's no accounting for manufacturing strategy’

Part 10: Performance measurement

Part 11: Changing roles and things to do NOW!

Last edited: 11/03/10

© Tangram Technology Ltd. 2001

Our standard disclaimer regarding Internet data applies.