Epicycles has been developed in part to support that life cycle.

Its origin is with R.H. Hayes and S.C. Wheelwright in two articles:-

• "Link Manufacturing Process and Product Life Cycles", Harvard Business Review, Vol. 57 1979, No. 1, pp133-140
• "The Dynamics of Process-Product Life Cycles", Harvard Business Review, Vol 57 1979, No. 2, pp 127-136

We've taken the liberty of adding some of the Systems dimension (in red) to the original matrix as suggested by Silver and Peterson.

If we have understood it there are several aspects of the matrix which are of interest to the systems designer:-

• Process pattern classification and product mix are related and are a result of strategic business decisions about capital, scale and markets, not a fundamental of how things are made.
• The management challenges for a product mix - process combination are predictable.
• The production and inventory control systems are appropriate to the position in the matrix.
• The long term consequences of utilising a system from one part of the matrix in another part is system failure, high cost and/or business failure.
• Position in the matrix is not static, both product mix and process evolve as business strategy changes.
• The tendency is for businesses to evolve towards continuous processing of commodity products through product and process life cycle maturity.

The initial value of the matrix to us was in confirming the origins of traditional MRP within a particular area of the matrix and the consequences of misapplying it. It seems that limiting the computer based system design to a small area of the matrix creates difficulties in applying it in other areas. Also, since the evolution of the business is typically (not always, as described below) from job shop to continuous flow, that business strategy aspires to continuous flow,  it seems that an appropriate systems approach is to first consider all requirements then differentiate where necessary rather than consider one area of the matrix first and evolve to cover other areas.

Our simple conclusion is that cementing in place a traditional MRP system (or any other system localised with the matrix) will inhibit the evolution of a business through its product and process life cycles. The design of computer based production planning and scheduling systems is as strategic a decision as that of process and product mix design.

While it is always possible to change the system there may be some difficulty in finding the right time or finding the time without distracting the business. It seems to us that this is waste, far better to have a system which can provide for business evolution.

The unusual but short story is that production began differentiating the product (because they could, and were aware of the considerable market potential and premium prices), marketing reluctantly followed and began selling differentiated product. Market acceptance was swift as it rippled through to consumer products. General acceptance and excitement followed within the first 6 months. Initial systems changes within production were a reversion to manual systems, it took much longer before the computer based systems caught up. After 20 years the full implications of product differentiation in that business environment are still being discussed and further related computer based systems development continues to occur. The strategic decision to differentiate product which was eventually formalised could be viewed as essentially a decision to initially step backwards up the product process matrix, and allow the disruption to occur, then evolve.

An outcome of the matrix for us was to highlight the relationship between production and inventory control and cost (management) accounting. MRP and Job Order Accounting have similar origins while Just In Time and Process Costing have similar origins. The possible lack of relevance of management accounting systems in support of the underlying production process have been well documented by the proposers and practitioners of activity based costing.

In the same sense that we have questioned the distinction between materials and capacity we are also led to question the difference between Job Order Accounting and Process Costing. We believe they have more in common than they have differences. Activity based costing expanded the understanding of cost drivers rather than allocations and distributions.

We very briefly touched on maintenance above. As maintenance processes have evolved from breakdown to reliability we have observed a greater need for data. It is a source of concern to us that in a large number of situations there has been a separation in individual systems of downtime and uptime. The inevitable consequence is that rarely do they add up to 24. Surely our systems can support the one physical piece of equipment and its operator in a better way than that. We may like our maintenance and production personnel to cooperate in maximising the throughput of our plant, we suspect having independent systems inhibits the desired behaviour.

Although Epicycles hasn't yet addressed the maintenance and production systems issues in depth rest assured the facility exists for a single picture of machine time and machine state.

One may ask why a computer based system may be required to support lean manufacturing. It occurs to us that while lean manufacturing provides a very visual indication of materials requirements at an instant in time on the shop floor it has little forward planning visible. The "trick" is therefore to provide a lean planning system which is capable of providing just the minimum level of planning to more easily accommodate business change.

We believe that developing system concepts which are capable of wider use within the product process matrix will remove some of the barriers to continuous business improvement and evolution.

We have described another consistent aspect of "Lean. Quiet. Seamless".