Productivity Metrics

How does one judge the productivity value of an AMHS? Traditionally, AMHS was purchased simply to comply with threshold values of moves/hour. After fab layout, the required moves/hour were computed between tools, and the AMHS system selected was to have capacity to deliver this requirement at a minimum cost of ownership. Cost of ownership excluded productivity considerations which the AMHS could impart to the fab, because it viewed the AMHS as a passive device. This was contrary to facts. Each factory is characterized by its operating characteristic. The curve is generally derived through experience and it takes into account the aggregate variability of tools and the variability of WIP delivery logistics. Variability is the coefficient of the OC curve. Thus, dispatch systems play an important role, contributing to this coefficient through reducing the variability of WIP arrival rates. It is universally recognized that more efficient dispatch can be effected with a conveyor network infrastructure (than with discrete delivery technology - manual or vehicle based), because of the conveyors superior capacity and immediate availability to transfer.

It is thus accepted, that productivity metrics of an AMHS should include its ability to contribute to fab efficiency (i.e. to fab capacity through the utilization of its tools, or its dependent variable, cycle time). Because, in this most important fact are the applications of conveyor networks based. When used, a fab can deliver a higher throughput with the same tool set, as opposed to discrete transport methods. Or conversely, deliver a shorter cycle time at a fixed throughput rate.

Current experience with discrete vehicle transport systems in the 300 mm environment show some capacity constrains. This dictates the use of large transport lots which by themselves are a hindrance to efficient fab operations (throughput or cycle time). This fact and commercial requirements for small lot production aggravate problems with discrete vehicle technologies. As we increase the wip moves, the variability of WIP arrivals is pushed up, unnoticed, before we even reach the point where some transport capacity constraints might show. A silent breaking force on fab capacity and cycle time will operate. To show this, simple fab simulations are employed (not transport system simulations). Only through this true metric, can transport technologies be judged for their active role in fab efficiency, and thus their real cost of ownership.

An important focus today is the capability of small lot production. At the time transition to small lot production occurs, the conveyor transport network will need no expansion. The above arguments illustrate this fact.