AAOM Handbook

ST.04 Calibrate Performance Model

Context

The stakeholders in a business (shareholders, employees, community and government) collectively define the expectations that the business must meet in order for them to continue to support its operation. These expectations typically encompass safety, environmental, social and economic dimensions and translate into three distinct characteristic. These are; • Effectiveness (units output per time period), • Efficiency (a ratio between resources used and units of output) • Sustainability (the material and social requirements for ongoing effectiveness). The performance of a process over any period of time is not a single value, but rather it is a range of values with differing probabilities of occurrence (i.e. as can be represented in a capability histogram). In addition to the obvious constraints that exist when there are elements of the Business Structure that have capacity limitations, the variation that occurs in elements of the Business can interact to produce additional significant constraints to the throughput of the process. To effectively Set Performance Targets we need to be able to understand the constraints and opportunities within a process, and hence how the capacities and random variation within a process interact. This requires a statistical model of the process, i.e. the Business Structure. A statistical model will only be helpful in understanding process performance and setting performance targets if it is calibrated so that it produces reliable forecasts of performance. This can be achieved by adjusting the model parameters so that when historical data is input to the model, the outputs of the model produce a match to the historical outputs. The correct calibration of the model requires consideration of: • The logic/mathematics of the model. This can change if the process design or operating (production and service) strategy is varied, and hence may require adjustment if either of these changes occur. • The parameters that describe the probability distributions for the performance of each element of the model. The minimum, maximum, most likely and mean values, plus the shape factor of the distribution, may require adjustment to account for; o the degree of data randomness (independence), o a shift in process operating point on the performance/effort curve, and/or o the size of the historical data set. • Non-statistical input parameters (e.g. maximum levels of stockpile or other physical parameters)

Purpose

To calibrate the Business Structure Performance Model.

© McAlear Management Consultants 2006

Operational Planning: Set Performance Targets

Updated: August 2018

Page 28