AAOM Handbook

[OFFICIAL]

INTERNAL USE ONLY

2. MODELLING AND STATISTICAL TECHNIQUES 2.1 Software used to build BSP model It is important to note that a cloud-based Digital Operational Planning platform has now (as of 2023) been adopted across 8 Operations for BSP modelling. Microsoft Excel and @Risk Microsoft Excel and @Risk are the software tools that have been used to develop the Business Structure Performance models since the start of the Operating Model in 2014. @RISK performs risk analysis using Monte Carlo simulationto show possible outcomes and probability of those outcomes. @Risk is an add-in of Microsoft Excel and is very easy to use in the Excel environment. 2.2 Statistical Modelling Monte Carlo simulation is a computerized mathematical technique that allows for variability and risk to be accounted for in a range of possible outcomes and probabilities. Monte Carlo simulation performs risk analysis by building models of possible results by substituting a range of values from a defined probability distribution for any output measure that has inherent variability. It then calculates results over and over, each time using a different set of random values from the probability distributions but with the value selection frequency weighted by the distribution shape. Probability distributions are a realistic way of describing variability in output measures. @Risk has the capability of defining best fit for continuous distributions and associated parameters fromdata as well as discreet distributions. There are a range of distributions in @Risk. Theaccountable roleholder must either fit a distribution using @Risk over the historical data thatthe roleholder has selected or use discretion in selecting an appropriate distribution according to the historical data. The recommended continuous distribution to use over a representativeprocess stable period is a PERT distribution. • PERT- defines the minimum, most likely, and maximum values, just like the triangular distribution. A four parameter PERT considers the spread. Values around the mean/mode are most likely to occur. An example of the use of a PERT distribution is to describe process throughput rate or duration of atask in a maintenance service. • Binomial – describes probability of discreet random events to occur. During a Monte Carlo simulation, values are sampled at random from the input probability distributions (this is random within the boundaries of the input parameters but with the value selection frequency weighted by the distribution shape, different scenarios with different input parameters can be tested as required in PS.08 for combined strategy effects). Each set of samples is called an iteration, and the resulting outcome from an iteration is recorded. Monte Carlo simulation does these hundreds, or thousands of times asspecified to give possible outcomes.

Operational Planning: Building a Business Structure Performance Model Page 6 of 39