Six Sigma. Design for Six Sigma. Total Quality Management. Statistical Methods. In today’s competitive business environment, quality is more important than ever. Enter @RISK, the perfect companion to any Six Sigma or quality professional. This powerful solution allows you to quickly analyse the effect of variation within processes and designs.

@RISK enhances any Microsoft Excel spreadsheet model by adding the power of Monte Carlo simulation. Monte Carlo simulation is a technique which examines the range of possible outcomes in a given situation and tells you how likely different scenarios are to occur. Six Sigma is defined as identifying – and minimising - the percentage occurrence of errors in a process. @RISK is perfectly suited to do just that.

Use the @RISK functions to customise your models with the performance measures that are suited for your project.

Applications of @RISK & RISKOptimizer in Six Sigma
DFSS/ Statistical Tolerance Analysis
Process characterisation; add variation to experimental design models
Project selection with risk and uncertainty
Improve manufacturing, queuing, or customer service processes
Refine and optimise inventory systems to minimise costs
Extend your six sigma programme to business processes
Experiment with different strategies instantly, saving time and cost
Cost estimation of new, untested projects or products
Reliability and failure analysis
Service quality analysis
Resource allocation optimisation

@RISK and Six Sigma: Step-by-Step

A Six Sigma analysis with @RISK would likely consist of four basic steps:

1. Define your model in Excel. Outline the structure of the problem you are trying to solve in a spreadsheet format. You might be an airline identifying the number of bags of luggage lost in year, an auto maker seeking the frequency of failure of a particular part, or an electronics manufacturer looking for circuit boards that exceed given tolerances. Whatever your business, you can describe your quality issue in Excel. Not sure how to structure your model? Palisade consulting services can help.

2. Replace uncertain factors in your model with probability distribution functions. Uncertainty is everywhere, from weather delays of airline flights to material costs in manufacturing. Most companies use an oversimplified “worst case, best case, most likely” approach to uncertain factors. Many use only a single-point estimate! These approaches ignore the entire range of possible values these uncertain factors could take. Using historical data or expert judgment, you can define probability distribution functions in @RISK to more accurately describe uncertainty in your model. Not sure which distribution functions to pick? Palisade on-site training and consulting can help.

3. Identify your bottom line, and simulate. Choose the cells you are interested in tracking – your output cell. This could be the number of service complaints in a year, a target dimension, or the number of defective parts per batch. Then run your simulation. @RISK will recalculate the spreadsheet model thousands of times – each time choosing values at random from your input distributions and recording the resulting outcome. The result – a look at all possible outcomes, and their probabilities of occurring. @RISK results are exactly what Six Sigma analysts need to improve the quality of their products and services!

4. Examine your results and make a decision. In a single mouse click you can have graphs of your bottom line cells that tell you not only what could happen, but how likely it is to happen. A wide variety of graphs and statistical reports – as well as all the data from your simulation – are available. And, @RISK offers sensitivity and scenario analysis that identifies which factors – or combinations of related factors – contribute to the results you see. This information is crucial for focusing your efforts in the correct area.

The Next Step: Add Optimisation

Use RISKOptimizer with @RISK

The Industrial version of @RISK includes RISKOptimizer, the cutting-edge simulation-optimisation tool for Excel. RISKOptimizer is a genetic algorithm-based optimisation tool that finds the best solution to complex, nonlinear real-life problems. Examples include:
Optimise your process settings to maximise yield or minimise cost
Optimise your tolerance allocation to maximise quality levels
Optimise your staffing schedules to maximise service levels

RISKOptimizer’s genetic algorithm-based engine finds solutions other optimisers would miss – such as Excel’s Solver. And it finds them faster than other non-linear optimisers on the market. But what sets RISKOptimizer even further apart from the pack is the Monte Carlo simulation engine that is also included. RISKOptimizer supports @RISK functions in its models, allowing you to identify and account for uncertainty in your optimisation problems. You can use your existing @RISK models with RISKOptimizer!

Example: Optimising Resource Allocation

Say you’ve modeled your manufacturing process using @RISK and determined that your yield is 99% (10,000 PPM defective), and your company goal is 99.99% (100 PPM defective). You can use the @RISK sensitivity analysis tool to identify the parameters or components that are the largest contributors to the variation driving your quality yields, and subsequently use RISKOptimizer to optimise your nominal component values or process settings in order to maximise yield, minimise percent defective, or minimise cost. For example, sensitivity analysis may have shown that key drivers leading to these defects include the amounts of Materials A-D that are combined to produce the product. You can take your existing model and tell RISKOptimizer to vary the amounts of Materials A-D that are used to minimise the percentage of defects occurring. Set constraints on the quantities of each material available as appropriate, and even set a constraint on the yield that meets your company’s target (i.e. not below 99.99%). Then let RISKOptimizer find the optimal allocation of materials to achieve your quality goal!

Features of @RISK & RISKOptimizer for Six Sigma
Easy, accurate definition of variation using 38 probability distribution functions
Sensitivity Analysis to identify key factors which drive variation and uncertainty
Distribution fitting to allow you to quickly define your data set
Scenario Analysis determines which combinations of input variables led to different outcomes
Fastest Monte Carlo simulation engine on the market saves valuable time
Ability to use multiple CPUs in a single machine to speed up large simulations
Correlation of uncertain inputs to reflect real-life dependencies between elements
Risk analysis to determine the extent of quality issues and identify the key drivers
Optimisation to generate a viable solution to meet your goals
Seamless integration of risk analysis and optimisation lets you perform multiple analyses on the same models

Model: Electrical Circuit Analysis
     See a working model! Electrical engineers use @RISK to model an electrical circuit and simulate performance.

Model: Design of Experiments
     See how @RISK can be used to ensure the quality of experimental design.

Model: Design of Experiments with Optimisation
     See how @RISK simulation and RISKOptimizer optimisation ensure the quality of experimental design.

Case Study: DecisionTools Safeguards Precious Metal Refiner
     Met-Mex Peñoles simulates manufacturing processes to conserve materials and reduce errors.

More About @RISK, The World's Leading Tool for Risk Analysis