Let’s start working with Six Sigma & what we refer to as the breakthrough equation.
Y = f(X) + E
In the above equation: Y = the desired results or outcome; X = input factor(s); F = function (the way the inputs are transformed); E = the presence of error or uncertainty;
So, let’s take the example of baking bread.
Bread = Mixing + Baking (Flour, Yeast, Salt, 300 Celsius) + Random Error Rate
We can assume that we can create the desired outcome by configuring and controlling inputs in a specific manner. We can analyse every output of a process and determine the input, processes and error behind it. Once you know the root cause of the issue, you can control it next time. For example, variability in bread may be down to humidity in the room, which may be caused by the time of day and the month in which you bake it. So, we can put measures in place to control the humidity in the room. Before you confirm the link between humidity and your output variability, we need to first rule out other influences – once we understand variation, we can control it and minimize its impact.
So, controlling variation is the name of the game. We have natural / common variation, which is stuff that is difficult to control. So, think about wind on a wind farm, we can’t control how often and how hard the wind blows. We then also have special cause variation which we can control – things like network connectivity issues delaying a process.
We should always seek to fix the special cause first. In the above example, we can fix the humidity issue by utilizing a dehumidifier or other technology.
We must always assess variation & categorise it before taking any action. It can be short term variation, which is where the variation is within certain tolerances and doesn’t impact the process or long term variation, which is outside of the tolerances and does impact the process. We don’t need to remove variation that doesn’t impact us, so let’s focus on the long-term variables.
Now, in order to improve our processes it’s vital that we can measure our X’s and our Y’s. We should always start with the Y’s. Let’s look at an example. If you’re goal is to have £10,000 in your savings account, you can measure this easily by logging on to your bank and checking whether you’ve hit your goals. When the answer is no, we can start measuring our X’s. In this case, the X’s are income and outgoings. We can then work on a plan to either increase income or reduce outgoings to meet our goal output (Y).
The above example is very simplistic and isn’t the reality of most processes. Most processes have a degree of qualitative data which is hard to measure. However, we must put a value against our qualitative data. For example, customer satisfaction can be quantified through the use of surveys. Once we have our data for a subset of customers, we can apply a general rule that says ‘where X’s and E’s meet this condition, the customer satisfaction level = 5, by doing this, we can analyse the cause & effect of the resulting Y.
As part of this measurement process, we need to understand the leverage of each of our X’s. That is to say, how influential is the X on the output Y? Pareto law suggests that 20% of inputs account for 80% of the impact on an output. With that in mind, we may say that when baking a cake, the flour and sugar has higher leverage than the sprinkles as it has a bigger impact on the quality of Y. In most cases, only a few factors will carry the weight to significantly impact Y (that is to say, a few key X’s cause performance problems and defects with Y).
So our Job as Six Sigma analysts is to identify the most powerful variables, those that impact Y most heavily & focus our efforts on reducing their variability.
The Six Sigma framework we use throughout improvement projects is DMAIC.
- Define: create a problem statement, objective statement and key metrics
- Measure: Understand the current process, validate data accuracy
- Analyse: Determine the relationship of Y=f(X)+E and narrow down the key variables (those with most leverage)
- Improve: Develop, test and validate changes
- Control: Monitor performance and ensure that changes are sustainable
Six Sigma can be applied to 4 main domains:
- Thinking for breakthrough
- Processing for breakthrough
- Designing for breakthrough
- Managing for breakthrough
Thinking for breakthrough is where we create a set of rules / guidelines to fuel the culture for change in an organization. Getting the masses speaking the Six Sigma language for change and creating an air of enthusiasm will help the project succeed.
Processing for breakthrough focuses on optimizing performance of existing business or operational processes to bring them to their entitlement level – the best case performance that can be achieved by the process if all the variables align nicely. As Six Sigma analysts, we should be working to ensure that the process performs at its entitlement level all of the time, without overhauling the design.
Designing for breakthrough is about optimizing the design process using DFSS (Design For Six Sigma). This is an approach for planning, configuring, qualifying and launching products, services, processes and systems to move quality upstream.
Moving quality upstream is the process of anticipating manufacturing issues before they occur. We plan appropriately to ensure we avoid those issues from happening. Designing this way avoids a lot of the costs associated with variability.
Finally, managing for breakthrough is about selecting and training the right people, installing the Six Sigma infrastructure across the business and establishing management practices for Six Sigma.
Implementing Six Sigma
Six Sigma must be implemented in a top-down fashion to be truly successful. The senior leadership team should have their full weight behind the initiative in order to make the culture shift required within the business to adopt the new way of thinking.
The core Six Sigma team should be comprised of: executives, a Six Sigma deployment leader, business unit leaders and senior representatives from finance, training, IT, HR and communications.
This team should:
- Ensure that Six Sigma is made to be a highly visible initiative.
- Install some sort of measurement system to enable progress to be tracked.
- Benchmark the company products / services against the competitors to show the true market position.
- Set stretch goals for the business to ensure large performance gains. Promoting change vs tweaking processes.
- Educate all levels of people within the organization.
- Evangelize Six Sigma success stories from other organizations.
The Six Sigma Lifecycle
The Six Sigma Lifecycle is:
- Initialize: establish the required infrastructure and goals for Six Sigma.
- Deploy: train and equip staff to work effectively in the Six Sigma environment.
- Implement – make changes to improve performance.
- Expand – implement Six Sigma in additional business units.
- Sustain – re-train staff regularly and evolve the use of Six Sigma in the business.
Content based on study of the Six Sigma Black Belt course and Six Sigma for Dummies