LEAN
https://www.twi-global.com/technical-knowledge/faqs/faq-what-is-lean-manufacturing

LEAN

Lean essentially means without any excess. Without any ‘waste’. Waste of material, waste of effort or waste of time, all eradicated. Lean sees anything that is not value-adding as waste. Anything that we don’t require to move forward, should not be pursued or even entertained. But how do you identify this waste? Well, believe it or not, almost all of this waste comes from our assumptions, and that’s where you start the scrapping.?

First Principle Thinking

Is it actually important or do you think it is important? The key question is “Why do you think this is important?” and if you keep digging with this shovel, sooner or later you will find the treasure. The core assumption on which it all depends, the only truth that matters. And putting efforts and resources into anything else is waste. This questioning and rigorous filtering of waste retains its importance throughout the journey. And a good way to keep away from assumptions is to do more and hypothesize less.

Action-bias

Eric Reis has majorly used the playfield of a 0 to 1 journey. How waste-elimination is critical in product strategy and initial business growth. He sums it up as a Build-Measure-Learn loop, establishing a bias towards action as compared to planning by starting the loop by building.?

The key rationale for this is that you cannot plan what you cannot control. And hence to plan for such things just delays what actually holds importance. Remove this waste and you will find an accelerated value creation. Identify what the 1 important thing is (or what you believe it is) and do it. Hit or miss, doing it will definitely give you learning for moving forward. These learnings add actual value.

Value-Addition

Stagnant waters breed diseases. Progress is a necessity that, contrary to popular belief, is not about just adding things to a basket but about figuring out a basket that everyone desires. Whether it's an app, a good, or even your own portfolio, adding features might look like building but features are not the value; learning is the value. It is pointless to be stuck in adding something which hasn’t been proven to be of importance by the learnings we gain. And hence it is equally important to adapt a system of constant feedback, hypothesis testing and prioritizing the voice of the customer.

6σ - What is it.

6σ (Six Sigma) got formalised as a methodology on the Motorola floors. It is a collection of methodologies and statistical tools that aim to improve quality by reducing errors.?

The guiding concept is pretty simple in essence. Our aim is to optimise our process so that that σ (Sigma) level of the process increases. What is the σ level? Well it is a scale used wherein a process produces only 3.4 defects/errors per million produce/goods/opportunities.

How is this calculated ?

Let's take it step by step.

Disclaimer: Please read about Normal Distribution and Central Limit Theorem before diving in.

Let's say a factory produces 10,000 nails a day which have to be 10 cm long. The Lower Specification Limit (LSL) and Upper Specification Limit (USL) of 0.6. Which means customers accept our nails if they are between 9.4 to 10.6 cm long. We will say our ‘population’ is 10,000 nails.

We can't measure all 10,000 nails to check how many are acceptable. So we take out samples of say 100 nails and check their mean. We do this process for some time and plot the statistical distribution for these means.?

Now the central limit theorem dictates that this distribution will approximate to a normal distribution whose mean and standard deviation can be used to approximate the population characteristics.?

So let’s say the mean comes out 10 cm and standard deviation (or σ) as 0.3 cm

Now as per the LSL and USL, we have a window of 10 - 9.4 = 10.6 - 10 = 0.6 cm. (If both windows are not equal, we take the smaller one.)

Or, we can say a window of 0.6/0.3 sigmas i.e. 2 sigmas.

This dictates that following a normal distribution, a span of 4 sigmas (2 on both sides) will be compliant to customer requirements, i.e. approximately 95%

https://en.wikipedia.org/wiki/Normal_distribution#/media/File:Standard_deviation_diagram_micro.svg


We can say that the process is 2σ optimised.?

To bring this to , we aim to reduce the standard deviation to 0.1 which will make it 0.6/0.1 = optimised. And in turn will mean approximately 99.99% of our yield will be of quality and use.

Now six sigma also take in the long term effect by introducing the 1.5 sigma multiplier and the final values of DPMO (defects per million opportunities) corresponding to the sigma level can be found here -?

https://6sigma.com/sigma-level-table-its-all-about-quality/


DMAIC / DMADV

Now comes the hard part. Reducing the standard deviation. The Lean Six Sigma methodology suggests following a cycle of DMAIC or DMADV which is - Define, Measure, Analyse, Improve/Design and Control/Validate. Depending on the product, either can be used and both share the same backbone underneath.

Define phase is wherein the main objective is to formulate a good problem statement. A good problem statement contains the metrics you want to chase, the timeline and the scale of the problem in a way that is understandable and non-disputable for all stakeholders. Also read: SIPOC Analysis

Measure phase begins with collecting the information you need. Talk to customers, get production data and anything and everything you need. Get now, throw later. At this stage it is highly important to use the correct tools to get actionable insights from this data. It is important to carefully to convert the customer VOC to Critical factors which affect business.?

Analyse phase is where you start building a story. The transition from Measure to Analyze is not that demarcated as you most likely will still continue collecting data while parallaly trying to make sense of all of it. Figuring out correlations & causality is our aim. So that we know what to improve. Where to work ?

Improve phase is the action phase. From a choice of a number of methodologies and tools, use the one that suits your case. Tinker out the flaw, build that feature, you know what it takes, now do it.?

Control phase is your failsafe. You’ve done so much work to make improvements you wouldn’t want it to go waste. This is where you put in measures to protect and continue with the improvements you’ve made.?

Tools -?

Design of Experiments (DOE)

Poka Yoke

Statistical Process Control Charts


要查看或添加评论,请登录

Kushagra Ahuja的更多文章

  • Strategic Management

    Strategic Management

    In this article I have tried to put forward a consolidated view of my understanding of strategic management that I have…

社区洞察