Six Sigma concepts into Lean 8 Wastes

This is about introducing basic Six Sigma Concepts in the analytics of process wastes in Lean manufacturing. A very interesting integration of two powerful tools to boost productivity in any manufacturing floor, or even service businesses in general.

First, let's clarify some basic concepts before getting into serious matter. The basic knowledge of any lean practitioner is that there are 8 wastes. Others also understand another concept called the "MUs". Muda, Mura, Muri, each respectively meaning engineering waste, process interruptions, and Process or machine stress. All 3 MUs are strongly connected in cascade. Muri, a stress or overburden in the process, shall eventually result into the process breakdown and interruption, Mura, which in turn will cause one ore more of the following wastes, Muda: idle and wait time, in-process inventory (at least where the break occurred), some over processing, and probably defects an rework. However very few connect the dots and know that Toyota basic understanding of waste lies in the 3 MUs. Each MU is a category type waste.

According to Taiichi Ohno, the farther of the Toyota Production System, the unique fantastic goal is the absolute elimination of waste. However we all go wrong about that. A waste assessment will not reduce waste. The basic principle of continuous improvement is that you cannot improve what you do not measure. Therefore waste needs to be measured, not only assessed. And this introduces the whole new concept we are developing here: Lean Waste Analytics.

Lean Waste Analytics is about defining a waste measurement system to support the decision making process for management in a Lean environment. Quantified waste metrics makes it easier to control and manage your lean transformation journey. In order to quantify waste, the organization must first separate what we call “Underlying waste” from “Functional waste”.

Underlying waste are waste that are not visible, mainly passive, caused by poor work organization, poor layout, poor designed work station, inaccurate standards or specifications etc. Functional waste are more active, and result from the bad performance in the work being done, for instance downtime due to machine breakdown, not meeting customer specifications, poor production planning resulting into over production and inventory etc.

The underlying waste measurement should be established for a chosen value stream. The right data collection at a high level become important for an accurate measurement. This data collecting process can be time consuming, therefore we would recommend it to be performed periodically, and avoid large time consumption. The following data are required to quantify the underlying waste:
*CT - Cycle Time of all process steps in the value stream, where Cycle Time is the time needed to finish one process step of a single product unit before initiating the next product unit.
*NVA - time of Non-Value Added steps of each process, where Non-Value Added steps also include pure waste and are considered as Muda
* TT - Takt Time of a value stream, where Takt Time is the available production time divided by average customer demand for the same time interval

Once these data are collected, we may calculate Underlying waste as followed:

MURA =Standard deviation of operators CT within the value stream
                        Average operators CT within the same value stream    

Mura metric represents the variation in workload between operators in a single process (in case more than one operator) or a value stream.

MURI = Absolute value of TT - CT  x 100        
                                      TT

Muri metric represents the potential overload and/or underutilization of employees.

MUDA = time of activities  NVA  x100
                              CT

Muda metric represents the percentage value of non-value added activities performed by each operator.

The following data are required to quantify the functional waste:
* dDL - daily Direct Labor Time of each process, where Direct Labor Time is the actual time spent on value added and non-value added process steps.
* ST – Standard time of each process, where Standard Time is the expected time needed to finish a particular production process.
* AT – daily Available Time of each process, where Available Time is the number of shifts multiplied by the number of hours (by law) in a workday minus planned breaks and downtime.
* DT – daily unplanned Downtime of each process.
* RT – daily Rework time.
* WIP - daily Work-In-Process inventory in a process.
* SWIP – standard daily Work-In-Process in a process.
As noticed the functional active waste are much more complex to quantify than underlying waste

MURA = Standard deviation in daily direct labor time dDL within a week.      X 100
                             Average daily direct labor time dDL within a week  


Mura metric represents the variation in production volume represented by direct labor time. Similarly as for the underlying waste, it is necessary to use the coefficient of variation formula.

MURI = Absolute value of AT - ST  x 100        
                            AT

Muri metric represents the actual overload or underutilization of employees on each process.

As for MUDA in the functional waste analytics, It must be decided which MUDA is being quantified, if not all, since each type have a slightly different way of calculating. The functional MUDA types can be:

1. Flow MUDA.  = |SWIP-WIP| x100
                                SWIP

2. Utilization MUDA.  = (DT/AT) x 100

3. Time performance MUDA  = AT - dDL.  X 100
                                                  AT

4. Efficiency MUDA  = dDL-ST  x 100
                                     dDL

5. Rework MUDA = (RT/AT) x 100


We recommend that MURA is calculated on a weekly basis, while it is possible and strongly recommended to calculate MURI and MUDA on a daily basis. Once waste is measured and kept as a Lean Manufacturing KPI metric, the organization will have a better understanding of the origin of the waste (the data collected speak for itself) and thus should be more capable of working towards its reduction and/or elimination. In fact elimination is not possible as one may think because of the fundamental concept of Standard deviation itself.

Understanding the formula for each category type, it is now more evident why a Muri will eventually generate a Mura which in turns cause one or more Muda, as we established the strong connection between them at the beginning.