Home > Forums
Topic Title: OEE
Topic Summary: Calculating
Created On: 04/29/2013 06:34 PM
Linear : Threading
Send to a Friend Send to a Friend
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
04/30/2013 10:41 AM
Print this message

Author Icon
JamesCM
Cindy James



Currently we do not measure the "off quality' production. The off quality just gets hauled away and mixed in with the off quality from all of the other lines and then gets recycled to make more product. We do measure uptime and speed. We also track the quantity of saleable product off of the line.

To get started - I am proposing that we calculate the potential production at 100% OEE.......and then divide our actual production by the potential. This would give us an OEE. We could then back into a calculated amount for offquality/waste. This would help us see where our biggest opportunities are (out of 100+ lines)

We are working on getting systems in place to weigh the offquality - but also difficult because the unit of measure off the end of the line is cases yet we throw away packages and/or individual units.

Just wanting to be sure that I am not missing anything here.
04/30/2013 02:02 PM
Print this message

Author Icon
duecesevenOS
Kris Hallan



I'm not quite sure what you mean by "off quality". It sounds like it is bad product or product that falls out of the system that can be reworked by running through the process again.

This is a great and blaring example of why you should not EVER use OEE on anything other than the bottleneck of the value stream. You should not be using OEE as a way of looking for your "biggest opportunities". It's only (very limited) use is to focus value stream wide attention on your already identified "biggest opertunity" (aka the bottleneck).

If the process you are measuring is (as advised) a bottleneck, then the consequence for making bad "saleable product", losing speed on equipment, losing time on the equipment, and "off quality" is all the same. If it's the bottleneck, the result of all of these losses is lost sales to your end customer. This is true because there is no making up for lost production at the bottleneck. If you are making a part every minute on your bottleneck then every minute wasted is a lost part, regardless of how you waste it.

If the process you are measuring is not the bottleneck (they can't all be...), then how you should handle "off quality" wil vary dramatically depending on your process. If, for instance, this process has 500% of the capacity of your bottleneck and you are able to rework the product without adding any labor, then "off quality" isn't really a big deal. All instability will force you to add inventory and creates waste, but if your factory is full of instability, this one isn't your top priority. If on the other hand, your "off quality" forces you to do labor intensive rework and forces you to make huge batches that get run on overtime and completely messes up FIFO and part identification, then you got a real problem. So coming up with a single metric that is going to somehow tell you what your opertunity is by comparing areas is a waste of time. Spend the time on the floor and do your value stream mapping.

Overall, OEE is pretty much useless. I can never understand why it is the topic of conversation on so many supposed lean blogs and forums. It has got to be the biggest ddistraction from making a lean conversion that could possibly be out there.
04/30/2013 02:02 PM
Print this message

Author Icon
KeithL
Keith Lodahl



What are you hoping to gain? The final number has little value to me. Collecting the data that goes into the calculation tells me what is keeping the process from being effective. That is where OEE adds value. The team can focus on the areas that are keeping the process down and fix them.

I would use time instead of quantity to determine OEE. How much time is used making off quality?
04/30/2013 03:30 PM
Print this message

Author Icon
Boeing_Lean
Ken Hunt



Kris,

I agree with you 100%, especially your last paragraph. OEE is but one area that people tend to focus on when they shouldn't. Sometimes it seems like people hear about or see some tool and if they see it referred to often enough, they say "We gotta go do that", without understanding what it may or may not do for them. I'm not implying that this thread is a result of that, because I am not involved in this particular situuation.

At the risk of being repetitive, do the VSM first. This will make the bottlenecks obvious and will provide focus on what constraints are in most in need of eliminating. Then MAYBE OEE can come into play in some cases but I bet there are other measures that will tell the story, like creating a Future State Map and laying out a plan to make it a reality. That said, NOTHING can take the place of a Gemba walk to truly understand what is happening, and finding ways to make improvements.

Ken
05/06/2013 06:07 PM
Print this message

Author Icon
LION
Emmanuel Jallas



Cindy,

Here's one former reply to the same question in an other thread about OEE.

Hi Pierre, You are asking for a book about OEE. Kaizen Express from John Shook will tell you that OEE is a tool to look back in a mirror, an accountant's tool, not an operational tool.
John will explain you how to use operational availability to drive in real time your processes and how to fight the 6 major losses.
Bonus. You'll learn a lot more on attitudes, methodologies, tools in this book.
Enjoy.
Emmanuel

As John states in Kaizen Express p62 - 64,
OEE = Availibility rate x Performance Rate x Quality rate

Availibility rate = time the equipment actually run / time the equipment was supposed to run

So, by definition you know availibility rate, AFTER the shift, and it is a mean measure. It can be analogized to a mean speed of a car between two cities. You know it after your trip. And you can't do anything about what happened. You can report it, as accountants report books. It tells you the story when it's done, hiding somme "unimportant" facts.

Another flaw is that "the time the equipment was supposed to run" contains production non balanced by orders. Tipically management ask to USE the equipment to lower costs of production. The more the equipment produces during an amount of time, the lower the the price. Thus it contains itself the waste of overproduction.

Operational availibility depicts how you are using your equipment right now. It is your speed indicator.

Imagine you are on an interstate highway, allowed speed is 70 Mph, you look at your speed indicator, it tells you 30 Mph. Within seconds you're going to react, and find what is going wrong.

Operational availibility contains only the production balanced by orders = required quantity,

so it avoids the waste of overproduction. I'm sure John will tell us much more about ka dou ritsu in its "make money" meaning. Emmanuel


Hope it helps

Emmanuel
05/07/2013 02:43 PM
Print this message

Author Icon
Jeroen2011
Jeroen van Deursen



Hi Cindy,

I'm in doubt...

In general I agree that OEE is not the best of indicators. As the others write it can be a usefull indicator for your bottle neck, but even then it is not the holy grail.

On the other hand, at this moment you don't have any measure for 'off quality' and what you want to do is an improvement.

for the measurement, may be it is possible to provide a data sheet to convert packages and cases to individual units..

My advise, continue, evaluate and change/improve where necessary
05/07/2013 02:43 PM
Print this message

Author Icon
KeithL
Keith Lodahl



As I said on another thread about OEE, the final number is not one to run the plant by. I have found OEE important because it gives the operators a clear picture of where the Muda is in their processes. I won't use any of the computer programs that automatically catch the downtime. I want folks in Gemba tracking the downtime and understanding what it is. The individual losses are data that goes into planning countermeasures and getting the machine up and producing. I have heard the argument that more can be learned by going to Gemba, but frequently people living in Gemba have become so accustomed to poor performance that they no longer notice it, or realize things could get better. Seeing and trackng the data is a good way to point it out.
05/08/2013 04:13 PM
Print this message

Author Icon
Mike_Thelen
Michael Thelen



We never used OEE as a metric. We used OEE as an evaluation tool to help understand what we needed to do:

When someone said they needed more equipment to meet product demands, we conducted OEE assessments. Suddenly, you would find all sorts of opportunity to increase throughput at the process to meet those demands.

When someone said they couldn't "do flow manufacturing" because they didn't have capacity, we would to an OEE assessment. Suddenly, you would find all sorts of opportunity to meet flow demands.

OEE is a tool to assist with the rules in use - 1) highly defined work, 2) simple and direct flow 3) clear and binary connections and 4) continuous improvement to the ideal state.

Why use OEE to measure quality? Right-First-Time or Perfect-Order-Index or First-Pass-Yield are really more well-established quality measures. You could simply measure by tick sheet the number of products that flow through a workstation with no errors and the number that are rejected or reworked. There's your quality metric, driven at the station, by the people.

OEE itself is not a useless measurement, it just has to be used correctly. Too many lean leaders/consultants use it incorrectly and have no idea how (AND WHEN) it should be used. It is EQUIPMENT EFFECTIVENESS - not a quality rating. If your value stream map shows that you have flow issues out of a specific area/process/machine, then you use OEE to determine if you need better utilization, better machines, or more machines. A machine running 30% OEE that is moved into a line demanding throughput from it at 85% OEE isn't going to make the line run correctly. You have to either improve the OEE or select different equipment.

In lean worlds, the line is set up to maximize flow through the entire value stream. Having unreliable equipment put in the line will doom the line - that reliability is the only way to achieve reduced inventories, reduced lead times and improved quality. Thus, the importance of (and the how/why to use) OEE.
05/09/2013 11:13 AM
Print this message

Author Icon
Robert_ELSE_Inc
Robert Drescher



Hi Cindy

OEE is another of those metrics that try to combine multiple areas into one metric. The only real place it serves any real value is in very high volume, low mix (generally only one product per production line for long periods of time) and very stable processes that have good quality to begin with. There it can help draw attention to developing issues in equipment wear and tear.

For most organizations it is often far more beneficial to analyze and understand the three separate variables that make up OEE. Especially in your case where you have so many lines, it is very likely that lines with identical OEE could actually not share any similar problems, that is why the underlying information is of much greater value, it will help identify those issues you can tackle the cheapest and fastest first, while also allowing you to plan for the other more complicated fixes.

As far as determining your line quality you could do it backwards, but if the lines are multi-step and multi-machine line quality still won't tell you which step or machine is the cause. A simple visual record at the individual machines where defects are removed would help more than even knowing the exact numbers for the whole line.

Good Luck
Robert Drescher
ELSE Inc.
05/09/2013 03:54 PM
Print this message

Author Icon
duecesevenOS
Kris Hallan



Mike,
It doesn't sound to me like you are promoting the use of actual OEE. You say to leave quality out of the equation so you are actually talking about some other measure. The equation to calculate OEE has three parts to it: Availability, Speed, Quality. If you aren't using the three parts than you are not using OEE. You are promoting the use of some other metric. It sounds to me like you are actually agreeing with me that OEE IS useless but another metric that ignores the quality portion of OEE can be useful in a few instances.

I agree with you. My plant actually just measures the availability and speed legs of OEE and we call it Operational Availability (OA). It's basically Actual Production / Scheduled Production on a given piece of equipment. OA is an interesting measure for point analysis like you said. When someone says that they are the bottleneck or that they are capacity constrained, an interesting starting point is to balance out their actual capacity and show the percentage of the time that they run to that capacity. It can be a good measure of the opertunity available before implementing good systems of TPM.

OEE as calculated according to it's definition is pretty much useless.
05/10/2013 03:07 PM
Print this message

Author Icon
Mike_Thelen
Michael Thelen



Hi Kris!

How are things? It's been a while since we were in the same discussion!

Sorry, I wasn't very clear. We don't (didn't) leave Quality out of the OEE calculation. I just think CIndy is trying to measure Quality via OEE. OEE isn't the most effective, or easiest, measurement of Quality, from what I understand she is trying to determine. Why use OEE, when a simple tick sheet provides the data. OEE is about a piece of equipment's overall performance ability to meet customer demand, not providing merely an 'in-spec' piece of work.

We did use OEE, including the Quality component, but not on a regular, defined basis. We only used it when we were evaluating a problem determined through a conflict with one of the rules in use. If your machine is producing the right number of parts, but not producing them in quality, it leads you into TPM processes such as tool life, centerlining, and condition of the machine/tooling. My point was that it ISN'T what I would use to measure QUALITY, but rather EQUIPMENT ABILITY (I really wish we could bold or italicize - I hate CAPS!) If the OEE is low, based on the Quality component, you need to upgrade or problem solve the equipment to get an acceptable quality level. However, if the OEE is low due to an operator always wondering off and the machine is sitting idle - it doesn't determine your quality problem.

Why you use OEE is to force equal consideration between: is it available (Availability)? Is it optimized (Speed)? and is it capable (Quality)? Otherwise, all we worry about is 'making parts' or 'making parts fast', not 'making the best parts quickly and efficiently'.

HOWEVER, like you, we use(d) OA far more. It was an effective way to reduce lost time (or potential available time) on a piece of equipment. All you are looking at cycle-time vs. out-of-cycle-time. When a machine isn't turning, it isn't providing value (unless demand dictates that it not run!) This drives you to look at why the machine is idle for 5 minutes out of every 20 minutes (often load/unload time, QA checks on the parts, clarification on specs, etc). But, that is because we measured quality by 'right first time' - meaning a simple tick sheet that showed how many parts flowed through without requiring rework of any kind.

Have a good weekend everyone!
09/18/2019 08:25 AM
Print this message

Author Icon
jaidynmoore
jaidyn moore



Hi, Check out the following link, it will definitely help many (as it has also helped me to track OEE and Downtime in my yogurt factory). https://www.downtimecollectionsolutions.com/thrive-oee-lp
FORUMS : Manufacturing : OEE

Topic Tools Topic Tools
Note: These forums are moderated by the Lean Enterprise Institute. All posts are reviewed within 24-48 hours prior to appearing on the site. Views expressed in these forums do not necessarily represent the views of the Lean Enterprise Institute.