Posts Tagged ‘machine truth’

shopTALK – Episode #2: OEE for Continuous Improvement

 

 
 

QUESTION: How do you use OEE as a measurement for Continuous Improvement?

ANSWER: A lot of people are using OEE, overall equipment effectiveness is what it stands for. However, you can calculate it without having a tool like Plantnode, you can take just the total number of units that have been produced, divide it by the amount of time you’re planning to run and the expected rate the machine should be running at so you can get an OEE number. Though, it’s really a target, it’s something that you’re trying to hit, and it’s a percentage, a percentage of the expected time you were supposed to be running, and a percentage of the run speed rate that you’ve entered.

After all that’s been calculated the real benefit of OEE is to understand who can I make my OEE better? How can I improve? Ultimately OEE is an efficiency measure it’s a way to measure how am I doing against where I expect to be, our best customers actually move the bar frequently throughout their lifecycle of adoption of OEE so today they might be at 95% OEE and they might change the expected rate to be higher than what it is today so their OEE drops the next day their 75%. So the OEE of “World Class 85%” or whatever the world class happens to be for your industry is merely a measurement point it’s not an end goal. You can improve and improve on where you are, you can always get better, and the way you get better is by making your standards harder and harder to achieve.

What a lot of our customers that use OEE really like about the Plantnode tool is that it provides more information and a breakdown of where your losses are coming from. OEE as a number is really not that meaningful unless you understand the elements that make it up. There are those three elements of: Availability – which is your uptime percentage, your Performance – which is your run speed as a percentage of the target you are trying to hit. And your Quality ratio – Which is your good product divided by the total product, or a measure of scrap or first pass at yield. What’s interesting is, when you look at the individual components: they are mutually exclusive and collectively exhaustive, you can’t hide one really good result with another really poor one, or cover up a poor result with a really good one because the OEE will still be reflected when you multiply the two.

Based on our customers experience using OEE, the value that they get out of our tool is that they’re able to use it to diagnose why their OEE gets low or why their OEE is where it is currently, they can also universally apply that same standard across all their plants in their entire enterprise. So we can help them diagnose the issue: if the OEE is low, is it because of a run speed problem?, is it because you have a lot of downtime?, and if it is downtime, what’s the reason for the downtime?, is it shortstops?, is it changeovers? Is it other unexpected maintenance problems you weren’t planning on happening? Those types of things are what really makes OEE useful is once you understand what the components are that are causing your OEE to not be what you want it to be.

 


 

 

Michael Dedrick

Michael Dedrick is the Marketing Campaign Specialist at Shoplogix Inc.
Follow us on Twitter @shoplogix
Join our Machine Truth Blog on Linkedin



shopTALK – Episode #1: LED Display Boards

QUESTION: How does operator engagement and LED display boards affect an organizations productivity?

LED Display Board

ANSWER: One of our customers was clearly able to articulate how operator engagement and LED display boards were able to help their organization achieve record productivity levels. The customer was a discreet job shop: with assembly cells, fabrication cells, and welding cells. They had been working on self directed high performance teams in attempt to balance the throughput to achieve production shift targets. They also incented the employees for production that exceeded the targets of the overall shift, rather than just the individual cells. The difficulty was they were spending a large amount of supervisory and management effort trying to balance the workload and move resources around from one cell to the other to achieve the production targets.

LED Display Board

When Shoplogix was implemented the LED display boards were clearly able to articulate, as each of the work centres throughput was displayed on the boards. Individuals were able to see if an area was falling behind and required some assistance and were able to redeploy themselves to achieve the targets of that cell and of course maximize the output of the entire shift. They proceeded to set record production throughput levels upon the first shift after implementing Shoplogix.

Michael Dedrick

Michael Dedrick is the Marketing Campaign Specialist at Shoplogix Inc.
Follow us on Twitter @shoplogix
Join our Machine Truth Blog on Linkedin


Machine Truth in Context

In a previous article, we discussed what Machine TruthTM is and its importance. Gathering this data automatically is undoubtedly important and a certain amount of work can be done with this data. But it is just machine data. The next step with Machine TruthTM data is to put this data in context by associating the data with additional non-machine data. The non-machine data to use is really dependent upon the business and strategic goals of the organization. A common set of non-machine data to use includes the customer for whom the job is running, the operator running the machine, and the suppliers of materials. Another important item is the production standards used by the organization.

Why is non-machine data important? Consider a purchasing agent that secures a better price for the paper used in the process. By every measure used by this organization, the purchaser has done a good job in saving the company money. However, at the shop floor, it is found that this paper curls a little bit more under the humidity conditions of the plant and leads to the machine jamming which in turn results in more downtime which leads to lower production and increased costs. By associating Machine TruthTM data with supplier information, the loop between purchasing and production can be closed. This data can give the purchaser the ammunition they will need to go back to the supplier to ask for improvements or to change suppliers all together.

Associating the Machine TruthTM data with the labour used to produce the product is important on several accounts. First it provides us with the production per labour hour metric. It may not always be advantageous to increase production by increasing the labour as the increase in costs will eat into the margin. Secondly, associating production with the labour allows the organization to determine which operators are struggling and in which areas in order to allow the organization to target the appropriate training programs to bring all their operators to the required level.

Finally, let’s consider the association of Machine TruthTM­ data with organizational standards. Among other things, these standards describe how long it should take to set up a machine, the expected run speed, the expected labour hours, and the expected production.

Consider the following diagram. With conventional metrics of uptime and production, Operator B would be considered as having the better day. However, the success of Operator B’s day should really be measured against how well he performed against the standards developed by the organization. If the standards indicate that this operator should have had 70% uptime and produced 700 cases, then he did not have a good day.

Conversely, Operator A had less uptime and produced fewer cases but had to endure more changeovers. If this Operator performed better against the standards provided then he would have had the better day.

The important take away is that Machine TruthTM data by itself may be useful it can only take you so far. It is necessary to put this data into context in order to get the full impact of what this data is telling you.


Building from Machine Truth

Machine TruthTM is very simply the automated gathering of data directly from the factory floor machine in a manner that accurately describes what that machine is doing at that time. Types of Machine TruthTM data include the run speed, production count, scrap count, etc. Machine TruthTM can be gathered from sensors or directly from the PLC. The important thing is that it is automatically gathered, it is accurate and unbiased, and it is the absolute truth of what the machine is doing at all times.

Why is this important? Often times we hear from Plant Managers that they know what is going on at the factory floor because they gather data manually and record it in their ERP system or their Excel Spreadsheet for processing. They feel this gives them the visibility they need. Unfortunately they are wrong. Our experience has been that manually collected data is invariably inaccurate and this inaccuracy is costing the plants real money.

Here is a straight forward example. This diagram shows what was collected by the operator – the blue area is the set up time, the green the time running time and the red the downtime. Now the production manager realized they were not producing what they should have been. His conclusion was that they were not running fast enough. So he asked his team to run the machines faster.

untitled

In this next diagram, the automatically collected data, for the exact same time period, is shown. The difference is immediate. The setup time was longer; there same two major areas of downtime are visible but more importantly there were a number of smaller slivers of downtime that were not recorded. There problem was not run speed, it was downtime. This customer was spending real CI dollars on trying to improve in the wrong area. They were spending money on improving run speed http://www.shoplogix.com/wp-admin/post.php?post=768&action=edit when they should have been focused on downtime. They were chasing CI ghosts.

untitled2

Why was the incorrect data manually recorded? Was it the operator deliberately attempting to protect himself? Possibly, but this hasn’t been our experience. Another plant was in a similar situation. They were packaging gum and they believed they could produce more but were unsure as to why they could not. In a similar manner to the above, they collect data automatically and found they had more downtime than expected – their shift was littered with slivers of downtime. When the operator was approached as to why he did not record these slivers of downtime, he responded it was because the machine jams and it had always jammed and he never recorded it because he thought that was just how the machine operated. When they investigated as to why it was jamming, they determined it was because the gum was too big. The plant manager moved up the line to ask why they were making the gum too big. The operator responded that they were making the gum to the specification provided. When asked if they could make the specification tighter, the simple answer was, yes of course. The result: the slivers of downtime disappeared and production increased. Machine Truth TM provides the visibility necessary to make the right decision.

Having Machine Truth TM enables key personnel to ask the right questions and to stop spending CI dollars in wrong area.


Employee engagement in OEE. It is critical to sustainability.

Mistake proofing, or poke yoke, is commonly known as a technique for ensuring an assembly process is done in the correct manner however there are other “mistakes” that happen on the shop floor with equally serious consequences.

Over production, below-rate run speed, prolonged setup times, material-handling delays, etc. are mistakes that cost a manufacturing organization time and money.

Engaging your employees through the provision of real time machine truth data drives change in the correct direction and shines the light on the real issues impacting performance and profitability, e.g. Downtime, Idle, Setup.

TAKE THE BLINKERS OFF.


 

Recent Posts

Contact Us

Shoplogix Inc. 2626 Argentia Rd. Mississauga, Ontario Canada L5N 5N2 Tel: 905.469.9994 T/F: 877.469.9994 Fax: 905.469.9970 info@shoplogix.com

About Us

Shoplogix is the leading developer of manufacturing performance management solutions designed to enable manufacturers to reduce operating costs, increase manufacturing profitability and drive rapid time to value.

Our patented Plantnode® solution is the market’s only
integrated technology solution, combining the power of
software analytics with the strength and stability of
embedded technology.

Seeing our customers uncover their hidden potential to
realize dramatic performance improvements inspires us to continue to innovate in order to help them to further escalatetheir success.

Shoplogix was founded in 2002, and is headquartered in
Mississauga, Ontario Canada.