Data Driven Manufacturing Equation

What is Manufacturing Data Analytics / Data-Driven Manufacturing

Let’s get on the same page about what Data-Driven Manufacturing / Manufacturing Data Analytics / IIoT is. This will be brief. We dig into this topic more in other blog posts.

Please note: I use the phrase Data-Driven Manufacturing interchangeably with IIoT (Industrial Internet of Things) and Manufacturing Analytics. There is a whole semantics discussion in there. However, we’re going to skip that semantic discussion today.

Data-Driven Manufacturing

At its core Data-Driven Manufacturing is about making decisions on manufacturing based on facts, not guesses and opinions You get that capability of making decisions based on facts by combining 3 things:

  • Manufacturing Data Analytics
  • Data-Driven Culture
  • IoT

Data Driven Manufacturing Equation to assist in Manufacturing data analyticsIoT is pulling data from devices and sensors on machines, making that data valuable, and getting that new information in front of the right people to use for making decisions and taking valuable action.

Being Data Driven means that all of the people in that company that require it have access to the right data, can use that data to make decisions, and most importantly are empowered to take valuable action based on that data….typically corrective or Continuous Improvement actions.

Having a Data-Driven culture at your manufacturing company is really important. Some companies may not be prepared to take on a Data-Driven Manufacturing solution because they lack a Data-Driven culture. People need to be open to change, improving operations, and using data to do it. At some companies that isn’t the case. In those cases you’ll often hear, “We’ve been doing it this way for the past 30 years…” and on and on.

That might be the case, but what worked yesterday may not work today. A large number of manufactures have their own certain way of doing things and have gotten accustomed to using the same process. Not too long ago we did not have this access to all of the data that we do have access to today. So it was acceptable and common practice to make decisions in a number of different ways, without any of those ways really being based on the data, the facts. The world is changing, technology is advancing every day. One major way the world has changed and continues to change is not only the devices that are being developed and more readily available, but also the consumer capabilities with these platforms. We can gather more data than ever before, and you need to know how to use it to your advantage. Data-driven manufacturing is the way to stay competitive. You can completely eliminate the tough decisions, and guessing what to do next. Industrial IoT systems will pull the data from machines that you need to see so that you can analyze that data and make your decision based on completely factual information.

Implementing a data-driven manufacturing system does not have to be a tough process. A lot of companies are already using data in their Continuous Improvement projects, gathering Lean data, having daily production meetings. So, they’re already using data. Data-Driven Manufacturing simply takes them a couple of steps further…it’s not a leap light years into the future. Data-Driven Manufacturing simply provides more accurate, additional valuable, and real-time information to use in the existing normal processes of daily production meetings, Lean, Continuous Improvement, etc. Once the Data-Driven Manufacturing solution is in place and used in some of the existing processes and once it’s proven to be really valuable, then the manufacturer can move to expand the Data-Driven Manufacturing solution across more machines, lines, and plants.

With the definition of Data-Driven Manufacturing covered, and a little extra flavor for how it can fit into a business, let’s talk about the Journey of putting in place and experiencing Data-Driven Manufacturing in our next post.

Machine Downtime and OEE: Example, Tech Choices, and Recommendations

Machine Downtime and OEE: Example, Tech Choices, and Recommendations

In previous articles, we have discussed machine downtime and OEE including why machine data is valuable, how to use the data, and how to get the data. Here we will provide an example and some other helpful information on monitoring machine downtime and how to improve overall equipment effectiveness. To calculate your own OEE, you can use this OEE calculator.

 

Example

In this example, we will use Kepware as a recommendation. Kepware is a very powerful and flexible tool. It is well accepted across industries and very well known. They have great support, can connect to many types of protocols and devices, and is well maintained.

Using Kepware to connect, you will:

  • Set up a channel over which to communicate with the device driver (MTConnect, Bacnet, Modbus, etc), and the network card to connect
  • Add a device or machine to connect to including the IP address and set up various data settings
  • Then add the Tags or fields with names and addresses in the machine
  • Open the Quick Connect tool
  • Connect to the machine
  • Test to see the data coming through

When pulling data you can either have Kepware push the data directly to a destination such as a database or a setup another system to pull the data from Kepware. You will view this data in an IoT software platform. What you view is depending on the IoT software platform you’ve selected including its capabilities, what is required to set it up and connect to data sources. Viewing data will also depend on how you’ve set up the logic in the system to process data and how you’ve set up screens to display it.

Choices of Tech

There are several options available out there to choose from and deciding which is best for you can be tricky. While you will need to decide which best suits your needs, here are a few options for machine downtime data gathering and processing we recommend:

Kepware

      • Industry-standard product, widely recognized
      • Is a PTC company
      • Translates data from multiple protocols including for CNC machines: MTConnect, FOCAS for the GE FANUC controllers, and others
      • Great support
      • Can fit into multiple solution architectures

ThingWorx

      • PTC is a well known and trusted brand in engineering and IoT solutions
      • Very flexible development environment for solutions
      • Multiple products to greatly extend the solutions
      • Wide support by many partners, of which we are one

Microsoft

      • Household name brand
      • Database, software development, and cloud tools are industry standards, including in manufacturing
      • Flexible tools provide multiple options for solution designs

Best Practices and What’s Next

Use best practices for LAN design, security, database structure, naming conventions in the data layer, etc. Use open standards: OPC, OPC UA, MTConnect, ISA 95, for example: OPC can be useful because it creates intelligent rules for how to collect data and how to verify have right data.

The next steps are to monitor the machine, look for issues that are valuable low-hanging fruit, address them, then watch the numbers over time to see if there are improvements.

Final Recommendations

Be thoughtful for now and the future with Industry Best Practices. Solve the problem now and at the same time set up a foundation for growing the solution in the future. Setup a framework with a data model, and a network that will standardize how to interface with any machines in your plant. This will greatly simplify setting up new machines and simplify how to access the data.

Reduce Manufacturing downtime to keep optimal production flow.

Manufacturing Downtime and OEE: How to Use the Data

Reduce Manufacturing downtime to keep optimal production flow.How to Use the Data

How can you use data from your machines like manufacturing downtime data to improve your production capacity? How can data be used to improve scheduling, reduce waste, use, or materials efficiently?

  • Start with collecting the data and converting it to useful information such as downtime and OEE
  • Then provide it to your team on a tablet, a computer next to a machine, or even a big screen tv in the plan.

You’re now looking at real-time data about your machine equipment states and events. You’re getting a real view of your capacity performance constraints measured against baselines with manufacturing downtime, starving, idle, and shortstops.  When looking at the data, machine patterns will often reveal themselves and you can gain a new understanding of actual capacity, product materials, throughput, and promised delivery metrics.

You’ll understand 1) where your actual capacity and production is at, and 2) how machines are actually performing to enable that capacity, including the types of manufacturing downtime. You can then use the actual machine data to troubleshoot and solve particular issues. For example, it’s possible in some cases where getting data on OEE and then solving problems to improve OEE by 10% can result in an increased capacity of 19% with better efficiency and throughput. That can then result in increased operational net revenue which results in improved product margins.

Two key issues to consider are education and culture:

  • Education: Companies will need to ensure teams are educated such that they accurately understand what the data means, and what actions they are empowered to take which will have a positive result on the machine data they see.
  • Culture: Awareness of the culture on the plant floor is important. There can be a perception of a big brother when operators and shop managers know that management can more accurately see what they’re doing and not doing. However, the culture can be developed such that staff can welcome opportunities to do a better job, to improve production capacity, quality, etc.

Manufacturing Downtime Use Cases

Sometimes data concepts can seem abstract, so here we will provide two real use cases.

Food Plant

  • Situation: Food production plant processing 980K lbs product per day to make broths and fat products. They used 6 different boxes and fluidized dryers in the drying process.
  • Problem: The plant was steadily losing capacity over multiple months that seemed to point to excessive downtime in their dryers and evaporators.  Data was collected manually showed a 14-day schedule with 6-minute downtime cycles but upon investigation, they discovered excessive downtime with their dryers and CIP processes.
  • Solution: Invested in a real-time data collection and process visualization system with OEE to track manufacturing downtime with performance measure metrics.
  • Result: Collecting real-time data and OEE showed improvements could reduce downtime by 25-50% with an OEE index of 60%. This converted to savings on the evaporators and dryers between $760K-$1.5M in cost avoidance over a 12 month period with a recovery of 25%-50% in plant equipment capacity and 10-20% product material loss recovery.

How to Use the Data

How can you use data from your machines like downtime data to improve your production capacity? How can data be used to improve scheduling, reduce waste, use, or materials efficiently?

  • Start with collecting the data and converting it to useful information such as downtime and OEE
  • Then provide it to your team on a tablet, a computer next to a machine, or even a big screen tv in the plan.

You’re now looking at real-time data about your machine equipment states and events. You’re getting a real view of your capacity performance constraints measured against baselines with downtime, starving, idle, and shortstops.  When looking at the data, machine patterns will often reveal themselves and you can gain a new understanding of actual capacity, product materials, throughput, and promised delivery metrics.

You’ll understand 1) where your actual capacity and production is at, and 2) how machines are actually performing to enable that capacity, including the types of manufacturing downtime. You can then use the actual machine data to troubleshoot and solve particular issues. For example, it’s possible in some cases where getting data on OEE and then solving problems to improve OEE by 10% can result in an increased capacity of 19% with better efficiency and throughput. That can then result in increased operational net revenue which results in improved product margins.

Two key issues to consider are education and culture:

  • Education: Companies will need to ensure teams are educated such that they accurately understand what the data means, and what actions they are empowered to take which will have a positive result on the machine data they see.
  • Culture: Awareness of the culture on the plant floor is important. There can be a perception of a big brother when operators and shop managers know that management can more accurately see what they’re doing and not doing. However, the culture can be developed such that staff can welcome opportunities to do a better job, to improve production capacity, quality, etc.

This article is just one part of a four-part series, check out our article on how to get the data.

statistics on OEE and machine data, use this information to keep productivity optimal, and reduce production downtime.

Production Downtime and OEE: Why Machine Data is Valuable

Production Downtime and OEE: Why Machine Data is Valuable

Smart Manufacturing is the use of one or more technologies including IoT, automation and robots, big data, artificial intelligence, and modeling to optimize manufacturing. An efficient manufacturing operation means, among other things:

  • Product flows through the plant seamlessly;
  • Machines have the highest asset utilization possible to drive as much production and revenue as possible; and
  • The plant’s machines, operators, and systems are providing data that is shared across the organization to enable visibility into the factory floor for making valuable decisions throughout the organization.

Optimizing of Manufacturing

Companies can optimize manufacturing in product concept generation, flexible operations for changing customer needs, manufacturing efficiencies, machines becoming self-optimizing and self-diagnosing, and creating a dynamic manufacturing supply chain.

Manufacturing Efficiencies

Manufacturers can work on developing more efficient and flexible operations. This greater efficiency and flexibility enables manufacturers to attain better asset utilization, production capacity, and lower overall costs of production. The higher production capacity and asset utilization, and flexible operations enable them to handle more dynamic customer needs and a dynamic supply chain. They can then dominate their market as a more flexible and consistent low-cost producer with solid quality and delivery to dominate their markets.

Need Data to Improve Company

To reach these levels of better production capacity and low-cost production the key is people must have access to data about machines and how they’re operating… uptime and downtime, asset utilization ratios, the productivity of the machine, quality of the products it creates, it’s running condition, how the operator is running the machine, etc. With this data, companies can then monitor production lines and pieces of equipment, find issues, and make improvements.

Production Downtime

Unplanned production downtime is the time a machine is down due to circumstances that weren’t planned ahead of time. This includes times when:

  • The machine is available but not running: the machine could be used and there is a product to produce but no operator is available, or there is an operator but no raw materials to produce a product.
  • The machine is not available: The machine is down because it had a failure and is under repair.

Production downtime is one of the biggest “killers” in production efficiency. It leads to material loss, resource loss, uptime or asset utilization loss, and capacity loss. Additionally, it reduces a company’s ability to deliver products to quality standards which leads to missing promised delivery dates and lost trust with a customer. In the end, it all results in higher costs and less revenue.

Benefits of Tracking Downtime

Tracking downtime gives you the data you need to both find root causes of downtime and make improvements to users, machines, and processes. When downtime improves then production capacity and asset utilization increase, costs of a product are reduced, revenue increases, and margins increase.

Downtime Study

In 2017, GE ServiceMax commissioned a study that surveyed 450 decision-makers across manufacturing and other industrial verticals. They found companies had issues as a result of unplanned downtime including:

      • 46% of companies couldn’t deliver services to customers
      • 37% lost production time on a critical asset
      • 29% were completely unable to service specific equipment

They also found that production downtime in a factory affects 62% of its productivity with scheduled resources and equipment. These numbers indicate unplanned downtime is common and has a big impact on companies.

statistics on OEE and machine data, use this information to keep productivity optimal, and reduce production downtime.
Downtime Fallacy

Companies who don’t track production downtime of their machines may believe they know what the equipment downtime is. However, they are often off by as much as 50%. Without the actual data to measure and analyze, many companies are not seeing the whole picture, and taking all information into account when measuring downtime – and this can greatly impact their business.

This article is part of a series, if you found this information insightful, check out our next article on downtime and how to use the data.

 

Why Should I Calculate OEE Manually?

 

Why Should I Calculate OEE Manually?

OEE (Overall Equipment Effectiveness) is a valuable KPI for how well machines on the plant floor are being utilized. (More information on what OEE is can be found in the Intro to IoT blog post) It is often thought to be a complicated value to calculate, and sometimes it can be. However, it can be greatly simplified. Given the value and the potential simplicity, it is recommended that operations staff including their supervisors and managers should know how the calculation works, what OEE is, and how to use it effectively.

Why is OEE Valuable?

OEE has a direct correlation to the revenue the machine provides the manufacturer. If the machine isn’t utilized well, i.e., if it’s down or if it’s producing poor quality parts that cannot be sent to a customer, then the machine isn’t generating as much revenue as it otherwise would.

It’s also valuable because breaking the value down into is parts can help hone in on the cause of issues with a machine. It is made up of three separate percentage values: Availability, Performance, and Quality. One can look at the component values and determine where to start their analysis to find issues with the machine.

Why Calculate Manually

Why should operators and related staff calculate OEE manually? Two reasons:

  1. To understand what OEE really means and what it’s made up of, and
  2. To get a solution calculating OEE in place when an automation solution isn’t available.

Understanding OEE: One can only understand another person’s perspective when they’ve walked a mile in their shoes. In the same way, a person can only understand a concept once they’ve recreated it themselves and worked with it. The nice thing is that OEE, or at least some parts of it, can be calculated easily on paper.

Automation isn’t available: In many manufacturing companies, OEE will likely be calculated automatically by various systems such as IoT software platforms (these are the solutions we often implement). These solutions are becoming more predominant as companies look for more ways to pull valuable data from their machines to complete…all under the moniker of Smart Manufacturing. However, some companies may want to start utilizing OEE before budgets can be made available for these solutions. So, why not start when pen and paper!?

How to Calculate Manually

A deeper discussion of the calculation of OEE can be found in this How OEE is Calculated blog post. So we won’t cover the details here.

To calculate OEE all you need is a pen, paper, maybe a clipboard, and diligent effort to record and calculate the data regularly. It is also helpful to have Excel available. Don’t need anything fancier than that.

You can start with Availability, one of the components of OEE. Put the paper, pen, and clipboard close to the machine to measure. Then every house has users mark what happened with that computer regarding its availability: If it was up and running at normal speed mark a “G” for Green (can include planned downtime if appropriate). If the machine experienced some setup or adjustment time during that hour, mark a “Y” for Yellow. If the machine was down, mark “R” for Red. You could also use color pens/markers and/or a whiteboard.

Then count the total number of G’s, Y’s, R’s. Subtract the total number of Y’s and R’s from the total number of G’s to get the number of hours the machine was available to run. Count the total number of hours overall on the paper no matter the condition for the scheduled operating time. Then divide available time by the scheduled operating time and you get the Availability Rate % value.

Do this for each day, record the results in Excel, maybe even create a simple chart, and you’ll start to see a trend for where the OA (Overall Availability) for the machine is. If you do this for multiple machines you can then get some solid, objective data on the status of the various machines, and which might need some attention.

You can also repeat this process recording similar data for Performance and Quality.

How is OEE Calculated?

OEE (Overall Equipment Effectiveness) provides an indication of how effectively machines are utilized on a manufacturing plant floor. It’s valuable for identifying and removing production constraints and for driving up revenue earned by every machine.

The calculation has some mysticism surrounding the complexity of its calculation. We’ll simplify the calculation to a level where the reader can immediately start calculating it by hand.

How to Calculate OEE

OEE for any machine is made up of 3 parts multiplied together: Availability (up-time) x Performance (production speed) x Quality (widgets produced correctly the first time). Below is a brief explanation of each part and how it’s calculated.

Availability Rate

Machine’s up-time, the percentage it is ready to produce products and is working properly, excluding changeover and setup time.

Formula: (Scheduled Operating Time – Downtime) / Scheduled Operating Time

Note that Scheduled Operating Time is not “total time”. Availability refers to when the machine could be running based on when it is needed or planned to run. There could be reasons it won’t run when it is needed (i.e., setup, breakdown) and this calculation must account for those downtime reasons.

Performance Rate

The rate a machine actually produces products relative to it’s best known or standard production rate.

Formula: Actual Output / Standard Output

Note it’s critical to come to terms at your plant with what the standard production rate is for your equipment. It’s best not to use the specified or design production rate by the vendor of the product. In this case, if the machine were running faster than the design production rate and you were using the designed production rate this could mask quality or availability issues in the OEE calculation.

It’s also worth noting that the production rate will highlight losses due to idling, slowdowns, and minor stoppages.

Quality Rate

The rate the machine outputs good parts.

Formula: Right First-Time Output / Actual Output

Note that products produced by the machine which require rework or any sort of adjustments, along with scrapped products, are not counted as a quality product.

Example

Now let’s run a quick example calculation for one shift:

Availability: 

Scheduled Operating Time: shift 8 hrs or 480 mins, 20 mins planned downtime, 0 mins breaks; total 460 mins

Downtime: breakdowns 30 mins, setups and adjustments 15 mins, minor stoppages 15 minutes; total 60 mins

Available time = Scheduled time – Downtime = 460 – 60 = 400 mins

Availability rate = (Scheduled Operating Time – Downtime) / Scheduled Operating Time = (460 – 60) / 460 = 87%

Performance:

Actual Output: 400 parts or 1 part/min for 400 mins of Available time

Standard Output: 800 parts or 1/2 part/min for 400 mins of Available time

Performance rate = Actual Output / Standard Output = 400 / 800 = 50%

Quality

Right First-Time Output: 400 parts – 20 defective parts = 380 parts

Actual Output: 400 parts

Right First-Time Output / Actual Output = 380 / 400 = 95%

OEE = Availability x Performance x Quality = 87% x 50% x 95% = 41%

Additional Considerations

One must keep in mind that OEE is not a value that can be used to compare the performance of many different machines, different production lines, or even less so various plants. As you can see above, there is a lot of uniqueness built into the calculation for each machine. Each machine is unique for the schedule required of it, it’s inherent production rate (by its design), and other factors. Therefore, OEE is really only meant to be used as a metric to improve the performance of each individual machine. Monitor the OEE value for the machine, break it down into its component pieces, and hone in on the problem, fix, and then watch OEE to see if it improves after the fix.

Even if the above is true, OEE can be used to compare machines and lines, but only if they are used in a similar fashion and have similar demands.

Comparison of OEE across different lines and plants can be done but should only be done 1) when using an average OEE value, and 2) with the understanding the value will not be highly accurate.

Introduction to OEE (Overall Equipment Effectiveness)

Many companies have jumped on the bandwagon of using OEE (Overall Equipment Effectiveness)

as a KPI for their manufacturing equipment, lines, and plants. This article will introduce what OEE

is and briefly discuss how it should and should not be used.

What is OEE

OEE indicates how effectively machines are utilized on the manufacturing plant floor. Another way to put it is OEE is the percentage value the machine is performing up to its true potential.

OEE is made up of three parts, each of which is calculated as percentages, and the total value is calculated by multiplying all three parts: Availability rate x Performance rate x Quality rate. Ex: 70% Availability x 80% Performance x 90% Quality = 50% OEE

Each of these three parts can be further broken down as follows:

Availability (uptime of the machine)

  • Equipment failure or breakdowns
  • Setup and adjustment

Performance (speed producing widgets)

  • Idling and minor stoppages
  • Reduced speed of operation

Quality (products produced correctly the first time, without rework)

  • Process defects (scrap, repairs)
  • Reduced yield (from startup to stable production)

Why is OEE Valuable?OEE

This is very valuable because the utilization of an asset has a direct correlation to the revenue the machine provides the manufacturer. Of course, you want to make sure that your machine is operating as efficiently as possible. paying attention to OEE will help you to keep your machines running more, with less downtime. OEE can also assist in making sure that your machine is making a good quality product to be sent to the customer. If your machine is down or making poor quality products that cannot be sent to the customer, then the machine is not generating as much revenue as it otherwise would.

Having machines down or just not working at an effective pace can greatly affect the overall production of the plant, and cause you to miss deadlines. That means less happy customers and a higher chance of them taking their business elsewhere. Happy customers mean more business, more money, less stress, and a great strong reputation to retain a good customer base and push your growing business up to the next level.

The bottom line is that your machines across the plant floor will be more productive, making better products. Making sure that your machines are being used to their full potential will lead you to generate higher revenue.

How to Use it

Here are 3 ways we recommend using OEE if your company is new to this manufacturing KPI:

  1. Find machines that need attention: A good manufacturer will attempt to remove all constraints in its production process whether or not that additional capacity is currently needed. Users can find constraints or issues with machines by looking at the individual OEE values for each machine and determine which are performing effectively and which are not. Lower OEE values will indicate which may need some attention. Addressing issues with machines and thereby improving their operating effectiveness can then increase the revenue the machine produces.
  2. Find the cause: Once a machine with a lower OEE is identified the staff can look at each of the three components of OEE (Availability, Performance, and Quality) and find a leading cause of the OEE difficulties. Additional analysis is then required to find the cause and fix. Then setup improvement projects to address the issues.
  3. Confirm improvement: Once the issues are addressed with the machines continue to monitor the OEE values for those machines. If the OEE numbers go up, especially in the parts where the issues were addressed, then the efforts to fix may have been worthwhile. Otherwise, the machine and its issues may warrant a second look.