How the Right Production Monitoring Strategy Can Improve OEE

Manufacturers that use production monitoring have better OEE Performance, study finds.

Can your factory be smarter?

Are you ready for Industry 4.0?

Industrial software vendors and trade media love to raise these questions, but for manufacturing professionals, the real question is: Do I really need Industry 4.0 technology to be competitive?

To answer that question, engineering.com recently conducted a survey comparing the use of production monitoring with production performance, as measured by overall equipment effectiveness (OEE). The study found that better OEE is correlated with better monitoring capabilities. More importantly, the study found that what system you choose, and how it’s implemented, may make a big difference.

The 254 survey respondents included Director and Manager-level personnel, as well as engineers and technicians in manufacturing industries, including automotive, food and beverage and aerospace.

Source: Research Report: The Connection Between Production Monitoring and OEE

Source: Research Report: The Connection Between Production Monitoring and OEE

Part of the analysis included grouping the respondents by their self-reported level of OEE performance, then comparing the production monitoring strategies of those different groups. The full research report titled, “The Connection Between Production Monitoring and OEE” is now available for download at no charge.

The data showed that while more advanced production monitoring correlated with better OEE in a broad sense, pulling actionable insights out of the data can be a real challenge.

“OEE is one of the KPIs that people pay a lot of attention to when evaluating production monitoring,” said Bill Boswell, VP Marketing at Siemens PLM. He broke the concept down into three parts.

Connect and Monitor

“Most people tend to start at the connecting and monitoring phase by connecting and syncing assets to a system like Mindsphere (Siemens IoT platform),” explained Boswell. “Once they begin collecting data, they tend to do some rudimentary utilization analytics on it. This will typically help them get started in asset management and then move into condition monitoring. Once they move into the performance management of the asset, you can begin taking action with measures like OEE.”

Unlock More Advanced Capabilities

To paraphrase Boswell, the next step forward in IIoT solutions includes more advanced applications such as predictive maintenance, AI and machine learning. These and other more advanced strategies of analysis and prediction help optimize processes and prevent failures.

Upshift into Full Digitalization

The third level involves a shift toward full digitalization that can transform business processes. At this phase, companies begin to take advantage of a “digital twin” of their process, not only within a plant but across the supply chain, across different facilities, and in the field.

That vision for the potential of Production Monitoring technology stands in some contrast to the reality revealed by this survey.

Fact: More Robust Data Collection and Analytics Correlates to Stronger OEE

In this graph, the purple bar represents respondents who reported strong OEE performance. As the data shows, operations that use analytics in conjunction with networked production monitoring are much more likely to report strong OEE than those with less developed monitoring strategies (the teal colored bars).

Does that mean any manufacturing operation can simply tack on an IIoT monitoring system and expect to see benefits?

According to Jim Brown, President of Tech-Clarity, achieving superior OEE requires a results-based approach to planning.  “Start with your business strategy and understand what you’re trying to improve in your plant,” he advised. “IIoT and equipment monitoring can be an incredible way to improve OEE if you look at IIoT as a way to achieve your OEE goals instead of starting with an IoT system and looking for ways to measure its success. It’s important to put the business first and not the technology.”

Problem: Actionable Insights Can be Difficult to Access

asdf

In this chart, the x-axis represents self-reported OEE. The y-axis represents the number of respondents.

This graph shows that it’s not enough to collect production data. The real value is added when analytics can provide actionable information. For example, an industrial oven may report temperature to a central database, with the data accessible on a dashboard interface. The dashboard can alert operators when the oven temperature is out of spec, enabling them to adjust parameters and improve product quality. This is basic data collection.

However, if that real-time data is combined with historical data from the oven, along with the product being baked and compared to other ovens in the plant, an analytics system can reveal previously hidden insights. For example, it might show that oven temperature consistently drops when a certain product is baked. Simple monitoring is reactive. Analytics opens the door to proactive solutions to production problems.  

“What’s really ground breaking, I think, is the IoT capability to collect more data and combine it with data from other sources—for example, from an ERP system or from environmental data—and to take a broader look and apply analytics,” Brown said. “That’s where you begin to achieve the goals that have been out of reach in terms of improvement through your normal process control. If you take the time and you gather that more advanced data, that can feed back into your existing continuous improvement program, allowing you to find new problems and their root causes. That’s how IoT analytics enables even higher productivity improvements.”

Boswell agreed. “We’re seeing some customers do interesting things, such as looking at whether a device is operating in the field the way that it is expected to, not just the way that it operated over some arbitrary amount of time,” he explained.

“One example of that is a company called Ham-Let Group that builds IoT-enabled high pressure valves. They’ve built valves that are used in everything from electronics and semiconductor manufacturers, to oil and gas. The devices are IoT enabled right from the valve. With this technology, more data is accessible than simply the open/closed state. For example, the data can show that the valves operate best under certain operating conditions, enabling engineers to actively avoid conditions where it would freeze in the past.”

All of this data collection and analysis requires an IIoT platform. So, how should you select the right one?

Solution: Select the Right IIoT Platform

asdf

IIoT platforms come in three main flavors. Manufacturers can build a solution in-house that’s designed to address their immediate concerns; they can choose an off-the-shelf option provided with a machine tool; or they can evaluate purpose built IIoT platforms such as Mindsphere, Thingworx, Predix et al. This survey revealed that those who chose the third option had the most success with their production monitoring system.

“One of the ways to get more value from IIoT data is by including more data from a variety of sources and finding the hidden correlations between them,” said Brown.  “Having a platform that’s built to pull data in from multiple sources and analyze it will do that, so I’m not surprised to hear that that’s providing more value. That definitely fits with our experience.”

However, it’s important to recognize the value of systems built by machine providers. In some applications where the platform is not required to “play nice” to a complex degree with third party machines or software, this could be the best option. Alternately, as Boswell pointed out, if you have a complex environment, the solutions from machine vendors may not address all of the potential connections required.

Now that purpose-built IIoT platforms are becoming more common, there is less reason for a manufacturer to develop their own system. While in-house systems offer the ultimate in customization, they are typically less robust and don’t come with support. Also, the project can be larger than initially scoped.  “If you build a system up from scratch inside a company, you will need to build all that plumbing that sits on top of the infrastructure’s service provider,” Boswell emphasized. “Working with [a specialized vendor] allows you to focus on creating and deploying those specialized production applications that you need, without having to reinvent the wheel every time.”

 “It’s all about the industrial domain knowledge,” explained Boswell. “How you apply analytics to the availability of service on a high-speed rail line is different from how OEE is calculated for machine tools, or industrial analytics to manage a smart city environment. So, a vendor that works across all those different industries is going to get you a result that’s more applicable to your industry and your problem set.”

Next Steps for Production Monitoring

Brown’s analysis takes a pragmatic approach to Industrial IoT.

“Find whatever that hairy problem is in the plant—maybe it’s a piece of equipment that continues to go down, or a piece of equipment that tends to produce off-spec product, for example. Find a problem that the normal continuous improvement program hasn’t been able to solve on its own,” Brown said. “Then, figure out what can be done to gain more intelligence, to gain more knowledge about what’s actually happening based on monitoring and analytics and crack that nut.”

“The hardest first step is just getting started and identifying the things you are going to monitor, and which key parameters impact your business,” said Boswell. “Oftentimes, a company will have an “a-ha” moment and find out that while they set out to do one thing, but the real benefit they got out of their IIoT implementation was slightly different.”

You can now download the full report on The Connection Between Production Monitoring and OEE to see all of the insights that connect production monitoring and OEE.