The Internet of Things (IoT) conversation has grown very quickly. A recent press release stated that the IoT in the manufacturing market will grow to over $13 billion by the year 2020. With this extreme growth, everyone is jumping on board for a piece of the profit, and many have jumped too quickly without carefully examining the risks. In the manufacturing industry, that can lead to major consequences for companies.
IoT in the manufacturing world is about data and how to get more of it. We can achieve this by connecting smart devices to our networks and analyzing the inputs. Unfortunately, a lot of what we have seen through mass media relating to IoT in has been marketing to push companies toward spending more money. This drive is solely related to the substantial growth estimates.
The Internet of Things (IoT) is not a new concept; we have lived with the IoT for more than 15 years. People use their smartphones to remind them of an appointment or for information on the most expedient traffic route. Applying smart devices to our lives allows people to be more efficient without much added risk. That strategy doesn't work inside process manufacturing environments.
IoT manufacturing benefits
Although there are benefits, the IoT is not some unseen sci-fi super power that drifts in and out of your factory and provides 100% overall equipment effectiveness (OEE). Any implementation of connected devices should be carefully considered beforehand. Involvement in the "IoT revolution" should be based around using strategic data capture to make more money for a business instead of just trying to be a part of the movement.
Of course, anyone in the industry can associate more information with additional benefits. Increased revenue areas based on increasing data collection include OEE metrics, uptime, reliability, manual input, energy output, and the ability to calculate the cost of production per equipment. The benefits are quite endless because of the many production processes and technological advances that occur on a daily basis. Overall, companies need to ask what they could do in their facility with improved process visibility, extended equipment life span, and reduced total cost of ownership (TCO).
Discrete vs. process implementation
Discrete manufacturing is like building a Lego Death Star from Star Wars; the company assembles pieces and parts to make a product. These discrete facilities have led the way into the Big Data mentality because of the ease of execution via limited controlled parameters. While it works for discrete manufacturing, don't be fooled into thinking that implementing these virtualized solutions is a one-size-fits-all approach across all manufacturing disciplines.
Process manufacturing would be more like making a batch of whiskey. For example, one element of the process is the addition of yeast, a critical ingredient. Yeast starts to decay when exposed to air or moisture, so timing is crucial when handling. It also can be affected by ever-changing conditions, which need to be monitored. Overall, the final product depends on hundreds of variables that have to be taken into account for that repeatable perfect batch.
By identifying and analyzing stable common cause and special cause variations, process engineers create control practices that help in increasing the ever challenging repeatability. With the introduction of more data, those variations can be tuned even further.
Get more from the data gathered
It seems that most of the focus on the IoT's primary objective is on real-time visibility. While visibility is important, we should not stop there. Companies should continue using that visible data to create predictive models and algorithms that mitigate future production discrepancies to an acceptable level. Preventive action is the name of the game in modern manufacturing.
Is your company reactive or proactive in its process? If the company always is responding to the last issue, the only way to improve production is to respond faster and work harder. It's easier to implement high-performance graphics with predictive modeling than only have the ability to react and generate a report of why we still achieved low outputs.
In that high-performance environment, the operator can see the red flags as they happen and, as the severity is displayed, the predictive control is programmed to take action on pre-approved faults and request operator intervention on the others. Bring this model to the plant's operators and discuss it with them. If they can create a list of "When that happens, I do this..." then the company should be able to increase its production.
The IoT is growing, and there is no way to know to what extent. But many people are calling it a movement that will be the fourth industrial revolution. However, instead of just jumping on the bandwagon, some more research on the benefits of IoT should be completed first. Yes, more real-time data can be used to generate increased productivity in your plant effectively; however, as data increases, security threats do as well.
It is important to discuss manufacturing "wants/needs" with qualified engineers and system integrators that have experience improving processes and implementing solutions surrounding larger data capture. We always have believed that the more data we can obtain, the more efficient we can become in production. That is likely true, but the relationship between the return on investment (ROI) and data capture is not linearly dependent, and can change drastically with the influence of IoT threats.
Eli Jenkins is an account manager with Cross Company. He has a background in chemical manufacturing, control system integration, and consultant sales. This article originally appeared on Cross Company's Integrated Systems blog. Edited by Chris Vavra, production editor, Control Engineering, CFE Media, cvavra(at)cfemedia.com.
See additional stories from Cross Company linked below.