The Industrial Internet of Things (IIoT) is a buzzword we hear all the time. (If you’re still looking for a straight answer on what exactly IIoT is, I recommend the excellent article The IIoT in a Nutshell by my colleague, Kagan Pittman.) But for many manufacturers, the question isn’t “What is IIoT?” but rather, “How do we use it?”

“We all know we can get data out of machine controls. We have done this in automotive since the eighties,” says Karl Rapp, Applications Engineering Manager, Automotive and Machine Tool at Bosch Rexroth. “I think that in the future, every company will have a data scientist. If we know what the data means, and where it’s coming from, then we can improve the process, whether it’s a machine, maintenance or anything else.”

In large multinationals, especially in process manufacturing industries like petrochemical, it clearly appears beneficial to have each facility, each boiler, and even each pressure sensor connected to a powerful central data processor so that management can maintain control. But what about smaller companies? What about parts suppliers and contract manufacturers? Is IIoT and machine data monitoring technology worthwhile for small to medium sized enterprises?

I spoke with several IIoT industry experts to find the truth about this question, and the only thing they did agree on was that yes, there are achievable, measurable benefits to be gained by collecting and analyzing machine data, no matter the size of the operation. What they didn’t agree on was the best way to go about it.

Types of IIoT Software

MES

Manufacturing execution systems (MES) act as a bridge between your enterprise resource planning (ERP) software and machine data. Memex Inc., the company of one of the experts featured in this article, specializes in MES software. MES  is all about sharing information into the corporate network. This can enable tools like dashboards, summary information, reporting, email alerts and other network-related tasks.

SCADA

Is IIoT Worth It?

How small does an operation need to be before IIoT becomes irrelevant? This is one of the most common questions on the subject: Is my company too small to see benefits?

For smaller operations, such as an auto parts supplier or job shop, presentations from large automation companies may seem intimidating. A sales manager presses a button simulating machine stoppage, and the system pings his smartphone with an alert. Other presentations show an online dashboard plastered with line graphs and bar charts, showing everything from temperatures to total monthly downtime to spindle speeds.

But what exactly is the point of all this data? How can it be used effectively?

Dan O’Brien, an IIoT expert at Honeywell Connected Plant, said this question is a driving focus of Honeywell’s IIoT business.

“It’s a really legitimate question: how is connectivity going to drive competitiveness, visibility, operational certainty? There’s a million different words floating around about IIoT, cloud and this type of stuff. But what are the opportunities afforded by the collection of equipment or process data?” O’Brien said.

According to John Rattray, senior VP marketing and business development at Memex Inc., if you have even just one machine, it is possible to find actionable data metrics to turn into ideas for continuous improvement and increased productivity. However, it’s worth examining the level of control you maintain over your plant without the system.

“I have one customer, Lynch Fluid Controls, who uses Memex MES in a high mix, low volume environment. They’re doing setup changes. They have highly flexible production. It’s tedious to do all the data collection because of all the changes. So, they have found that the automation of their data collection has improved scheduling, costing, production runs and other benefits,” Rattray said. “I’ve seen a customer with just six machines take the data and use it effectively to create business growth.”

“If you’ve got 15 or 20 machines and you run one shift a day, you probably don’t need a MES system, because you, as the owner or manager of the business, can walk around the plant and make sure everything is running optimally. On the other hand, if you’re running multiple shifts and you aren’t able to maintain that level of visibility, MES can collect the data, have the history, and have the reporting done,” Rattray said. In some ways, the question depends on company culture and the nature of the business.

Micah Statler, Operations Manager at Advanced Technology Services (ATS), agrees.

“If a company is analytics driven and they’re going to do something with the data, then they should dive in head first,” he said. “I think the biggest challenge I see out there today is that there are individual engineers that really would love to see the data and be able to make on-the-fly changes and improvements and do the analytics. But organizations overall are a little more tentative to make an investment. When they do, it’s not always used to its full potential, and that’s really a shame. Because there’s so much power in the numbers.”

Karl Rapp brings up another interesting point: machine data is of particular interest to the machine OEM. By collecting data on the machine such as vibration, temperatures and other sensors, the OEM can identify areas of improvement for their machine.

“The machine builders do not have data on how good their repeatability is, if they’ve never recorded it,” Rapp said. “Once they start adding sensors in there to record it, the OEMs can actually improve the quality of the machine they built. If you record the motion data and the data analytics tells you, ‘the signature of this is completely different,’ you can ascertain that it could be due to improper torque on a bolt, for example.”

According to Rapp, that’s where the opportunity of IIoT is for machine builders.

“It may take you one or two years to really understand your machine. The mechanical engineer that makes his FMA analysis, if he has vibration sensors in the machine he can verify in the real world how it really behaves, and where the resonance nodes are. So, he can compare that information with his models and analysis and make modifications to the machine to improve it.”

Armed with real-world machine data to verify simulations and models, the machine builder could create a ‘health index’ of the machines, which can in turn be used to reduce downtime.

So, if you’re an engineer who sees the potential power in the data, how can you convince management to invest in an MES system?

Getting Started with IIoT in Your Factory

This is the strategy to use when you don’t know where to start, or even how exactly IIoT data will benefit you.

“For instance, you take a critical machine for which you know you want to maximize utilization. This could be a single machine, a cell of machines or one production line. Focus on getting connected, on getting the information from that machine,” Rattray says,

“With the IIoT connections in place, utilize that data and drive the benefits out. Any one percent savings on that critical machine is probably going to translate into hundreds of thousands of dollars. So, you really want to get started in an area in which you can create some quick benefits,” he added.

If executed correctly, Rattray’s strategy will serve you a gold-plated ROI pitch for getting funding for IIoT upgrades. When you do get the green light, here’s your next step:

“Bring out your gut feel. Chances are, there are machines in your plant which, in your gut feelings, you know should be doing better in terms of productivity and efficiency. But, you just don’t know what the issue is and why the machine isn’t running optimally. With IIoT technology, we can pinpoint down to the second everything that goes on with that machine. Now you can start to analyze, categorize and summarize all the information and start to make informed, effective decisions on it,” advised Rattray.

Micah Statler offers similar advice to engineers who know the data they are looking for but are concerned that an IIoT system will inundate them with data they aren’t prepared to use.

“One of the key success factors we see customers using is knowing what questions you want answered by data, rather than just going out and collecting it. Part of our process is really sitting down with the production staff and engineers and trying to figure out what questions they want to answer,” Statler said. “It’s very easy these days to put a sensor on anything that collects data, but if you don’t know why you’re doing it, then you miss the whole point.”

Dan O’Brien urges smaller companies to embrace the data analytics power of IIoT systems as well:

“There are process companies that have so many engineers analyzing every bit of data, and maybe their lower tier counterparts don’t have access to that much expertise. So, if they can use connected systems to make up for their lack of experts on staff, they can benefit just as much as that multinational trying to squeeze the last half-percent of uptime out of their production,” he said.

How to Bring Legacy Equipment Online

The first step is to bring equipment up to a minimum modern standard: digital controls. If you’re running antiquated equipment or software, you can’t get the support you need when things fail. For example, Microsoft ended support for the Windows XP operating system in April 2014. The operating system no longer receives security updates and Microsoft does not provide any technical support for XP. If your system crashes, there may be very little support available, even from third-party services, to get things back up and running. The same holds true for the machines themselves. What is your machine’s OEM? Is it out of business? Does it still support the boards, processors and controls of your model?

“The first avenue is where there is good iron on the floor. The physical machine may be capable, but the controls are not. The first way we can improve the reliability and performance of this equipment is by updating the control systems so that critical spares are supported. In addition, newly-educated technicians are familiar with newer systems,” said Statler.

With the controls updated, the machines can connect to ethernet, and therefore IIoT and SCADA platforms. 

The next step is to bring safety controls up to modern standards. Safety standards have changed significantly in the last twenty years, and chances are good that if your machine controls are not up to modern standards, your safety controls aren’t either.

The third avenue is to take advantage of the first avenue—PLC upgrades—as an opportunity to incorporate new sensors and data feedback devices as inputs to the new PLC for extraction through a SCADA system. This opens the door to advanced data analytics and actionable metrics.

Benefits of IIoT: Just in Time Manufacturing, Predictive Maintenance and More

Machine Repeatability

Karl Rapp indicated another benefit of machine sensor data: repeatability. If you record high-frequency sensor data, collecting data in millisecond increments, you can monitor things like vibration, speed and temperature as you run a machine operation. For example, if you are cross drilling holes in a block of steel, you can monitor the torque of the motor and refer to the data to determine if the holes were crossed straight enough by comparing the drop in torque.

For example, this locally processed result information can then be used to inform the machine controller automatically to, for example, have the workpiece marked “good” or “defective” and shuttle a “defective” workpiece to an inspection station instead of the next manufacturing stage.

Such local (and fast) data analytics at the machine offer a faster ROI than lower frequency data collection for long-time predictive maintenance analytics. Naturally the local data analytic server can also allow configuration of desired data for preventive maintenance, locally buffer  it, pre-analyze it before it gets passed to a cloud system for further analytics, e.g. across multiple machines and locations.

JIT and PdM

Most manufacturers use their systems to track basic metrics like throughput, fail rate, uptime and downtime. By analyzing these metrics, it becomes possible to make continuous improvements to personnel or processes to tune up your production. In addition to these incremental improvements, Just-in-Time (JIT) and predictive maintenance (PdM) are the real revolutionary opportunities enabled by IIoT technology.

Statler gave a great example of the power of IIoT for these two things:

“When the last hurricane came through and decimated Puerto Rico, our customer set up 23 large diesel generators to run three plants. The problem with these generators is that they were originally there for intermittent power outages. When it became clear that Puerto Rico would be likely without power for up to six months, the customer called ATS. He said, ‘We need to know what’s going on with these generators at all times, because when and if they go down, it’s going to affect not only the productivity of that plant but also the productivity of five other plants that they feed.’ So, we went to Puerto Rico and put together a package that, through satellite connectivity, allowed them to monitor the status of each of those generators at any given time, from smart devices. IIoT is not micromanagement, and it’s not just data. It’s better communication, and better control.”

Since those plants in Puerto Rico were running “just-in-time,” it was necessary to use advanced manufacturing technology to keep that supply chain running. This example shows how PdM stabilizes and goes hand in hand with JIT. This is useful information for large multinationals, but critical for suppliers.

Why? Because if your client runs an advanced IIoT system, there’s a good chance you will be, too.

Connecting Remote Assets

“At Honeywell, we mainly work with process manufacturing, like petrochemical,” said O’Brien. “For this IIoT purpose, we’re thinking about things that have been more isolated or stranded in the past. Now, we’re able to remotely connect the asset awareness. So, what’s the measurement? What’s the condition? What’s the asset health?”

Sensor data such as vibration, temperatures, forces and torques are now remotely accessible. The ability to centralize the assessment and control of equipment brings greater control and monitoring of remote assets.

MRO

IIoT insights can have a significant impact on maintenance, repair and overhaul (MRO). Dan O’Brien provided an example:

“We can open that data up to the producers of the assets, because they run a little bit blind, too. For example, they build a compressor, ship it out and get it back for repair. The OEM says, ‘what were you doing to it when you broke it?’ and the producer would just say, ‘I don’t know, just fix it.’ With data collected from the assets, we can obtain a better picture of what’s going on.”

Can a Tier One Stay Competitive Without IIoT?

“If you’re a tier one supplier, I don’t think you have a choice,” said Statler. “I think you’ve got to get on board, because your customer is going to continue to push for more distinct delivery, leaner production. They’re going to keep pinching you on costs, and I think the only way you’re going to improve efficiency and communication with the customer is by getting more keen with the data and diving into the analytics. It will also give you the benefit of being able to clearly communicate what your work has done to influence their productivity.”

Dan O’Brien of Honeywell agrees.

“Certainly, there have been successful companies who have run without the cloud. However, there is a lot of opportunity to help companies run their asset-intensive operations more efficiently, with better actionable insight. There’s emerging technology with big data and remote connectivity. Of course, it’s possible to do manufacturing without IIoT, but there are tremendous benefits available to those companies that use the technology,” he said.

“We’re starting to see it taking hold strongly in some industries, but just think where it’s going to be in ten years,” O’Brien added. “I think that ten years from now, companies that said, ‘we’ll never connect’ are going to be at a severe competitive disadvantage.”

 IIoT Security: Is Industrial Internet Secure?

While most manufacturers aren’t running unsanctioned nuclear testing, security should be a major concern for any company dealing with intellectual property and trade secrets.

Karl Rapp, pointed out that despite the cloud-based technological developments from IoT vendors, industry players are not open to many of these options because of IT security concerns.

“Cloud IoT is actually, I would say, unrealistic today,” Rapp said. “Everybody has the idea to stream the data to Azure or another cloud service. The problem is, you are not getting there—because every manufacturing plant that is of a bigger size has an IT department with security concerns, and they limit access to the production floor from the outside. Most customers, in my experience, do not allow you to send their data to the cloud, and they do not allow a remote tele-service to their production floor.”

Rapp points to further issues with cloud storage in terms of geographical and political location:

“With cloud services, what you are paying for is the data transfer that goes to and from the server. When I make, say, a Fitbit account, I have no idea where the data is stored—and I don’t care. But for manufacturing data, your legal department will want to make sure that the data stays here, and doesn’t go to China, for example,” said Rapp.

However, Statler disagreed.

“I think people are, in some cases, a little over-concerned,” he said. “I think it’s safe to just pipe the blue ethernet hose all over the plant and keep everything in-house, but the good thing about SCADA platforms today is that they’re not integrated to the machine control in such a way that you’re able to pull programs or things like that. There are tons of applications in place that can really dismember and isolate data before dumping it to a server that then talks to the cloud. Data is repopulated at the analytics end. So, if you wish to keep the data on site, that’s good, but they may lose out on multi-plant functions, such as the ability to compare data and do remote diagnostics.”

John Rattray, of Memex, brings up another good point: even if you blue-wire your plant and use a local data storage system, as Rapp suggests, you are still vulnerable to cyber-attack.

“We have one small customer in upstate New York. They had a service technician come in to service a machine, which is a routine occurrence. The technician put a USB stick in the machine. Unknowingly, the stick infected the machine with a virus, and the customer’s entire corporate network went down. Repairing the damage, and upgrading security, became a mission-critical, primary business focus for this customer,” Rattray said.

He went on, “I was at a plant in South Carolina, and we got talking about some of their security issues. The plant manager looked at me and said, ‘We’ve been known to see drones fly over our plant and we wonder if they’re picking up wireless signals from our machines.’”

Rattray emphasized his statement. “Ninety-nine percent of manufacturers don’t want to go wireless on the shop floor.”

Dan O’Brien acknowledges the risks as well. “I don’t think that physical system and cyber security is a myth. It’s just a big reality with the way the world is today. When I travel to customers, they tell me most of their security vulnerability comes from physical intrusions within the systems of the plant. A vendor brings in a memory stick, for example. A lot of the security breaches have been on-premises originated, not cybersecurity originated. I think you need both.”

Industrial IoT Data Security Solutions

While these solutions reduce the risk of cyber-attacks, they also supplement the work of your IT department and the security hygiene of all employees. For example, does your IT department enforce regular password changes? Are employees trained not to load unknown USB sticks or email spam? Is there a sign-in process for external contractors? Without these security basics, all the IIoT budget in the world can’t make your data secure.

That being said, here’s what the experts recommend for keeping data safe. Karl Rapp favors the industrial PC storage and local intranet approach:

“In order to take advantage of IoT, we have data on the analytics server. This can be used as a database tool. So, we transfer the data from the controller or sensor into this database, but in order to benefit from IoT, we need to do something with this data,” Rapp said.

“In this database, we can add machine learning functions, MATLAB, Simulink or other software to read the data. Once the customers see the benefit, your company will have more willingness to store all the data in the cloud. But, often you cannot start with the cloud, for example because the IT guys will restrict or block access,” Rapp added.

This advice echoes the above suggestions from Rattray and Statler on getting started in IIoT in the first place. However, it adds the idea of avoiding potential security risks before your team has familiarity with the technology.

Many companies offer devices designed for cyber security protection. Mazak, for example, offers a product called a smart box, which connects as a switch between the machines and the controls. The smart box has various functions designed to protect the machines from cyber-attacks.

For example, says Rattray, “It’s got safeguards and controls inside. It prevents intruders from getting access to the machine, it prevents machine control from going off-site. If someone sticks a USB stick in the PLC, the smart box shuts the machine down. This is network switching control and network security.”

Honeywell offers security solutions as well. “We have a separate cybersecurity practice with a bunch of products to help keep things safe,” said Dan O’Brien.

Conclusion: Drive Out the Benefits of IIoT in Your Plant

If you’re an engineer who sees the immense opportunities for both incremental and large-scale benefits by implementing MES, SCADA, or other IIoT technology, follow the advice given in the ‘Getting Started’ section of this article. A good first step is to contact the experts at ATS, Memex, Honeywell Connected Plant, and Bosch Rexroth. These companies did not sponsor this article, but they did provide the expertise.

Ultimately, like all emergent industry 4.0 technologies, the engineers of the manufacturing industry are still seeking, refining, and developing the best ways to use IIoT to its full potential—and the question of the cloud’s role in it is no exception.

“It’s not a pure, ‘cloud or no cloud’ discussion,” adds O’Brien. “My argument is that it’s a matter of connecting the dots to improve your asset reliability and economics. I think IIoT and cloud are all really important way that customers can become more competitive and attract more generations of industry workers that we need to attract.”