The Internet has become an essential part of our lives and it is quickly becoming integral for competitive manufacturing.
The Internet of Things (IoT) has grown from a consumer desire to connect all our smart devices together – in manufacturing, this means connecting machines and automation across a facility to a single digital platform or infrastructure. We call this the Industrial Internet of Things (IIoT).
The historical approach to manufacturing has been a “run-to-fail” model, as mechanical failures have been impossible to accurately predict until something is already beginning to break down. Modern sensor hardware and computing software can allow manufacturers to analyse captured data from machines to detect approaching mechanical failures. This data can be stored online in digital data banks, like the Cloud.
Machines, automation and other assets throughout a factory can be connected to an IIoT platform to maximize scale and digital intelligence.
For digestibility, plant operators, process and equipment engineers and managers can monitor incoming data through HMIs (human machine interfaces) on the factory floor, computers on-site and even mobile devices such as smartphones and tablets. Known as edge devices, these can aid field technicians and onsite maintenance personnel by eliminating the need for routine diagnostic inspections, thereby reducing costs, labor and time to action.
Predictive maintenance is only one possibility with IIoT-connected factories. Insights found through analytics can even lead to the discovery of problems that have previously slipped under the radar.
“We can now monitor more things than we ever could in the past because of the infrastructure that’s come from the Internet and Wi-Fi, but the biggest change has been cost,” said Wade Wessels, sensing and IoT business leader, systems and applications, at Honeywell.
Wessels points out that the dropping costs of sensor technology and the increase in computing power has expanded how much and what kinds of information manufacturers can collect from machines.
“With things connecting to the Cloud and data getting cheaper, we can scale in a way we haven’t seen before,” Wessels added. “Not only can I have a connected system across a factory, but across factories.”
The Industrial Internet of Things and the Connected Factory
ARC Advisory Group defines the IIoT in their whitepaper, Process Automation and the IoT: Yokogawa’s VigilantPlant Approach to the Connected Industrial Enterprise, as being comprised of four key parts:
- Intelligent assets
- A data communications infrastructure
- Analytics and applications to generate information from raw data
“Intelligent assets” are defined as sensors, controllers, application software and security components capable of built-in intelligence, self-diagnosis capabilities, connectivity and support for analytics.
(Image courtesy Sandvik Coromant.)
Intelligent assets and edge devices send information directly to data communications infrastructures such as the Cloud, where it is turned into actionable information on overall equipment effectiveness (OEE), for example.
This actionable information can be used for predictive maintenance, as well as comparative analysis and process optimization.
Data collected over time can be combined to find patterns indicating greater faults or process limitations. These patterns can also be used to analyze how a human worker or robot performs a task.
“You can feed information back into robots to change what they’re doing, or change how someone runs a particular machine – maybe you’re actually overloading a machine and not even realizing it, but you can get that information in feedback,” said Wessels.
Additionally, employees can master IIoT platform interfaces without an increased technical knowledge or extensive training. This is a significant advantage for today’s shrinking skilled workforces, as more of the previous generation’s experts retire.
Ideal IIoT platforms include tools which allow engineers to visually explore data and decrease the need for understanding complex algorithms or programs, when trying to solve problems. This reduces the reliance on Information Technology (IT) expertise.
(Image courtesy Sandvik Coromant.)
Information can then be institutionalized, keeping critical knowledge within the workforce despite changing faces, skill-sets and levels of experience. This could even include useful tips and tricks from external experts.
With the IIoT, data can be sent to experts, no matter their location – if a machine breaks down in “Factory A”, a tech expert in “Factory B” can help those at “Factory A” fix the issue. Information can also be sent out to suppliers or partner companies responsible for the maintenance of machines.
However, every factory and plant is unique in its application requirements. This means IIoT platform architectures may vary significantly between facilities with different processes.
“It’s all about what you want to measure and what you want to see,” explained Wessels.
“With intelligence, scale and cost, you can bring together more pieces of information to make decisions. For example, how clean the air is in a factory that makes silicon is directly related to temperature and humidity, which is directly related to their yield. It’s not ridiculous now to deploy sensors in the factory that monitor particulate content, humidity, temperature, barometric pressure, et cetera and you can feed those insights directly to an environment control system. This system will make sure the factory air is always clean, at the proper humidity, temperature and pressure so you get the maximum yield from your fabrication line.”
Working with the IIoT
We know the IIoT is designed to collect info and be easy to interact with, but how does it change the engineer’s or technician’s experience on the shop floor?
Sensors will become commonplace, but the key to mastering the technology is not in the data collection itself, but rather in how a manufacturer derives insights from that data.
Jeff Rizzie, senior manager of business development at Sandvik Coromant stresses this point when approaching the IIoT from a machining application standpoint.
“Today you might want to take data from something as simple as whether the machine door is open or closed, whether the spindle is turned on or off, what RPM it’s running, the horsepower of the spindle or the temperature of the spindle bearings,” Rizzie explained. “We're even starting to add sensors in tools. It gives you insight and the chance for decision making on issues that operators would not have detected quickly enough or know what needs to be done.”
CoroBore Tool connects to mobile devices through Bluetooth, streaming performance data while the machine is active. (Image courtesy Sandvik Coromant.)
The collection of basic data across varying brands of CNC machines has been universally standardized through platforms like MT Connect, Rizzie explained. For more detailed information, Rizzie stressed connecting to the PLC of the machine:
“If you really want to do adaptive control, then you must plug into the PLC of the machine for access to higher level data, much faster. Think of the data coming from the machine tool in three different levels. You can have a hot stream of data to take immediate action. A warm stream of data is not as critical in real time and is instead used for historical analysis. A cold stream of data can be taken from multiple machines or devices and moved to the Cloud and isn't critical in real time, so it can be saved for enterprise-level analysis.”
However, not everyone in the facility will need the same information. For example, what the machine operator will need to know versus what the operations manager will need to know will not always be the same.
Machine operators will most often want to know why a machine misbehaves or stops running, such as if a machine overrides a command due to an overload or other anomaly or if it needs repairs. If a machine suffers from significant downtime due to repairs, an operator can determine trends through an operator portal in the HMI or a connected mobile device, for example.
“With an operator portal, you can categorize downtime to determine when a problem is not an operator issue,” said Rizzie. “This can help you point to greater issues when speaking with a plant manager.”
The IIoT Does Not Equal “Big Brother”
Because this information is available to plant managers, the technology may create a “Big Brother” mentality among operators, believing they will be monitored just as closely as the machines.
Rizzie counters this idea.
“This is here to help provide you insight and knowledge about what’s going. You may get some pushback from operators saying their bosses will know every time they open the door or the machine’s not running and they get a little nervous about that. We're not bringing this technology in for your managers to keep an eye on you, but to help you run the operation better.”
Despite this, Rizzie has experienced customers who installed an IIoT platform saw about a 10 percent increase in production across their machines after installation.
“Machine operators just become more aware of themselves – that's not the design of the IIoT platform, but it's a benefit. The real goal is to look for certain anomalies.”
Rizzie explains how one customer that partnered with Sandvik Coromant to connect their facility to the IIoT used their new monitoring capabilities to determine the root cause of inconsistent machining times. Sometimes a part would take 15 minutes to machine, sometimes an hour and never with any consistency.
The root cause was discovered to be no fault of the operators, but instead was due to frequent machine starting-stopping.
“The machine was just stopping itself for no purpose and this feature was likely put in there when the program was originally written and just slipped by troubleshooting.” Rizzie explained. “All that had to be done to get the machine going was that the operator would press the button to get the program running again. If the operator was standing by the machine when this happened, they would press the button right away, but if they were doing a setup on another machine, they weren’t around to get things moving again.”
A machine operator analyzes readings from a CoroBore connected machine. (Image courtesy Sandvik Coromant.)
Sandvik Coromant offers it’s CoroPlus system to connect different areas of a value stream, using software and hardware to connect CAM programmers, process development engineers and others to the information they need from across their machines.
“We’ve basically developed some software with an API that connects directly to your CAM system, so you never leave your environment,” Rizzie explained.
“With the CoroPlus Tool Guide and the CoroPlus Tool Library, it essentially becomes a module to your CAM system or your manufacturing systems. The other part of the platform is what we call machine connectivity. We’ve introduced three products underneath that, based on sensor intelligence.”
The CoroPlus system can be used with CoroBore tools to streamline performance information via Bluetooth to iPhone, iPad or Android devices. This Bluetooth connectivity can also be added to machine controls.
“You can literally connect to these tools with an app and pull data from them while you’re machining, like vibration, temperature and a variety of other things,” Rizzie added.
Hacking IIoT-Connected Factories
By now, you must be asking “Are IIoT-connected factories at risk in today’s digital-age?”
Everyone with a laptop and an Internet connection has the tools they need to invade the privacy of others with easily available hacking software and coding knowledge. Factories seem no safer from this threat.
In 2012, Saudi Aramco, one of the worlds largest oil companies, was attacked with the Shamoon virus. The monster computer bug partially wiped or destroyed 35,000 computers, overriding their hard drives with junk data and seriously harming the company’s ability to operate.
The Shamoon virus reappeared as recently as January 2017, at the time of this article’s writing, provoking Saudi Arabia’s Computer Emergency Response Team (CERT-SA) to issue a warning fr companies to be on alert.
This sort of attack is only one of several possibilities, said Shankar Somasundaram, senior director of IoT at Symantec.
Large-scale attacks like the Shamoon virus may be more common between hostile nation states or be politically motivated, Somasundaram explained.
“If you’re a nation state and you want to bring down another nation, you can effectively attack their big manufacturing plants and cripple some of their critical infrastructure. What I think we’re going to see more of over time is ransomware attacks on manufacturing plants.”
Somasundaram explains ransomware attacks as seizing information, or taking control over production capabilities or other hardware and software, paired with a demand for payment within an allotted amount of time – essentially holding manufacturing plants hostage until the criminals are payed for it’s release.
Having attended conferences such as the Black Hat briefings, where IT, security and corporate experts meet to discuss and learn about cyber security and hacking, Somasundaram learned unintentional attacks are just as likely to occur, if not more so.
“At these conferences, people will say they landed on a device without knowing where they were,” Somasundaram explained. “I would say motivations can change, but a lot of cyber-attacks we see now are, I believe, unintentional.”
Furthermore, Distributed Denial Of Service (DDOS) attacks can be used to overflow traffic to a targeted source, overloading it and shutting it down.
“Industrial plants aren’t very robust,” Somasundaram said. “If you flood a network with a denial of service at one point, you could crash half of the equipment in a manufacturing plant.”
Cyber espionage tactics can even be used to infiltrate a manufacturers system to spy – quietly sniffing around your database, copying sensitive information for nefarious uses. These kinds of attacks can be in-and-out ordeals or a virus could be placed and left collecting data quietly for months or longer. You would never even know a hacker was there.
What’s worse, an isolated infection can easily and quickly spread to other systems, factories and networks through file sharing through USB drives and emails.
“Cyber attacks are a real danger for everyone,” Somasundaram said. “Bigger manufacturers may have brands that attract more risk, but SMEs dependant on their suppliers, with a smaller IT staff, have a high risk with architecture and security that may not be as well developed. From a core security perspective, every manufacturer is at equal risk.”
Security for IIoT-Connected Factories
Despite the horror stories and the possible what-if worst case scenarios, the IIoT, the vulnerabilities it creates and the security required to properly protect oneself need to be kept in perspective.
Rizzie warns manufacturers not to be scared off by hacker horror stories and not to let legitimate concerns turn into paranoia.
“I think what happens is that we get hung up on cyber security,” Rizzie explained.
“Is it possible someone can influence how your machines work or steal data from your machine and learn how you’re making a part? Yes, but ask yourself how realistic an issue is that for you? Cyber security is hugely important, but we shouldn’t let that stop us from taking the first steps of doing some basic OEE analysis, where you don’t have to go outside of your own network and move forward from there.”
Whether your network is isolated to a single factory or distributed across several facilities, there are some basic security precautions a manufacturer of any size can take to ensure their system doesn’t get infiltrated or infected.
Manufacturers should be wary of where they download software and how information is passed around. These precautions should even stretch to suppliers and other partners to ensure they don’t accidentally become a “Trojan Horse” through an infected USB stick, for example.
Somasundaram recommends working closely with your partners to ensure that everyone’s security precautions are in top shape and all devices moving between facilities are clean of any suspicious files or software.
“Take some basic processes and make some rules to set in place for suppliers and employees, like a list of devices allowed in the facility or required security software installed on all devices, keeping USB sticks clean and even putting additional levels of monitoring in their plant. This way, if something goes wrong, they are immediately notified.”
Somasundaram points to the NIST Cybersecurity Framework and the 62443 series of standards for industrial automation and control systems security as effective guidelines on what security precautions a manufacturer may need.
“Because most frameworks like these are just guidelines, you must look at the practical aspects,” Somasundaram continued. “How do you enforce these frameworks and work with them to make sure they are tactically feasible.”
For more information, visit the Honeywell, Sandvik Coromant, and Symantec websites.
To learn more about where you can start with the IIoT, continue reading with us about the Top 5 Assets You Need to Connect to the IIoT.