The concepts and technologies encompassed by the term “Internet of Things” are rapidly changing the world. Sometimes called “Industry 4.0” or the “Industrial Internet,” it is the idea that manufacturers will be able to connect everything in their operations, allowing information to be easily exchanged and enabling more autonomous plants.
However, to some extent, the objectives of smart manufacturing are not much different than traditional manufacturing goals. Processors will still strive to produce higher-quality products, give better customer service, increase productivity and optimize operations.
“On the surface, smart manufacturing plants will look very similar to today’s plants,” says Katie Moore, industry marketing manager for GE Digital. “The difference will be the underlying digital thread that seamlessly ties the flow of information from design through manufacturing to end-consumers including the lifecycle of the [food] product.”
Although no single, formal definition for smart manufacturing exists, some companies have come up with their own. “At Cisco, we like to refer to it as the Internet of Everything [IoE] because it encompasses so much more than just connecting ‘things,’” says Shaun Kirby, chief technology officer for Cisco Consulting Services. “The Internet of Everything is the networked connection of people, processes, data and things. By connecting sensors, machines, people and processes and applying real-time analytics to the data, food and beverage processors are able to gain intelligent insights they can use to improve all aspects of the supply chain, from how food is grown, to how it is processed, packaged and delivered to the store shelf, to how consumers purchase it.”
Actually, the word “Internet” might be a bit of a misnomer, since the connectivity of everything does not necessarily require access to the web. Consequently, ARC Advisory Group prefers to refer to IoT as the “digital transformation of plants,” according to Greg Gorbach, vice president of ARC.
“More machines and production systems are connected today, and even more will be connected in the future, but the Internet isn’t required for that transition; there are many other connection options,” says Michael Risse, vice president of Seeq. “The value is the connection—wired, cellular, Bluetooth, satellite, sigfox, ZigBee—not the specific protocol of it.”
Preparing for the future
To consider the evolution from traditional manufacturing to smart manufacturing, just look at the cell phone. In a relatively short period of time, phones have gone from being used only for calls to helping us control multiple aspects of our lives, such as remotely adjusting the thermostats in our homes. The question is: How do we make our facilities smarter?
“This requires the integration of existing, or wired, control and operations data with new wireless data from sensors,” says Risse. “Food and beverage companies can start getting smarter now with the data acquired onsite via sensors, connectivity and analytics.”
Because sensors can collect so much valuable data, they have become more pervasive throughout the value chain. When this data is brought into a central analytics environment, the result is better decision-making, regulatory compliance and more.
“Industrial automation has, for 30-plus years, allowed processes and people to be integrated,” says Maryanne Steidinger, director of product marketing for Schneider Electric Software.
For instance, manufacturing execution systems (MES) tie automation and business systems together, allowing the management of labor, inventory and equipment. Additionally, a human machine interface (HMI)/visualization system can offer the ability to monitor and control machines and processes. Providing graphical overviews of processes can make it easier for operators to understand them at a glance and manage any out-of-spec conditions.
“All the necessary equipment for IoT architectures is widely available, but two key components that are often forgotten are data acquisition and machine-machine connectivity,” notes Richard Clark, controls, process and security engineer for InduSoft. “Data must be collected and pushed to the cloud for it to be used effectively.”
To solve this problem, InduSoft recently released InduSoft IoTView, an HMI platform that is specifically designed to embrace new technology for connectivity. As the number of small devices used to collect data increases, Clark says so too will the demand for an easy way to send the information to the cloud for storage and analysis while also connecting the machines to each other. IoTView is an expansion of InduSoft’s HMI software with characteristics designed to enable the creation of connected intelligent systems.
Plant “personnel can evaluate what digital backbone will best serve the needs of the facility and begin working on the required upgrades,” says Ola Wesstrom, Endress+Hauser senior industry manager of food and beverage. However, to upgrade equipment and raise a plant’s level of connectivity, a company must first establish what aspect of the production process it is trying to improve through the use of additional data.
“When evaluating new equipment, manufacturers should think about what the piece of equipment will produce and determine what critical data and information are necessary to improve the process,” says GE’s Moore. “For example, if a bakery operation is looking to buy a new oven, it must consider not only the capacities, speeds and zone temperatures, but the things that it may not be measuring today, like humidity and airflow.”
“The value of data from the equipment itself is greatly enhanced by combining data from operational systems and the Internet through sensor fusion, which combines information from multiple sources to derive new insights or enable new ways of automating and streamlining business,” says Cisco’s Kirby. To take advantage of new sensors, data sources and algorithms, organizations should look for software that can be enhanced or evolved rapidly. “Proprietary, closed or stagnant software products will not provide the agility required to keep pace with the growing list of opportunities IoE creates to change the business,” adds Kirby.
Moore agrees open, scalable solutions are important to fueling growth opportunities. “The typical production environment holds automation and systems from multiple vendors, with varied adherence to standards like OPC-UA, ISA-95 and/or ISA-88,” she says. “Connecting disparate systems and equipment with an open and layered approach allows a company to take advantage of collecting critical, timely data from the shop floor.”
This information exchange allows a company to better understand its production capabilities and gain insight into data that could drive efficiency improvements, ensure food safety and mitigate risks. However, a company can only take advantage of connectivity to the extent it has employed technology in its plant. For instance, is there network connectivity? Is that network secure? Does it make data available?
“Manufacturers should begin by assessing the level of information technology [IT] and operational technology [OT] convergence in their plants,” recommends Dave Sharpe, director of consumer packaged goods industry for Rockwell Automation. This convergence is what Rockwell Automation calls The Connected Enterprise, in which “IT and OT make use of a shared architecture to establish a network for validating performance as more devices are connected, and improving how a manufacturer collects and shares relevant information across the enterprise.”
Processors also should invest in bandwidth to handle the volume of information needed across the plant, says Ron Pozarski, senior industry manager for Emerson Process Management. Plus, more intelligent field devices may be needed to provide additional data. Implementing controls systems will also provide advanced capabilities.
Data management
Of course, this digitization of manufacturing is heavily dependent on data collection. However, as a popular saying in the IT industry goes, some companies are “data rich, information poor.”
“Collecting data for data’s sake is not valuable,” says Moore. “It must be actionable, timely and delivered to the right stakeholders.” She advises a company starts by creating a data collection strategy linked to one or two key benefits, such as improving production efficiency, ensuring product quality or driving down energy costs. Then, the appropriate departments should receive the filtered data presented in an easily understood format to inform their decisions.
Moore recalls one well-known food manufacturer that was experiencing production inefficiencies due to manual reporting and a lack of visibility into its data. After the company started to use GE software to view real-time production data and make adjustments to production runs, it realized a 10 percent improvement in overall equipment effectiveness (OEE) over a seven-year time period. The company also saved $300,000 through production improvements, increased data visibility and higher efficiency.
“The key is using the data to provide insights for driving the optimization of outcomes,” says Seeq’s Risse. “Data analysis software that can work with a lot of data from batch or continuous process operations and many different sources is necessary to deal with the challenges of the data deluge from IoT and Industry 4.0.”
He stresses the importance of this software offering rapid insights employees can act on without the assistance of a data specialist for interpretation. Also, the application should present a variety of data in a useful and relevant context.
“Manufacturing and production data can come from many sources including controllers, HMIs, historians, asset management systems, production planning systems, etc.,” Risse says. “The data analysis software solution must be able to mix and match data from all of these and other sources in one visualization environment to enable quick analysis.”
Additionally, the software should allow access to information from any platform, including mobile devices and PCs, and enable insights and discussions among employees at different levels, as well as real-time decision-making.
“Manufacturing information is the basis for the deeper analyses used to improve operations,” explains Sharpe. “Integrated with the system-wide data flow, the FactoryTalk software suite from Rockwell Automation can provide visibility into the performance of each production line and how it fits within the plant and enterprise. Data can be viewed in real time on KPI dashboards and shared between users, teams and departments.”
This data also can help provide insights into a production OEE, which is measured by the availability of a machine or uptime, its performance or throughput, and the product quality.
“By deploying Industrial Internet of Things systems with predictive analytics, companies can schedule machine repairs in advance when failures are predicted, rather than encountering unplanned downtime when the machine fails,” says ARC’s Gorbach. “This increases uptime and therefore OEE. In principle, it could be applied to all kinds of industrial assets, such as conveyors, pumps or packaging machines.”
Endress+Hauser’s Proline flowmeters have a self-monitoring feature the company refers to as “Heartbeat Technology.” According to Wesstrom, this feature ensures up to 98 percent reliability. Additionally, the meters transmit variables that can be used to detect process anomalies, signaling a potential equipment problem.
“This type of data collection is at the heart of a preventive maintenance program,” says John Neeley, senior product manager for Fluke Connect. “If data can be collected and quickly analyzed, warning signs can be turned into work orders that allow scheduled maintenance rather than expensive, unplanned downtime.”
“For instance, one way to determine if a motor is starting to degrade is by testing the insulation on the windings,” Neeley says. “Tools are now available that can perform insulation tests quickly and accurately, and then transmit the readings over the Internet to the maintenance database. The result is a visual dashboard that displays periodic test results, and the motor is then flagged for service when the insulation readings degrade below a set threshold.”
Factories of the future
Due to the declining price of powerful computers, data storage and devices providing connectivity, IoT technology is becoming more ubiquitous. In fact, many food and beverage processors are already taking advantage of this digitization of industrial systems. For instance, Ohio-based SugarCreek, a processor of bacon, meatballs, sausage patties and chicken for the foodservice and retail market, recently used Cisco IoT technologies to connect devices, sensors and systems in its refurbished 418,000-sq.-ft. facility in Indiana.
Edward Rodden, chief information officer for SugarCreek, says the converged network infrastructure has generated cost savings and allowed SugarCreek to set up a way for vendors to remotely and securely access their machinery. The wireless network also enables the tracking of workers through the radio-frequency identification (RFID) tags on the bump caps all of them are required to wear in the facility.
The RFID tags offer three big benefits, he says, the first being the safety and security of the plant’s workers. In the event of an emergency, responders would be able to know and locate any employees still inside. Also, because the movement of workers can be followed, the company can gain more insight into how to best design jobs and workflow. Additionally, the company will be able to track labor and respond to staffing needs on an actual basis.
“The wireless [system] also allows us to extend our IT network to the plant floor and make use of Cisco Jabber,” says Rodden. (Jabber is a communications application for mobile devices.) “We’ve cut between $350,000 to $500,000 in expenses related to cell phone boosters because we can use wireless calling.”
Another Cisco client, Barilla, a maker of pastas and sauces, introduced an integrated tracking system called the Safety for Food (S4F) initiative to provide the company with more visibility into its supply chain. The system also helps consumers understand where ingredients have originated and how the end product reached the grocery store.
“Using data virtualization, intelligent networking and mobile technologies, Barilla has been able to integrate real-time data from external suppliers, distributors and retailers, and apply real-time analytics for compliance checking, tracking and tracing all its ingredients,” notes Kirby. Through this IoT technology, the company also can predict how certain decisions will affect specific steps of the entire process by measuring productivity and efficiency.
“Because the IoE allows manufacturers to measure operations remotely, travel time is removed from the equation, and they can respond almost instantly to course correct on a new product launch or seize an emerging opportunity,” adds Kirby.
OEE information can help processors sift through performance metrics and understand their plants’ production issues. For example, when Rockwell Automation client Barry Callebaut, a manufacturer of chocolate and cocoa products, wanted to address material usage variance, it pulled manufacturing information from several areas in the process, including raw materials and packing. The real-time OEE information allows the company to verify the correct quantity of materials goes into each product and react immediately to any individual box that is off weight.
“An OEE application lets food and beverage manufacturers drill down to root-cause analysis,” says Rockwell’s Sharpe. “Barry Callebaut focused on inefficiencies in a batch-filling process. Reducing each batch process by just two to three minutes results in significant time savings when running hundreds of batches per day.”
Organizational and workforce changes
With equipment getting smarter and real-time information becoming so readily available, it is no wonder IoT technology will have a major effect on manufacturing decision-making. Plus, the more immediate actions required for data-driven decisions might well revolutionize plant organizations, with top-down management losing favor.
“The point of [identifying issues] ‘sooner’ is to eliminate the approvals or reviews that turn ‘sooner’ into later,” says Risse. “This requires training and trusting front-line production personnel, as they are the ones best placed to make the quick decisions that lead to continuous improvement.”
SugarCreek already employs a high-performance work structure in its IoT plant, with employees put into groups that have minimal supervision. “All the skills are contained within the teams,” Rodden says. “For instance, we don’t have dedicated maintenance people responding to problems on the floor. Those [maintenance] skills are embedded within the teams.”
The model is highly dependent on collaboration and communication, but what enables it to work is having access to real-time data and analytics. The larger idea, Rodden says, is to empower every team member to make informed decisions and give each not just a job, but a career. This shift in thinking has changed the hiring process at the plant, with teams, not human resource managers, interviewing applicants to fill positions requiring higher skill levels.
Since smart manufacturing might require a more technologically savvy workforce, more programs are being created to address the skills gap. For example, Cisco and Rockwell Automation have developed online and in-person training courses that help employees learn how to install, operate and maintain industrial networks.
However, Cisco’s Kirby says devices are becoming easier to connect and program without deep technical skills. Since the more routine decisions will eventually become automated, he predicts this will free up workers to concentrate on the more creative, high-level areas of the business.
“Critical skills in this new world are clear, logical thinking about how to solve business problems, the ability to iterate quickly through successively better solution designs and a willingness to experiment with bold new ideas and learn quickly what works best,” he says.
“As sensing becomes more complete, and predictive algorithms improve, more and more operating decisions can be automated,” adds Kirby. “We foresee decision-making steadily evolving from fully manual methods to decision-making that is guided and advised by sensory systems with predictive algorithms to fully automated decision-making with integrated workflows for human approval at critical points.”
Soon, the machine-to-machine connectivity will allow equipment to make autonomous decisions using artificial intelligence. Machines communicating with each other could lead to “automatically ordering more materials when necessary or automatically configuring machines for different products without the need [for humans] to enter the information through an interface,” says InduSoft’s Clark.
Cybersecurity
How will smart manufacturing and the rising the level of connectivity in food and beverage plants affect cybersecurity?
Installing “business IT” types of security, such as anti-malware and host inspection agents, is not enough to stop cyberattacks, says InduSoft’s Clark. He suggests the “deployment of sophisticated systems incorporating security auditing and aggregation tools, specifically designed automation security systems and other methods of intelligent cyberattack prediction and prevention.” Moreover, he predicts insurance regulations will eventually require these intelligent safety-shutdown systems.
Rockwell Automation’s Sharpe recommends a holistic approach to industrial security that extends from the enterprise down to the plant level and out to end devices. He says manufacturers should perform three security steps:
- Conduct a security assessment to understand the risk areas and possible threats.
- Deploy a multilayered security approach, such as defense-in-depth, which establishes multiple fronts of defense, including policies and procedures, as well as physical, network, computer, application and device security.
- Leverage technologies, including software, networks, mobile, cloud and virtualization, in a way that optimizes security.
“The implementation of new technologies, including mobile, cloud and virtualization, involves a unique set of security considerations,” Sharpe adds. “For example, mobile device management should be used to restrict and monitor mobile access to the enterprise, while cloud computing services should offer proven security.”
However, the mere idea of a “secure perimeter” might be an archaic concept with this industrial digital revolution. Because more devices, people and facilities are connected, the canvas to attack is much broader. Cisco’s Kirby says this technology requires a different security architecture, one “that distributes intelligence through the network with a ‘single pane of glass’ for consolidated management.” He explains that, in this way, threats can be detected and stopped immediately using policies managed from one point and applied globally.
“Organizations should be aware that no amount of security can provide a 100 percent guarantee of immunity from attack and be prepared with plans for responding to security and privacy incidents at all stages, from before a suspected possible incident to during a newly detected incident to rapid recovery following an incident,” Kirby says.
However, the potential benefits of connectivity and the capabilities IoT technologies offer are too great for manufacturers to turn their backs on them. Consequently, companies must develop a solid cybersecurity plan now to manage the security and privacy risks, if they have not done so already.
For more information:
Dan Wright, Fluke, 425-347-6100,
info@fluke.com, www.fluke.com
Dave Sharpe, Rockwell Automation, 440-646-3434,
dksharpe@ra.rockwell.com, www.rockwellautomation.com
Greg Gorbach, ARC Advisory Group, 781-471-1000,
ggorbach@arcweb.com, www.arcweb.com, www.industrial-iot.com
Katie Moore, GE Digital, 855-968-7143,
katie.moore@ge.com, www.ge.com/industries/food-beverage-consumer-goods/
Maryanne Steidinger, Schneider Electric Software, 800-SQUARED,
maryanne.steidinger@schneider-electric.com, www.schneider-electric.us
Michael Risse, Seeq, 206-801-9339,
info@seeq.com, www.seeq.com
Ola Wesstrom, Endress+Hauser, 800-642-8737,
ola.wesstrom@us.endress.com, www.us.endress.com
Richard Clark, InduSoft, 512-349-0334,
info@indusoft.com, www.indusoft.com
Ron Pozarski, Emerson Process Management, 952-828-3500,
infocentral@emersonprocess.com, www.emersonprocess.com
Shaun Kirby, Cisco Consulting Services, 800-553-6387,
tac@cisco.com, www.cisco.com
