How five New York SMEs are putting Industry 4.0 to work—and why it's working

Picture it: A pristine, white warehouse where smooth, shiny robots silently move boxes and assemble products, all while a few lab-coated experts tap on tablets as they observe the fully automated, digitally synchronized orchestra at work.

If this sci-fi scene comes to mind when you hear “Industry 4.0” or “factory of the future,” you’re not alone. Since these trendy buzzwords first appeared, they’ve held a certain wow factor. But, as the world becomes digital inside—and outside—the factory, Industry 4.0 is less about what will be and more about what’s already happening. A catch-all for a large (and growing) array of tools and processes, Industry 4.0 is fundamentally about leveraging data to make manufacturing operations more effective using digital technologies. The most common of these are 3D printing, the Industrial Internet of Things (IIoT), cloud computing and analytics, robotics, and artificial intelligence (AI) through machine learning.

“Very often our first step is to help organizational leadership understand the expanse of Industry 4.0, busting the myth that it is out of their reach. The reality of building up Industry 4.0 capabilities is a lot more practical than many manufacturers who are new to it expect,” said Ajay Khaladkar, who is the technical program manager of advanced manufacturing and Industry 4.0 solutions at the Golisano Institute for Sustainability (GIS) at Rochester Institute of Technology (RIT).

Khaladkar leads RIT’s Industry 4.0 Transition Assistance Program, which works with smaller manufacturers—formally known as small and medium-sized enterprises (SMEs)—in New York State. The program is funded by a $3.75-million-dollar matching grant from New York State’s Empire State Development.

In 2021, the program received a $1-million-dollar-plus grant from the U.S. Department of Commerce through the Coronavirus Aid, Relief, and Economic Security (CARES) Act. Over the course of a year, this money was used to fund assistance projects for SMEs in the Central and Finger Lakes regions of New York State. Its goal was to build resiliency among manufacturers recovering from pandemic-driven challenges in these areas through the implementation of Industry 4.0.

The first step

“Digital maturity is a measure of an organization’s ability to both process and track data in real time, and how they use that data for quick, smart decisions to drive improvements,” said Scott Walker, who is the vice president of U.S. operations for INFICON Inc., an advanced technology manufacturer.

To determine INFICON’s digital maturity, the RIT team performed a site assessment of its facilities in East Syracuse, New York. This provided a baseline from which to build a roadmap for the firm that identified realistic opportunities for digitalization, a list that can be more than 100 recommendations long, depending on the complexity of a company’s operations.

Explosive growth in the semiconductor industry was driving a significant increase in demand for INFICON’s products. The required production ramp up meant a need for more frontline associates, so the first task that Walker chose to pursue was to modernize the firm’s work instructions in order to onboard new staff faster.

Walker engaged RIT a second time to transition INFICON’s work-instruction system to a digital platform. RIT’s team began by studying current challenges stemming from the company’s existing system in light of assembly, quality control, and testing. They then looked to better understand how INFICON’s work instructions were created and edited, starting with engineering and following the work flow through to floor release. Next, they documented all available data sources before sketching out ideas for an improved future process.

The RIT team created a list of options based on key—and often competing—priorities in line with INFICON’s business goals. For example, while all potential systems leveraged augmented reality (AR) capabilities that would allow users to overlay instructions onto live tasks, some used 2D rather than 3D imaging. Another important consideration was data security: Is information stored on a local server or is it cloud-based? Yet another factor came down to wearable equipment: What would work best for INFICON staff at a reasonable cost?

“By shifting to digital instructions we plan to improve our overall productivity and effectiveness,” Walker explained. “We want our digital instructions to be able to link directly to CAD (computer-aided design) models, be able to show both pictures and videos of processes, link directly to supporting documents, and interact with our ERP (enterprise resource planning) system.”

INFICON plans to implement one of the solutions that was discovered—and, importantly, validated—through RIT’s analysis as part of its Industry 4.0 roadmap.

RBW, a lighting design and manufacturing company headquartered in Kingston, New York, first explored what Industry 4.0 might look like for them in 2020. At that time, the company’s leadership anticipated purchasing high-end automated robotic systems as their first step towards digitalization. But that view changed after the firm began working with RIT to create an Industry 4.0 roadmap.

The roadmap eventually allowed RBW to improve its plan to transition to a 100,000-square-foot facility with eight times the space of its original Brooklyn location. With RIT’s recommendations and a larger floor plan, RBW was better positioned to begin pursuing digital automation of shop-floor processes. The firm also looked to automate enterprise-level activities, such as strategy and governance, workforce learning and development, and enterprise intelligence.

“We have a newer appreciation for automation in the broader context of repetitive work (of any type),” said Theo Richardson, RBW’s director of innovation, after the RIT Industry 4.0 team completed a full assessment of its operations. “We should focus our automation efforts elsewhere (data, office workflows) and explore cobots at a later stage. We are confident that the I4.0 assessment criteria we developed with help from RIT will lead us to in that direction at the right time.”

RIT’s assessment gave RBW the fresh perspective and information it needed to set in place realistic steps for connecting what was happening on the shop floor—such as product assembly—with other business activities, like design and inventory. Of note, technology and equipment upgrades were not at the top of RBW’s to-do list; while impressive leading-edge technology is without doubt a necessary lever of digitalization, there’s an even more important thread tying everything together: data.

Data put to work

For much of its 88-year history, employees at Giovanni Foods Company Inc. have recorded downtime incidents manually. Over that time, the way data was collected evolved from pen and paper to spreadsheets on a desktop computer where it could be tracked and analyzed. The resulting data records were far from perfect; staff used different terms for the same problem, entries were left unfinished because the operator was interrupted by something more urgent, or incidents weren’t recorded at all. 

This story isn’t unique to Giovanni; gaps and inconsistencies in operational data like these have long plagued plant managers and engineering staff across manufacturing. These often obscure the root causes for costly downtime and, consequently, make it hard for companies to plan for it.

Enter Industry 4.0.

Nick Leshkiv, Giovanni’s director of plant operations, saw an opportunity to apply Industry 4.0 to not only better track equipment downtime, but to predict when disruptions might happen. Using intelligent sensors that could communicate vibration, heat, or humidity data to an automated monitoring platform, Leshkiv wanted to make Giovanni’s assets “smarter.”

But to do that, Leshkiv first had to make sense of a year’s worth of downtime data that he had collected from the production lines at Giovanni’s facility in Baldwinsville, New York. Gerry Hurley, a technical program manager at RIT, worked with Leshkiv to “scrub”—a term for removing errors, biases, and inconsistencies from a dataset—the data. Hurley organized it into 30 unique categories and then assigned codes detailing the operators’ stated reasons for downtime, such as “breakdown repairs” or “lunch breaks.”

The data deep dive was just the beginning of an ongoing collaboration to discover Giovanni’s most cost-effective path to data connectivity, a pillar of digital manufacturing. RIT’s assistance has opened Giovanni’s management eyes to a wider view of how its equipment performs over time and why individual units go offline. With this improved visibility, the Giovanni team is currently selecting the assets that could be successfully integrated into a performance-monitoring system.

Importantly, Leshkiv intends the final product to be turnkey, using off-the-shelf components and open-source software. “I don’t want it to be only proprietary to Giovanni or somebody else. I want everybody to be able to use it,” he said.

This sentiment reflects one of the RIT Industry 4.0 program’s driving purposes; that is to foster innovation at the company level and to then translate it into knowledge and resources that are of benefit to New York’s manufacturing sector as a whole.

It's what you do with it

Faradyne Motors LLC, a manufacturer of submersible motors for water and fuel pumps, already had some automation technologies in place on its production and assembly lines when the RIT team reviewed its operations. The assessment uncovered opportunities for integrating Faradyne’s operational and technical process data into a global, unified information system, which is key to realizing “data connectivity,” a core Industry 4.0 concept. Such an improvement would allow the business’s managers to easily access data about core processes—like planning and scheduling, production, assembly, or final inspection—and quickly interpret it.

RIT and Faradyne are currently working together to identify areas where data-harvesting technologies, like sensors and edge devices, could be installed in the manufacturer’s operations. Once this landscape is mapped out, the two plan to explore how changes in Faradyne's production process might increase the quality and quantity of data that could be collected.

“Industry 4.0 is more about data-driven intelligence than it is about using any specific technology,” says James Meyer, a senior manufacturing engineer at RIT who worked on the Faradyne project. “It’s about how technologies are used to access data and make sense of it to make smart choices.”

Meyer’s initial analysis resulted in a decision matrix that outlined different steps that Faradyne could take, depending on weighted factors. These possibilities were ranked according to criteria like cost, data availability, security issues, and time. Ultimately, RIT’s goal will be to use the decision matrix to steer Faradyne’s leadership in designing a roadmap for achieving data connectivity cost effectively.

“Deploying Industry 4.0 appeared daunting,” said Dante Volpe, Faradyne’s president, after RIT’s consultation. “We were thrilled when we found that RIT was able to help guide us through this process. We are still in the early days of this endeavor, and with their help and expertise, we already see value.”

The more you know

Collecting and harvesting data is just one piece of the digital-transformation puzzle. How data is interpreted or analyzed also presents significant challenges for companies without staff dedicated to Industry 4.0 implementation.

Linton Crystal Technologies manufactures Czochralski furnaces, which are used to grow silicon wafers for the semiconductor industry. Even small deviations during the complex process can be costly.

Todd Barnum, Linton’s COO, is always in search of ways to differentiate Linton’s equipment in order to maintain traction in a fast-moving, increasingly competitive global industry. He partnered with RIT to discover whether data generated during the furnace’s operation could be used to create a model that would predict final part defects and prevent wasted production effort.

Abu Islam, the lead staff scientist at RIT who heads the project, is working with Linton to determine if machine learning could be used to develop the model. The approach being evaluated uses a convolutional neural network (CNN), which is a deep-learning method commonly used to analyze images. Islam is exploring how a CNN could be trained to use vision-system images to give Linton’s customers real-time visibility of the silicon-growing process. This would make it possible to anticipate potential failures and make adjustments to avoid costly errors.

“Deep-learning methods can be very reliably trained to detect facets in the growing silicon and, during runtime, can give alerts to the operator,” Islam noted.  

Future-proofing

The RIT Industry 4.0 Transition Assistance Program was launched in 2020 as an initiative of the Center of Excellence in Advanced and Sustainable Manufacturing (COE-ASM), which is based at RIT. Since its inception in 2011, COE-ASM has drawn on its Industry 4.0 expertise to develop solutions for manufacturers and to contribute as a partner to organizations like the Manufacturing USA institutes. The dedicated transition assistance program has allowed the center to dramatically scale up its efforts to drive digitalization among small- and medium-sized enterprises (SMEs) in New York State.

“We increasingly noticed common challenges that many smaller manufacturers had, such as access to labor, or the ability to resolve complex systemic manufacturing productivity and quality issues. Most SMEs don’t have the resources to connect these challenges to cost-effective technology solutions that facilitate automation and data-driven decision-making,” said Michael Thurston, COE-ASM’s director and also a research faculty member at RIT. “The need for Industry 4.0 and simultaneously the lack of Industry 4.0 literacy and competency was a recurring theme.”

SMEs lag behind larger, multi-national firms when it comes to Industry 4.0 adoption. They often lack the resources and expertise to support the development and deployment of new technologies. This adoption gap not only limits the growth potential of SMEs individually, but the manufacturing sector as a whole, too. And this matters: SMEs are the backbone of the U.S. industrial supply chain, accounting for 99 percent of all businesses in the United States and 44 percent of its total economic activity.

The RIT program serves another purpose that reaches far beyond any one company’s walls: future-proofing U.S. manufacturing for the global market. Over $900 billion are set to be invested in the digitization of manufacturing over the next five years by companies and governments looking to make the digital pivot.

Many countries have launched national strategies to close the digital-manufacturing implementation gap between SMEs and larger firms. China and South Korea already have technology-readiness initiatives in place—like Made in China 2025—to prepare their manufacturing sectors for Industry 4.0. More established industrial leaders, like Germany and Japan, have put in place policies—Industrie 4.0 and Society 5.0, respectively—to accelerate the adoption of Industry 4.0 among manufacturers.

A similar national effort doesn’t yet exist in the United States. But state-driven initiatives like RIT’s or Michigan’s Automation Alley are already showing how federal resources and expertise can be delivered at the regional and local level to strengthen the foundation of the U.S. manufacturing economy.

The RIT program will end a three-year pilot period in December 2023. Khaladkar and his team have used that time to distill the SME pain-points and needs they’ve discovered, analyze the solutions landscape, and adjust the program in anticipation of a more permanent, longer term initiative.

“Now we have a much deeper understanding of the challenges faced by manufacturers when it comes to digitalization,” observed Khaladkar. “Our foundation is solid and we look forward to scaling up to meet a need that will only become more urgent as time passes.”