Getty Images/iStockphoto

Digital transformation in manufacturing melds emerging tech

Partners work with customers to absorb a mixture of AI, IIoT, hyperautomation, 5G and cloud computing, while dealing with the challenge of highly distributed organizations.

Digital transformation in manufacturing calls for a mix of emerging technologies that tackle age-old industrial objectives: boost production, improve quality and reduce equipment downtime.

IoT, cloud and edge computing, 5G networks and AI rank among the IT developments manufacturers hope to harness. Enterprises typically deploy new technologies in combination rather than focus on a single field as the answer to their needs.

"It's not possible to pick just a short list of technologies and declare them to be the next big thing," said Max Ivannikov, solutions consultant at DataArt, a software engineering firm based in New York. "Any innovation project is always a combination of different technologies and thoroughly designed processes aiming to achieve business goals."

Getting there can prove difficult for manufacturing companies, which tend to have limited IT resources compared with other industries. IT service providers seek to fill that gap, offering consulting and implementation services to help organizations adopt emerging technologies. Partners lend expertise in fields such as data science and change management.

The distributed nature of manufacturing complicates their work, however. A large industrial firm may span multiple factories, myriad production lines and varied technology stacks. In addition, individual manufacturing centers operate with high degrees of autonomy. These industry characteristics make it difficult to scale a local deployment across an organization, presenting partners with a technical and cultural challenge.

Emerging technologies in manufacturing: IIoT, cloud and edge

Manufacturers, while a varied group, show some commonalities in technology use. IoT often serves as the foundation for transformation projects. Industrial IoT (IIoT) sensors attached to shop floor machinery generate a wealth of data on temperature, vibration, voltage, acoustics and cycle time, among other factors.

Ozgur Kaynar headshotÖzgür Kaynar

As a result, industrial firms find themselves "sitting on top of a huge amount of data," said Özgür Kaynar, general manager at Analythinx, a data science and managed services company in Istanbul, Turkey.

Adjacent technologies such as 5G will add to those data stores. Ivannikov said 5G provides the sensor density, low latency and connection speed manufacturers want, making even larger amounts of data available from IIoT devices.

Manufacturers tap yet another technology, cloud computing, to gather and store that data for analytics and AI.

"Cloud technologies make it much easier to grab sensor data," Kaynar said. The cloud can also house data from ERP systems, the core business applications for industrial firms. With access to broader data sources in the cloud, manufacturers can move from descriptive analytics to predictive analytics. The resulting benefits include timely equipment maintenance, supply chain optimization and workforce optimization, he added.

Cloud vendors have developed capable components for each phase of advanced analytics, Kaynar explained. Those include sensor data integration, self-service reporting, dashboarding and machine learning (ML).

On one project, works with a large manufacturing company in Turkey to collect sensor data and make it available in the cloud for business intelligence applications and dashboards. The data collected in the cloud helps the company identify anomalies in manufacturing processes and, based on that insight, boost production, Kaynar said.

While analytics takes place in the cloud, it also happens at the edge. Edge computing speeds up analytics for data-driven decisions. That's because the computing approach moves processing close to where data originates on the shop floor. In addition, sensitive data remains on local hardware and doesn't travel to the cloud, bolstering security.

Bruce McKinnon headshotBruce McKinnon

Manufacturers are beginning to use both edge and cloud for analytics, using the edge to gain a better understanding of local conditions and using the cloud for an overview of operations across factories, industry executives said. The data analytics continuum starts at the far edge, which includes programmable logic controllers on the factory floor and myriad sensors, continues to the near edge, where data is aggregated, and onward to the macro edge, which includes on-premises data centers, a remote site or the cloud, said Bruce McKinnon, chief strategy architect at Insight, a solutions integrator with headquarters in Chandler, Ariz. The cloud represents the fourth tier in this intelligent edge, offering compute and storage.

Cloud is well suited to the high volume of big data analytics, McKinnon said. Cloud, however, has higher latency than on-premises, local compute. For that reason, manufacturers should use the near edge for low-latency applications. He cited the example of worker safety, which demands near-instant processing to translate a safety warning into a stop-machine order or notification alert.

AI gains ground

AI and ML adoption have become more prevalent among manufacturers that have more data to exploit. AI and ML let companies automate processes and make them smarter, Ivannikov said. "It's hard to imagine any project nowadays without using AI/ML," he noted.

Paul Lewis headshotPaul Lewis

An increasing number of projects fall into the hyperautomation category. Hyperautomation describes a set of technologies used to scale automation within an enterprise. AI, ML and IoT fuel that trend among manufacturers, said Paul Lewis, CTO at Pythian, an IT services company based in Ottawa.

Hyperautomation runs the gambit from simple task automation for frontline workers, to process automation for production lines, to multi-plant business operation, he noted.

This field also includes process mining, which helps businesses discover bottlenecks in their operations. Accenture, for example, is using Celonis' AI-based process mining offering to improve processes at Mann+Hummel, a German company that manufactures filtration technology. Accenture and Celonis entered a partnership in January 2022.

The computer vision aspect of AI plays a role in manufacturing, with quality management as the pivotal use case. Computer vision, used with augmented reality and ML, can detect anomalies among items coming off a production line with greater precision than human quality assurance inspectors, McKinnon said, citing accuracy rates in the high 90% range for the method.

"The real draw is to improve accuracy," he noted.

Computer vision lets manufacturers compare a part to a known good record, which could be a picture, an assembly drawing or a bill of materials. Data scientists create an ML model for each use case to make that comparison.

Services rendered

Service providers help customers navigate the web of interrelated technologies, providing supplemental expertise to organizations lacking in digital transformation skills.

Max Ivannikov headshotMax Ivannikov

"Often, digital transformation is not a core skill for manufacturing companies," Ivannikov said.

IT teams may also lack skills in particular technologies. A manufacturer may not have any personnel devoted to advanced analytics or data engineering, Kaynar said. A tier-one bank, in contrast, might be able to draw upon 100 to 150 data science and engineering experts, he said.

Partners can address the talent shortfall with digital transformation and technology skills. But they also help customers build their own pools of expertise. offers an Analytical Center of Excellence Enablement service, which works with customers to identify candidates among in-house IT staff for data science or AI engineering roles and then provides two months of training.

The service also establishes a data and analytics strategy, which identifies business problems to solve with analytics, and determines the types of data and technology components to address them. also offers long-term support, supplementing customer personnel with its cloud/analytics engineers and data scientists.

Pythian, meanwhile, works with manufacturing customers to get the most out of their data, Lewis said. That could mean helping them make the right decisions on how to use data and technology, he noted. Or, Pythian may collaborate with a manufacturer to uncover customer insights that generate competitive advantage.

Consulting services -- developing roadmaps for a client's digital strategy -- and technical guidance typify partners' manufacturing industry offerings. In addition, partners translate strategy into an architectural design, build agile processes and provide deployment services. Change management -- helping a manufacturer's employees adopt the newly created digital processes -- also becomes important as transformation begins to take hold.

"When you change the process, you have to retrain people on it," McKinnon said.

Scaling success in distributed environments

An initial digital transformation project usually has limited scope within a single factory or production line. The benefits of advanced technology multiply if they can be scaled beyond the early foothold. That's easier said than done, however.

"The digital transformation process may differ even between two plants of the same enterprise, manufacturing the same products," Ivannikov said. The differences may not prove dramatic, but partners can expect to find something unique wherever they turn, he added.

Kerem Koca headshotKerem Koca

A large manufacturing enterprise could have 40 factories, a variety of in-house IT environments and machine producers, three or four cloud providers and different BI and analytics tools, noted Kerem Koca, co-founder and co-CEO at, a digital transformation solutions provider based in Tampa, Fla. Given that level of diversity, creates a reference architecture for one factory and then replicates that across other factories. The reference architecture includes technology building blocks, standards and reusable code.

The approach lets individual factories purchase products from different vendors -- provided they adhere to the architecture, Koca said.

Insight's McKinnon also encourages the long-term architectural vision, with manufacturers incrementally deploying technologies within that framework. Scaling becomes a continuing conversation between partner and customer as they pursue individual wins over time.

"You don't need an end-to-end digital transformation to see really significant improvement," he said.

Dig Deeper on Emerging technologies for MSPs

Cloud Computing
Data Management
Business Analytics