Reach or reality: Can converged manufacturing data save supply chains?

Is it a reach or reality to think that sharing the best thinking from manufacturing stakeholders in information technology (IT), operations technology (OT) and engineering technology (ET) can save today’s constrained supply chains? Depending on who you ask, supply chains are either on the brink of total collapse or poised for a complete transformation. Some experts see supply chains breaking under the weight of pandemic-related surges, declines and disruptions, while others envision an eventual stabilization and reemergence as supply chains become increasingly predictable, resilient, localized and data-driven.

While both sides remain adamant in defending their turf, perhaps they can agree that supply chains are not the only piece of the manufacturing puzzle that can benefit from meaningful data insights. In order for digital transformation to succeed, manufacturers need expertise from IT, OT and ET. Looming or lingering silos among these critical areas, however, have made it difficult to translate domain knowledge across expanses of corporate data centers, factory floors and engineering innovation labs.

According to a report summarizing a series of artificial intelligence (AI) workshops sponsored by the National Science Foundation (NSF) and the National Institute of Standards and Technology (NIST), “The manufacturing sector generates more measured, observational, operational, modeled and experience-based data than any other sector of the economy.” While the far-reaching benefits of predictive, computational modeling to control production quality are well-understood, data analytics and AI have been mostly limited to predictive maintenance and quality assurance.

Why? One reason identified by the workshop attendees is the prevailing manufacturing business model and “culture of secrecy” that “place[s] high value on expertise” and intellectual property. Another barrier is the lack of standard data formats and tools to ensure safe, secure data access and the ability to share non-proprietary data for AI training. According to a workshop summary from the NSF/NIST symposium, “AI has the potential to radically increase the value of manufacturing operational and product data by harvesting the implicit knowledge incorporated in it and harnessing its predictive, reactive and discovery capacity, again through datacentric modeling, machine learning, simulations and digital twins.” The most productive bounty of insights can be found at the intersection of IT, OT and ET.

ET is the missing puzzle piece
Digital transformation is at the center of converged IT, OT and ET strategies, according to ARC Advisory Group. For years, IT has had a place in manufacturing plants, with OT evolving alongside it to promote or prevent production-line changes. ET, however, has functioned separately, despite the fact that mechanical, electrical, design and industrial engineers possess rich and varied insights that can be translated and contextualized in powerful ways.

Successful digital transformations require collecting, assessing and taking advantage of integrated IT, OT and ET insights. Likewise, AI and machine learning (ML) need to be trained by a diverse set of teachers who understand the roots of predictability and resiliency, which also requires inputs from each group. Clearly, striking a balance between IT, OT and ET is essential, as this is an all-for-one or few-for-all scenario.

How do you find equilibrium among the groups? ET performs the deep dive into concepts, typically surfacing with discoveries that inform product design, technological innovation and customer differentiation. Engineers are visionaries who focus on the “what ifs” at every turn. Operational teams, in contrast, are focused on reducing risk, so they demand stable, predictable and precise processes.

Many engineers constantly question the status quo while calculating risk/reward to see how far they can push innovation before it breaks. Conversely, operations teams tend to be fastidious in developing and following specific rules because they know slight deviations can derail a production line. IT usually falls somewhere in between in both approach and philosophy. Among the trio, engineers are the most collaborative by nature, in part because of long-standing work with associations and standards bodies dedicated to moving entire industry segments forward.

The field of data science first gained momentum within IT as tech leaders saw the promise and peril of collecting, analyzing and extrapolating business-level insights from massive data stores. They realized pretty quickly that efforts to collect as much data as possible only resulted in data warehouses and repositories becoming chock-full of knowledge yet devoid of proper context. Amid the proliferation of aging data, any opportunity to unlock real-time value for the entire organization gets increasingly difficult if not unattainable. Unfortunately, this has done nothing but widen the chasm between IT, OT and ET as each group only uses what they can and ignores the rest.

Connected collaborations
Connecting the dots across data generated by information, operations and engineering technology teams is the only way to create an AI/ML foundation designed to improve manufacturing predictability and resiliency. First, this requires engineers to have a seat at the table, along with their counterparts in IT and OT. Next, the collective group must focus on looking for commonalities and reasons to collaborate.

The Information Technology & Innovation Foundation (ITIF), a think tank focused on formulating, evaluating and promoting policy solutions, believes alliances among semiconductor leaders are setting an example that the entire manufacturing industry can follow. ITIF envisions like-minded companies and nations collaborating on technology and ecosystem development, intellectual property and trade liberalization. The goal to do just that among semiconductor manufacturers is particularly appealing because this sector possesses some of the highest barriers to entry.

Yet this capital-intensive, highly specialized segment is bringing together scientists, engineers and researchers across the entire industry to collaborate, lead and innovate. Together, they are building sophisticated AI models that will drive competitive advantages while creating more transparent, resilient global supply chains for mutual benefits.

If we want to protect supply chains from continued disruption, we must remove organizational barriers, along with industry roadblocks as old as the manufacturing industry itself. Now is the time. Look for opportunities to share relevant industry data that elevates manufacturing while creating a high-impact “network effect” that reverberates around the world. Forbes