Manufacturing's labor shortage is no longer cyclical, but structural: data shows 74% of manufacturers report acute skilled-worker shortages and a projected global shortfall of 1.9 million manufacturing workers by 2033.
Companies are responding by investing in automation and workforce orchestration to drive productivity. But too often, these efforts are framed as a false choice: either equip workers with more digital tools and data, or rely on the frontline expertise that has kept operations running for decades. Manufacturers need both, and this overlooks the fact that the most valuable operational intelligence comes from frontline workers who routinely spot risks, inefficiencies, and failures before any sensor does.
By having smart wearables that capture audio and other human signals (the spoken communication, observations, decisions, and contextual insights shared during everyday work), knowledge that typically disappears at the end of a shift transforms into actionable operational intelligence. Wearables have the power to preserve judgment and expertise that becomes even more critical as facilities move toward increasingly automated operations.
Every retirement or departure can mean the loss of years of hard-earned judgment, from recognizing the early signs of equipment failure to spotting a safety issue before it turns into a major risk. That said, most manufacturers have no scalable way to capture that expertise before it has already walked out the door. The challenge is creating systems that capture and scale that knowledge rather than forcing workers to choose between doing the job and documenting it.
Outdated operational intelligence
For decades, industrial transformation has focused on extracting more intelligence from machines. Manufacturers have invested heavily in sensors, industrial IoT, manufacturing execution systems and analytics platforms to understand what equipment is doing and predict what it might do next. Those investments have created value, but they have also reinforced a narrow view of where operational intelligence lives.
Machine data has an advantage: it is deterministic, structured and relatively easy to capture. Human intelligence is different. What workers see, hear, say and infer is probabilistic, contextual and often fleeting. Many manufacturers assume that if something can't be captured by a sensor or entered into a dashboard, it isn't actionable data. For example a subtle change in a motor sound, or a recurring workaround becoming more frequent. Those observations may never appear in a dashboard, and without the right tools to capture them in the flow of work, they usually disappear at the end of a shift.
This is one of the central contradictions of modern manufacturing. Companies have more operational data than ever, yet some of their most valuable intelligence remains invisible. Humans on the frontline are, in many respects, the last frontier of enterprise intelligence.
This matters even more as experienced workers retire and manufacturers struggle to replace them. While the immediate concern with the manufacturing labor shortage is headcount, in the long-run, it’s a knowledge-retention problem.
When a worker with 20 years of experience leaves a facility, the organization can lose thousands of small judgments accumulated over decades: what to watch for, which anomalies matter, when to escalate and how to distinguish a harmless variation from the beginning of a serious failure or safety concern.
Traditional knowledge-management systems are poorly equipped to preserve that expertise because they depend on workers stopping what they are doing to document it. In fast-moving, high-noise, hands-on environments, that rarely matches the reality of work.
It requires the manager, after a long shift, to input data based on memories of what took place hours before into a tablet. It requires plant workers to leave their stations, remove protective garments or escape the noisy floor to press buttons on devices incompatible with their gloves or the sound levels in their environment. It requires them to enter data into form fields in ERP that may not accurately capture the nuance or specifics of the issues at hand. Or, it relies on broken translations to document critical steps taken in an emergency.
But what if all these steps could happen as the work itself progresses – in any language and be preserved for the future?
The connected worker promise fell short
Much of the operational intelligence generated on the factory floor never becomes part of the organization's decision-making and workflows. Small observations, emerging risks, informal workarounds, and contextual knowledge are often shared verbally, or not at all, and disappear once a shift ends.
Fulfilling the promise of the connected worker requires more than connecting frontline workers to digital tools. It requires connecting the knowledge generated during everyday work to the broader enterprise.
When communication and knowledge capture fit naturally into frontline workflows, manufacturers gain earlier visibility into operational issues, preserve valuable expertise, protect the safety of those working the next shift, and overall create a richer understanding of what is happening across their operations.
Automation makes human judgement more valuable
This becomes especially important as manufacturers move toward highly automated and, in some cases, near-lights-out operations. The common assumption is that more automation makes people less important. In practice, it changes what people are important for.
This starts with recognizing frontline knowledge as operational data. The observations workers make, the issues they escalate, the patterns they notice and the decisions they act on in real time are all part of how work gets done. When that intelligence is captured and processes are orchestrated around it, manufacturers lose visibility into the risks, workarounds and early warning or safety signs that shape daily operations.
As routine physical tasks are automated, the remaining human contribution shifts from dexterity toward judgment. Workers are now responsible for exceptions, anomalies, escalation and intervention — precisely the situations where context matters most and the stakes rise. A small problem can become an expensive disruption or dangerous situation before a centralized dashboard provides enough context to explain what happened.
The more automated the operation, the more consequential human absence can become. Manufacturers therefore need systems that do more than collect machine telemetry. They need to capture, connect and act on the intelligence generated by the people closest to the work.
For manufacturers facing a structural labor shortage, that is the larger opportunity presented by smart wearables: not merely improving communication or worker productivity, but preserving judgment at scale. Frontline knowledge is data, and technology must be designed around the realities of physical work and build systems capable of turning human signals into action.
The factory of the future may be more automated than ever, but human intelligence is far from obsolete. Rather, capturing it is more essential.
