How Industrial AI Strengthens Manufacturing Resilience

Forbes

11-06-2025

Jerry Dolinsky is the CEO of Dozuki , a leading connected worker solution for enterprise-level manufacturing companies. getty
When I met James, a veteran maintenance technician, he could tell from a faint rattle whether a gearbox would seize in two days or two weeks. He knew the precise torque sequence that kept a 30‑year‑old stamping press humming past its rated life.
When he retires next quarter, that intuition retires with him.
Multiply James by thousands of soon‑to‑retire front-line workers, and you begin to see the quiet crisis I've observed across global manufacturing: Critical know‑how is evaporating faster than we can capture it.
Knowledge attrition isn't an HR inconvenience anymore—it's an enterprise‑level risk to productivity, quality and safety. In BCG's 2024 AI adoption study, only 26% of surveyed companies had scaled value‑creating AI use cases. That gap reflects a failure to prioritize systemic problems like knowledge transfer over flashier pilots.
Meanwhile, a June 2023 McKinsey report estimated GenAI could unlock up to $4 trillion in annual productivity, with manufacturing among the top beneficiaries. However, if we allow expertise to walk out the door, that upside disappears. From Tribal Wisdom To Institutional Memory
In my work helping companies adopt AI on the factory floor, I've seen modern tools ingest everything from legacy SOPs to video walk-throughs and sensor data, transforming it into dynamic, role‑based guidance—instantly searchable on any device. I call this shift "process digitization at the speed of operation."
If you're considering where to begin, here's how we've seen companies move quickly without overhauling their infrastructure: 1. Start with reality, not theory.
Most plants already sit on a mountain of unstructured content—dusty binders, disconnected spreadsheets and endless ad-hoc videos. AI can now classify, tag and standardize that material in days, not quarters. The first win is unlocking the value of what you already have.
We've helped teams skip the burden of marathon documentation sessions by capturing short videos on the line—think TikTok for torque specs. These clips feed AI models that autogenerate step‑by‑step instructions, multilingual subtitles and safety callouts. This results in work instructions that are ready before the next shift clocks in. 3. Embed feedback loops at the point of use.
Operators know when an instruction doesn't match reality. With the right tools, they can flag discrepancies on the spot using voice notes or text. AI then reconciles those changes against global standards. This loop transforms compliance from a checklist into a real-time collaboration engine—and I've seen this reduce rework significantly. 4. Measure knowledge like any other asset.
Some of the most forward-looking companies I work with now track knowledge using formal KPIs.
• Percentage of critical tasks with validated digital standards
• Time-to-competency for new hires
• Number of front-line submitted improvements per quarter
Treating knowledge like OEE or EBITDA reframes it as a performance driver, not an afterthought. 5. Build a culture where expertise scales.
AI isn't a replacement—it's an amplifier. When companies elevate the voices of their most experienced workers and make their insights accessible to every shift, they build prestige around contribution. That attracts and retains the next generation of talent who expect consumer-grade digital tools on day one. Why Now: A Narrow, Urgent Window
A convergence is happening in manufacturing right now that makes this the ideal moment to act and the worst time to wait. Demographic pressures are intensifying. A 2019 Manufacturing Institute survey found that many manufacturers had significant shares of their workforce that were over the age of 55—meaning they would reach retirement age within the decade. I've seen firsthand how this exodus is straining training pipelines, thinning teams and exposing production lines to operational risk.
At the same time, the technology has matured. Cloud infrastructure is no longer a barrier, and the latest generation of large language models and computer vision tools has crossed the usability threshold. A small operational excellence team can now pilot in weeks what once required a systems integration army to deploy. Implementation is no longer about technical feasibility—it's about organizational will.
Finally, there's competitive pressure. In a 2024 Lucidworks survey (via Reuters), 58% of surveyed manufacturers reported increasing their AI budgets, even amid concerns around accuracy and ROI. Leadership teams recognize that waiting is a strategic risk.
To turn urgency into action, manufacturing leaders need a playbook. A Manufacturing-Leadership Playbook
To operationalize these ideas, I've outlined five strategic moves with targeted questions and outcomes: Elevate knowledge retention to an enterprise priority.
• CEO/COO: How much margin or uptime is at risk when veteran experts depart?
• Directors And Plant Managers: Which lines are most exposed to undocumented know‑how?
• Outcome: Converts an invisible liability into quantified exposure that merits investment. Launch a rapid-return lighthouse project.
• Operations And Quality Leaders: Which high-variance process causes the most scrap, downtime or safety incidents?
• Outcome: Demonstrates tangible ROI that can be socialized across sites and functions. Establish cross-functional governance.
• CTO And Ops Leadership: Who owns the digital standard once AI drafts it—ops, quality or HR?
• Outcome: Creates clear accountability and accelerates scale beyond a single pilot. Integrate front-line feedback loops.
• Plant Managers: How will operators flag improvements in real time, and who reviews them?
• Outcome: Turns compliance into collaboration, driving continuous process optimization. Align incentives with contribution—not tenure.
• All Leaders: How do we recognize the engineer whose AI-generated guide saves 1,000 hours?
• Outcome: Converts participation into prestige, boosting engagement and knowledge flow. Looking Ahead
In five years, the competitive gap won't be between companies that use AI and those that don't. It will be between companies that scale institutional knowledge and those that must relearn the same lesson every time a badge changes hands.
Industrial AI is the bridge between generations of expertise and the autonomous factories of the future. It's a resilience strategy. When James hands over his badge for the last time, the plant that captured his know-how will keep running like he never left—because in digital form, he never really did.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?