AI Readiness for Manufacturers

The Lesson From ERP

History Is Repeating Itself

When ERP systems came to manufacturing in the 1990s and 2000s, the pitch was transformation. Integrated data. Real-time visibility. Decisions based on facts instead of gut feel.

What actually happened: companies that had clean data, documented processes, and disciplined users got most of what was promised. Companies that didn't got expensive systems that were partially implemented, barely used, and quietly blamed for problems that existed before the software arrived.

The tool wasn't the problem. The foundation was. And the companies that got the most out of ERP weren't the ones who bought the most capable system — they were the ones who were operationally ready to use it.

AI is following the same curve. The manufacturers who will get real ROI from AI in the next five years are building the foundation now — not after the implementation fails.

The foundation is the same as it was in the ERP era: documented processes, trusted data, defined standards, and a team that understands what good looks like. None of that changed. The tool changed. The readiness requirements didn't.

What the Work Looks Like

Operations-First. No Vendor Agenda.

I don't sell software. I'm not affiliated with any platform. The work is diagnostic — a direct look at your actual operations to assess where you are and what needs to change before AI adds value rather than cost.

Manufacturing operations have specific patterns that matter for AI readiness: production data quality, process documentation depth, quality standards definition, and tribal knowledge concentration. We look at all of it.

• Process documentation audit — where are the undocumented workflows, the single points of expertise, the informal standards that only live in someone's head?
• Data quality assessment — is your production, quality, and cost data clean enough to feed AI, or will it produce unreliable outputs?
• Systems integration review — what connects to what, where the gaps are, and what needs to change before you layer more tools on top
• Standards gap analysis — do your teams have written definitions of what good looks like, or is quality evaluated by experience alone?
• AI readiness roadmap — what to fix first, in what order, and what AI applications make sense for your specific operation

Every engagement starts with a 30-minute discovery call. No pitch, no deck — just a direct conversation about where your operation is and whether this work makes sense for your business right now.

The OT/IT Problem Nobody Talks About Before Selling You AI

Your PLCs, SCADA, and MES Platforms Were Not Designed to Feed AI — and Most Vendors Don't Tell You That

Manufacturing AI runs on production data. That data comes from operational technology — PLCs, SCADA systems, MES platforms, quality inspection systems, and legacy ERP integrations. The problem is that most of this infrastructure was built for operational reliability, not data accessibility. The data exists. Getting it into a form that AI can actually use — cleanly, consistently, and securely — is the problem most AI vendors wave past.

There's also a security dimension that almost nobody addresses before signing the vendor agreement. When AI tools connect to your shop floor data, they're connecting to operational technology that controls physical processes. The OT/IT convergence creates exposure that traditional IT security frameworks weren't built to handle. What data is flowing where? Who has access? What happens during an incident? What are the vendor's actual security obligations in the contract?

An MS in Information Assurance means this isn't a checkbox — it's a material part of the assessment. The information assurance module included in every Systems Readiness Assessment reviews OT data governance, access controls, and AI vendor security posture specifically for manufacturing environments. For manufacturers, this is where the most unexpected — and most consequential — gaps tend to be.