It’s 2030. A major customer has drastically shifted its demand forecast. Simultaneously, a critical supplier in Southeast Asia reports a disruption affecting your supply chain. Your production lines are already running at 94% capacity. 

Before your operations team finishes reviewing the alerts, three things have already happened: 

Your AI system has analyzed demand signals, supplier data, inventory positions, and production constraints simultaneously. It has weighted dozens of variables in real time and identified three possible responses each with different costs, delivery impact, and margin implications. It presents them with the reasoning transparent and the tradeoffs explicit. 

Your operations team has eighteen minutes to decide. 

What happens next determines whether you satisfy the customer, preserve supplier relationships, or absorb the cost. What happens next is your competitive advantage. 

But here’s the uncomfortable question almost no manufacturing leader wants to ask: How would your organization actually respond to that situation today? 

The Uncomfortable Truth  

The optimism around AI in manufacturing is real, with investment and adoption continuing to grow. 

But here’s what the data actually shows: only 21% of manufacturers report being fully prepared for AI adoption. And 79% of companies that have adopted AI report struggling to reap its full benefits. 

That gap between investment and value isn’t random. It’s not about vendor quality or algorithmic sophistication. It’s about something more fundamental: 

Most manufacturers can’t keep pace with AI because their technology foundations can’t support it. 

The Real Crisis 

Manufacturing operates in a different environment than five years ago. And the pace of change is accelerating. 

Global supply chain disruptions increased 38% year-over-year in 2024. Those disruptions cost companies an average of 8% of their annual revenues. 86.2% of manufacturers have spent the last two years working to de-risk their supply chains yet the disruptions are growing more frequent, not less. 

Meanwhile, your competitors are moving faster. Some have already connected supply chain visibility, production planning, and customer demand signals into a unified data environment. When disruption hits them, they model response scenarios in minutes. Before your organization finishes assessing what happened, they’ve adjusted production, renegotiated supplier commitments, and communicated revised delivery dates to customers. 

The old response pattern – stabilize, analyze, decide, execute – takes too long now. 

What separates manufacturers that adapt quickly from those that struggle isn’t the size of their crisis team. It’s whether their technology foundation lets them see problems and evaluate options in real time. 

The Hidden Cost of “Just Keeping the Lights On” 

I’ll be direct about something I see repeatedly: the decision to maintain legacy technology systems rather than modernize is often framed as pragmatic. “It’s working. We can’t afford the disruption. Let’s integrate new tools instead.” 

The financial case looks straightforward. But the real costs hide in operational friction: 

Legacy system maintenance costs manufacturers roughly $30 million annually per system. That excludes productivity lost to manual workarounds, decision delays from fragmented data, and the opportunity cost of capabilities your foundational system can’t support. 

More critically: as AI becomes embedded in operations, the constraint shifts. It’s no longer whether you can implement AI it’s whether your data architecture allows it to work effectively. 

Legacy systems were designed as data vaults, excellent for record-keeping and historical reporting. They were never designed for the real-time data sharing, event-driven decision-making, and AI-driven automation modern manufacturing requires. 

In twenty years working with manufacturers, I’ve watched two distinct responses. 

One group says: “Our legacy system handles core processes. Let’s add point solutions around it and hope integrations hold.” 

The other asks: “What technology foundation would let us respond to disruption in real time? What if every decision was AI-informed?” 

The second group isn’t asking about software versions, they’re asking a business strategy question that leads somewhere specific: not incremental improvements to legacy systems, but a unified data architecture that allows AI to see across silos and respond to real-time events. 

The 2030 Test 

That scenario I opened with, customer change, supplier disruption, eighteen minutes to decide, isn’t futuristic. It’s happening somewhere in your supply chain every week. 

The question is: Can you see it when it happens? Can you model responses fast enough to matter? Can you trust the data you’re looking at? 

Upgrading to IFS Cloud isn’t about getting “onto a newer platform.” It’s about creating the foundational architecture that allows real-time, data-driven operations to actually work. 

It means: 

  • Unified data visibility across supply chain, production, quality, and customer signals 
  • Real-time event processing instead of batch cycles and overnight reports 
  • AI-ready data quality where information moves reliably across the organization by design 
  • Operational flexibility to adapt rapidly when disruption happens 

This is a technical decision. But it’s not primarily a technology decision. It’s a decision about what kind of business you want to operate. 

The Question 

The manufacturers who will thrive in the next three to five years won’t be the ones who adopted AI earliest. They’ll be the ones who built foundational infrastructure to actually use AI effectively. 

The global AI in manufacturing market is set to grow from $34 billion in 2025 to $155 billion by 2030.That growth is real. The opportunity is real. Look at that 2030 scenario, your version. What needs to change for your organization to respond that way? Where is visibility lacking? Where are decisions still manual when they should be data-informed? 

Then ask the harder question: Can your current technology foundation actually support that future operating model? 

If the answer is “not really” if you’re maintaining legacy systems and integrating around them, if data silos are baked into your architecture, if decision speed is constrained by how long it takes to get reliable data then you’re not choosing between “maintain” and “upgrade.” You’re choosing between adapting now and scrambling to adapt later. 

The manufacturers making the strategic decision to modernize their foundation today won’t be the ones debating whether digital transformation is necessary in 2028. 

They’ll be the ones running it.