Imagine spending millions building your dream house, complete with smart lighting, solar power and every modern convenience, only to discover it’s sitting on cracked foundations. Sooner or later, the structure will start to fail.
Industrial AI is heading down the same path. Too many organizations are trying to scale leading-edge intelligence on operating foundations that can’t bear the load.
NTT DATA’s 2026 Global AI Report: A Playbook for AI Leaders defines AI leaders as organizations with stronger AI strategies, greater AI maturity and higher AI-driven profits than their peers. They are more than twice as likely as laggards to rebuild core applications with embedded AI: 34.5% versus 14.5%.
That gap tells us where AI creates value. Leaders weave intelligence into the systems, workflows and physical operations that determine how the business performs.
For organizations where products are engineered, made, inspected, moved and sold through complex operating networks, the question is whether they’ve built the foundations to turn AI-generated insights into better decisions, safer actions and measurable results at scale.
Make the shift from experiments to operational intelligence
Organizations across industries have already shown that AI can improve individual tasks. Now comes the harder part: improving the connected work that stretches across production lines, planning teams, quality processes and supply chains.
In industrial environments, small gains can compound quickly. Better maintenance decisions reduce downtime, for example, while more accurate planning is good for inventory, production and customer delivery.
Because leaders recognize that these domains are interdependent, they redesign workflows from end to end, embed AI into operational workflows and give employees better information at the moment they need it to make decisions. That’s how AI moves from a promising pilot to a repeatable capability.
As Alonso Fernandez, Global Co-Lead: Product Industries at NTT DATA, says: “You can’t achieve operational excellence by optimizing one function at a time anymore. The biggest gains come when you connect the entire value chain, embed AI into everyday processes and modernize the core technology underneath it. That’s what allows decisions, data and work to flow together instead of getting stuck in silos.”
Physical AI adds intelligence where work happens
Physical AI makes this transition especially visible because it operates where industrial work happens: alongside machines, sensors, cameras, robotics, workers, edge infrastructure and digital twins.
A pilot can show that AI recognizes what’s happening, but production proves whether it can turn that understanding into trusted guidance that helps employees act faster and more confidently.
Quality inspection is one of the clearest examples. It depends on skilled inspectors spotting subtle variations in products and turning those observations into decisions. In paint defect detection, for instance, these manual inspections are time-consuming, resource-intensive and vulnerable to inconsistency.
NTT DATA developed a no-code AI solution on Google Cloud Platform to help Asian Paints detect paint defects by position, color, texture and defect type, showing how AI brings greater precision and consistency to a traditionally manual process. That kind of modernization allows intelligence to move beyond observation and become part of day-to-day operational improvement.
ALSO READ: Physical AI: Moving from cloud-centric AI to intelligence where work happens
Supply chains: Turn fragmented responses into adaptive execution
The same principle applies beyond the factory walls. Supply chains struggle when information is fragmented, decisions happen in silos and planning models can’t keep pace with volatility. AI can improve individual decisions, but the larger opportunity is to connect intelligence across the flow of work so organizations can anticipate change earlier and respond with greater confidence.
The case of a global beverage manufacturer illustrates what’s possible. We deployed GenAI tools, including agentic AI technology, to analyze vast datasets, recommend optimal retail trade spending and streamline decision-making for retail and marketing teams. The initiative uncovered about $10 million in new annual revenue opportunities and contributed to a $30 million revenue lift.
Those results come from making intelligence part of everyday commercial and operational decisions, not from creating another standalone AI application.
The goal isn’t just to make individual decisions faster but also to help the supply chain balance efficiency, cost, responsiveness and risk amid changing conditions. Once AI connects previously disparate datasets, organizations can adapt faster and more confidently while keeping the business running.
“Supply chains don’t need more isolated AI tools,” says Fernandez. “They need AI that’s connected across planning, operations and execution. Once you modernize the underlying applications, the entire business becomes more responsive and much better equipped to deal with constant change.”
Intelligent manufacturing becomes an operating capability
For manufacturing leaders, the opportunity is much bigger than automating individual tasks. It’s about creating production environments that get smarter over time.
Agentic AI, digital twins and predictive systems help factories make better decisions and deliver continuous improvements on the factory floor and in the broader business — but only when they’re supported by reliable data, secure connections between IT and OT, governed workflows and platforms that can be reused across sites.
Our AI report reinforces this point. It found that 85.6% of AI leaders use AI to support back-office and mid-office interactions. In other words, leaders are embedding AI deeper into the operational workflows that keep the business running.
When AI becomes part of production, planning, quality and maintenance workflows, organizations can respond more consistently, identify problems sooner and scale improvements across plants instead of solving the same challenge repeatedly.
Automotive and mobility: Proof at ecosystem scale
The automotive industry shows how far core modernization now extends. The opportunity is no longer limited to the plant. It now extends into the connected engineering, software and mobility environments that determine how vehicles are developed and updated.
For Continental Automotive, we used retrieval-augmented generation and agentic workflows to develop an AI companion for automotive product development, helping engineering teams manage requirements more efficiently.
The principle is the same. The value of AI grows when fragmented systems become connected enough for data, decisions and action to flow across the enterprise.
Take note of the leadership lesson
Building a solid foundation for AI is now a priority in your organization. Modernize your core applications, connect operational workflows, secure the infrastructure beneath them and establish the governance needed to scale AI responsibly.
Just as importantly, design these systems around experienced people, whose judgment remains central to every critical decision.
For manufacturers, automotive companies and consumer goods organizations, the next chapter of AI will be defined by how deeply intelligence is built into the operating core where products are made, moved, inspected, improved and delivered.
WHAT TO DO NEXT
Read more about NTT DATA’s approach to modernizing the core to see how we can help you build an AI-ready foundation for an intelligent, connected and resilient enterprise.