Does Manufacturing AI Have a Context Problem?
Manufacturers are not short on data. They are not short on systems, either.
A single product can generate information across engineering systems, ERP, manufacturing execution systems, quality platforms, supplier portals, maintenance systems, documents and spreadsheets. Each serves an important purpose. The problem appears when a decision requires information from several of them at once.
Consider something as routine as an engineering change.
The change may originate in a PLM system, but its impact does not stop there. Which parts are affected? Does the manufacturing bill of materials reflect the new design? Are there existing parts in inventory? What does the change mean for cost? Which suppliers are involved? Have work instructions and inspection requirements been updated? Could products already in production or in the field be affected?
Answering those questions can require teams across engineering, manufacturing, procurement, quality and supply chain to piece together information from different systems, using different identifiers and different definitions.
That fragmentation has always created friction. As manufacturers put more AI into operational and engineering workflows, it becomes a much bigger problem.
AI Can Only Work with the Context It Can Reach
The conversation around manufacturing AI often begins with the model: What can we automate? What can AI predict? How can generative AI help employees work faster?
But the usefulness of the answers depend on what the AI understands about the business behind the question.
Deloitte's 2025 Manufacturing Industry Outlook found that nearly 70% of manufacturers identify data issues, including quality, contextualization and validation, as significant obstacles to AI implementation.
That makes sense when you look at the questions manufacturers actually need to answer.
A design-to-cost decision may depend on product specifications, alternative parts, supplier pricing, inventory and manufacturing requirements.
A quality investigation may require inspection results, process parameters, supplier lots, engineering changes, nonconformance records and corrective actions.
A supplier disruption may force a team to determine which parts, products, plants and customer commitments are affected.
In each case, having the individual pieces of data is not enough. Teams need to understand how those pieces relate to one another.
The Real Cost of Fragmentation Shows Up in the Decision
Disconnected data is rarely defined as a "data problem." But it generally results in an engineer spending hours comparing records across systems.
It requires procurement to order a new component without visibility into equivalent inventory elsewhere in the organization.
It requires the quality team to manually reconstruct which supplier lot, process change or specification revision contributed to a defect.
It leads to the production team discovering that an engineering change did not make its way cleanly into downstream manufacturing information.
Increasingly, it results in the creation of an AI initiative that works well in a controlled pilot but becomes difficult to trust or scale once it encounters the complexity of real manufacturing operations.
The ABM strategy behind connected manufacturing intelligence highlights exactly these issues: fragmented product and supplier information can contribute to rising costs, limited traceability, inefficient change management, supplier risk and slower decisions.
The common problem is not a lack of information. It is a lack of connected business context.
Traceability Is about More than Finding a Record
Manufacturers have spent decades building systems of record. The next challenge is making the relationships between those records usable.
If a supplier alerts you to a suspect component, can you quickly determine every product configuration that contains it?
If a field issue appears, can you trace it back through production, quality, supplier and engineering history?
If an engineer changes a component, can the organization see the downstream implications for manufacturing cost, inventory, suppliers and compliance?
If an AI system recommends an action, can the person receiving that recommendation understand which data, documents and business rules contributed to it?
Answering these complex questions requires a connected view of products, parts, suppliers, processes, changes, quality events and operational activity across the lifecycle.
The ABM program centers on this idea. Better connections between engineering, manufacturing, supplier and quality information can support faster root-cause analysis, stronger product and component traceability, faster change management and better production planning.
What a Better Manufacturing Data Foundation Looks Like
Solving this problem does not mean forcing every function onto one application or replacing systems that already run the business.
A stronger foundation connects the information those systems hold while preserving the context needed to understand it.
That means being able to reconcile different identifiers for the same part or product. It means connecting structured records with documents such as specifications, certificates, work instructions and supplier documentation. It means understanding relationships between engineering designs, manufacturing processes, suppliers, quality events and operational assets.
It also means keeping lineage, permissions and evidence attached to the information, so that people, analytics and AI can use it with greater confidence.
The goal is not another copy of the data.
The goal is connected manufacturing intelligence: a business-ready view of how products, parts, suppliers, processes and operations relate to each other.
Start with the Decision that's Hardest to Make Today
Manufacturers do not need to begin by connecting everything. A more practical starting point is to find an area in which fragmented information is already creating measurable cost, delay or risk. The decision may involve:
Determining the impact of an engineering change
Understanding the true cost of a part across suppliers and production sites
Tracing a suspect component through affected products
Investigating a recurring quality problem
Start with Identifying the systems, records, documents and relationships needed to answer the question. Continue to find what evidence the business needs before it can act, then measure how long that process takes today.
That approach changes the AI conversation, as well. Instead of asking, "Where can we add AI?" manufacturers can ask a more useful question: "Do we have the connected, trusted context AI would need to help us make this decision better?"
Manufacturing's next advantage may not come from having more data or even more AI. It may come from finally connecting the information the business already has.
Want to See How This Approach Applies to Your Environment?
Discover how connected manufacturing intelligence can help you improve traceability, accelerate root-cause analysis and make AI-driven decisions with confidence.
