AI in Quality Management: Hype vs. Reality
By Eric Stoop
The reality is most manufacturers are still at the starting line with AI.
Artificial intelligence (AI) is dominating tech headlines across industries from retail to healthcare to financial services. In manufacturing, the excitement is just as strong.
Research from consulting firm McKinsey shows that manufacturing lighthouses (“the standard-bearers of manufacturing and supply chains”) are betting big on AI, with nearly 60% of top use cases among digital innovators relying on AI.
The results — like reducing defects, increasing productivity, and streamlining operations at a scale that was previously out of reach — are already here.
Or are they?
AI is Driving the Next Wave of Industrial Transformation — For Some
The McKinsey lighthouse research shows AI leaders are seeing results like 300% increased productivity and 99% reduced defects. These aren’t futuristic projections; they’re happening today for manufacturers that have invested in connecting their people, processes, and data.
Ford, for example, is using AI to reduce engineering cycle times by accelerating tasks like 3D modeling and stress predictions, while GM is using AI to streamline plant floor workflows.
GE Aerospace is leveraging AI tools to assist employees in finding the information they need, including solutions to quality issues. Schaeffler Group’s Hamburg plant has deployed an AI assistant to help track ball bearing defects and uncover potential root causes based on production data.
What’s Holding Others Back? The Foundation Isn’t There Yet
Despite these success stories, the reality is most manufacturers are still at the starting line with AI. McKinsey’s 2025 State of AI report, for instance, notes that only 5% of manufacturing functions had adopted AI as of 2024. (Unlike the lighthouse report, this study surveyed a more general manufacturing population across a range of regions, industries, company sizes, functional specialties, and tenures.)
Perhaps equally startling is the fact that most companies have yet to see value from these investments. According to a recent survey by Boston Consulting Group, another Big Three consulting firm, nearly three in four companies today are struggling to achieve and scale value from AI initiatives.
So which AI technologies are most prevalent in manufacturing quality today?
AI vision detection systems are the most developed use case, with manufacturers primarily using them for surface inspection and defect detection.
Other manufacturers are beginning to use AI in quality management to analyze large datasets, hoping to leverage insights from the thousands of variables that can exist within a given product.
That said, many are still struggling to make that data actionable.
For example, if you have 7,000 data points on a product, how can you relate data point #453 to #2671? Is it truly a causal relationship, or are they simply correlated? Connecting the dots between variables and assessing root causes still represents a big gap for manufacturers, requiring significant process knowledge and context to fill.
The key insight here: AI can’t work its magic without the right inputs and context. That’s why foundational readiness is so critical.
Why AI Can’t Fix Broken Processes
Here’s the uncomfortable truth: If your processes are inconsistent or undocumented, AI won’t save you. If anything, it might just automate bad decisions faster.
Imagine, for instance, a vision detection system that catches surface weld defects but can’t detect subsurface fusion issues caused by inconsistent operator technique. Without process verification, AI alone will fail to adequately detect quality risks, underscoring the need to verify critical process steps to prevent variation.
That means before relying on AI, manufacturers need to:
Standardize and document key processes
Verify that those processes are followed through routine checks
Make institutional knowledge accessible and sharable
Digital Foundation First, AI Second
AI works best when built on a strong digital foundation, but many manufacturers still face a major gap: the digital divide between systems, information, and people. Operators often rely on tribal or tacit knowledge, siloed paper-based processes, or verbal instructions; meanwhile, leadership lacks visibility into what’s actually happening on the plant floor.
To bridge that divide so AI can be used to its fullest potential, manufacturers need foundational tools that connect people to the right information at the right time, including:
Connected worker tools that deliver real-time guidance and capture tribal or tacit knowledge
Ongoing process checks such as layered process audits (LPAs) to verify critical-to-quality steps
On-the-job training workflows that ensure operators understand and apply new standards
Digitized data collection that makes insights accessible across shifts, teams, and locations
Without this groundwork, AI efforts risk becoming expensive science projects — impressive in theory, underwhelming in execution.
AI’s Promise is Real, But Only for the Ready
There’s no question that the rise of AI marks the most significant inflection point in the Industry 4.0 revolution to date.
But the road to real impact starts with process discipline, cultural alignment, and connected operations. AI in quality management doesn’t replace good processes — it supercharges them. The manufacturers best positioned for success are those that have established a solid foundation for AI to build upon.
