Why Industrial AI Scaling Stalls: Insights from Factory Leaders

Overview

Manufacturers are moving rapidly from discussing AI to experimenting with it, but maturity remains uneven across organizations. Workshops with more than 50 manufacturing leaders show a consistent pattern: executives view AI as a lever for throughput, working-capital performance, and strategic capacity planning, while shop-floor leaders judge AI by reliability, ease of use, and whether it reduces day-to-day friction. The primary blockers are not algorithms, but data readiness, organizational ownership, and workforce adoption. Companies that align IT and operations, invest in data foundations, and pilot against clear operational problems are best positioned to scale AI into measurable outcomes. This ARC Insight outlines the most common scaling barriers and provides pragmatic steps operations leaders can take to move from pilots to sustained impact.

Manufacturers are moving rapidly from discussing AI to experimenting with it, but maturity remains uneven across organizations. This ARC Insight outlines the most common scaling barriers and provides pragmatic steps operations leaders can take to move from pilots to sustained impact.

Key Takeaways

Industrial AI programs scale when IT and operations share ownership and decision rights.

Measurable value comes fastest from targeted use cases tied to throughput, downtime, and plan adherence.

Data governance and contextualized OT/IT integration are prerequisites, not phase-two tasks.

Workforce trust and workflow fit determine adoption more than model sophistication.

Context and Market View

ARC research and field conversations indicate that broad interest in AI is outpacing operational readiness. In a recent ARC survey, only 12 percent of European manufacturers described themselves as advanced in digitalization and already applying AI/ML at scale to drive measurable business outcomes. The same research suggests AI program ownership often sits in IT (32 percent led by CIO/CTO) more frequently than in operations (22 percent led by COO/operations leadership), creating a common mismatch between where AI is defined and where value is realized: on the shop floor.

Three patterns emerged consistently across roles, responsibilities, and proximity to daily operations:

AI maturity is fragmented, with different layers of the organization holding materially different expectations, definitions of success, and risk perceptions.

Scaling depends more on alignment than on tools, and initiatives commonly falter when strategy is IT-led while execution, adoption, and accountability sit in operations.

The most compelling value cases cluster around throughput, downtime reduction, planning stability, and expertise retention, but they materialize only when data and change management are addressed early.

COO/Operations Leadership View

From a COO and operations leadership perspective, the primary objective is to improve throughput and OEE, reduce stoppages, and convert operational improvements into financial outcomes such as delivery speed, working capital, inventory, and capacity utilization. Leaders also see meaningful upside in efficiency gains (participants cited potential throughput improvements up to 20 percent in selected contexts), faster and more transparent decision making, and better scenario planning for growth and constraints. At the same time, concerns center on incomplete or unreliable data, limited internal capability to operate and sustain solutions, integration complexity, workforce resistance, and uncertain cost/ROI. As a result, many favor a pragmatic approach that starts with pilots tied to high-pain segments, improves data quality and analytics first, and then scales based on proven value rather than a “big bang” transformation.

ARC Advisory Group clients can view the complete report at the ARC Client Portal.

Obtain more ARC In-depth Research Market Analysis.

Engage with ARC Advisory Group

ARC Services ARC MarketMaps Contact Us