Manufacturers Lag in Data Readiness as AI Tools Accelerate, Panel Says

Why It Matters

Data readiness is the linchpin for manufacturers seeking to harness AI-driven process optimization, predictive maintenance, and quality control. Without clean, connected data, AI projects become cost‑prohibitive, delaying the transition from experimental pilots to revenue‑impacting production tools. The industry‑wide push for shared standards and federated learning could unlock billions in efficiency gains, reduce waste, and accelerate innovation across supply chains. Moreover, the manufacturing sector accounts for a significant share of global GDP; its ability to adopt AI at scale will influence broader economic productivity. By addressing data silos now, manufacturers can avoid a future where only a handful of data‑rich incumbents reap AI benefits, preserving competitive balance and fostering a healthier ecosystem of suppliers and innovators.

Key Takeaways

• Data scientists spend 80‑90% of their time cleaning data, not building models.
• Manufacturers rely on siloed Excel spreadsheets, hindering AI deployment.
• University of Maryland launches an Industrial AI Center to create shared datasets.
• Federated learning proposed to protect IP while enabling cross‑company model training.
• Rapid + TCT 2026 highlighted the need for industry‑wide data standards and infrastructure.

Pulse Analysis

The panel’s diagnosis of a "data organization bottleneck" mirrors challenges seen in other legacy industries transitioning to AI. Historically, sectors that invested early in data warehouses and common schemas—such as finance and retail—were able to scale AI faster and at lower cost. Manufacturing’s reliance on ad‑hoc spreadsheets reflects a legacy of point‑solution data capture, which now clashes with the data‑hungry nature of modern machine‑learning pipelines.

The push for large knowledge models (LKMs) and federated learning signals a strategic shift from proprietary, closed‑loop AI to a collaborative, ecosystem‑wide approach. If ASTM and the Industrial AI Center can codify data formats and enable secure model sharing, manufacturers could achieve network effects similar to those that propelled large language models. This would democratize advanced analytics, allowing mid‑size firms to benefit from models trained on aggregated industry data without exposing trade secrets.

Looking ahead, the success of these initiatives will hinge on three factors: (1) the willingness of leading OEMs to champion open standards, (2) the development of robust governance frameworks that balance IP protection with data utility, and (3) the scalability of federated‑learning infrastructure across heterogeneous plant environments. Companies that move quickly to modernize data pipelines and join collaborative consortia will likely capture early AI‑driven productivity gains, while laggards risk obsolescence as AI‑enabled competitors slash costs and improve quality.

In sum, the data readiness gap is not merely a technical hurdle; it is a strategic inflection point that will determine which manufacturers thrive in the AI era.