RTA's Blog
Beyond the Algorithm: Why Standardized Data Models Matter More Than AI
Jun
What John Rinaldi shared during his Automate 2026 presentation on building scalable Industrial AI.
Artificial intelligence is dominating manufacturing conversations, but many organizations are focused on the wrong problem. Before AI can deliver meaningful insights, manufacturers need data that is accurate, contextualized and standardized.
During his presentation at Automate 2026, RTA CEO John Rinaldi explored why standardized manufacturing data models are the real foundation of scalable Industrial AI and shared practical strategies manufacturers can implement today.
AI Doesn’t Have a Data Collection Problem
Today’s manufacturers can collect more data than ever before.
The challenge is turning that data into information AI can trust. Raw PLC tags without engineering units, inconsistent naming conventions, disconnected systems and missing business context create data that is difficult to analyze and even harder to scale. As John explains, the problem isn’t the amount of data available, it’s the fragmentation surrounding it.
Three Bottlenecks That Prevent Industrial AI
According to the presentation, most AI initiatives struggle because of three common challenges:
- Raw, context-free PLC data
- Difficulty connecting data across disparate systems
- Poor or inconsistent data modeling
Until these bottlenecks are addressed, adding AI simply increases complexity instead of improving decision making.
Why Historians Have Become More Than Data Archives
Traditional historians were designed to collect and store data.
Modern historians need to normalize data, apply context, publish information to multiple destinations, support cybersecurity requirements and provide the foundation for analytics and AI applications. A historian should help explain what happened, not simply record that it happened.
Edge vs. Cloud Isn’t an Either/Or Decision
One of the key recommendations from the presentation is adopting a hybrid architecture.
High-resolution operational data belongs at the edge where it supports diagnostics and machine learning, while aggregated information can move to enterprise systems and cloud platforms for reporting and business analytics. This approach improves performance while reducing cloud costs and maintaining security.
The Biggest Takeaway
AI is only as good as the data it receives.
Organizations that standardize data at the source, govern it consistently across the enterprise and build scalable data models will be in a much stronger position to adopt AI successfully than those that simply collect more data.
Want the complete presentation?
Download the Automate 2026 presentation slides
