Answer extracted from the The Manufacturing IT Podcast — listen to the full episode below.
What advanced analysis engine is Sepasoft adding to simplify AI implementation in manufacturing?
Sepasoft is launching a production-ready analysis engine combining industry connectors like MQTT, APIs, and Kafka with an advanced analytics layer that automatically determines root downtime causes, calculates statistical process control metrics, generates machine learning predictions, and analyzes operator sentiment from notes. The system integrates with large language models and pushes actionable insights back to the plant floor, such as expected production losses, eliminating the friction that typically slows AI adoption in manufacturing environments.
From connectors to actionable intelligence
Most AI implementations in manufacturing fail because data integration and analysis remain separate, disconnected workflows. Sepasoft's approach as discussed in the episode solves this by bundling connectivity and intelligence into one product.
The engine handles the heavy lifting that manufacturers typically outsource or abandon: root cause determination automatically identifies why downtime occurred by correlating production logs with equipment states and operator actions. Statistical process control calculations track quality drift in real time. Machine learning models trained on historical data predict failures and anomalies before they impact production.
Beyond pure statistics, the system analyzes free-text operator notes using natural language processing to extract patterns humans might miss—sentiment shifts, recurring complaints, maintenance hints buried in casual remarks. This data feeds back into machine learning pipelines to refine predictions.
LLMs and the feedback loop
The integration with large language models marks a shift from isolated analytics to conversational AI on the plant floor. Analyzed data flows back as plain-language recommendations: "Expected production loss next shift: 2 hours" or "Machine 3 risk of bearing failure increased 15% this hour."
This transforms raw analysis into decisions operators and supervisors can act on immediately. A manufacturing leader detailed in this conversation noted that simplicity in implementation is what customers demand—and this feedback loop design removes the need for data scientists on the floor, making AI practical for plants with limited technical depth.
"Anything that simplifies the interactions between a company and our company to handle purchasing is what people are looking for."
Tom Heckman — Founder and CTO at Sepasoft, with over 20 years in manufacturing and automation. Heckman started in the 1980s at an Allen Bradley distributorship focusing on data collection and PLCs, later owned an integration business, and founded Sepasoft in 2010 to bridge the gap between MES and modern IT infrastructure. He recently transitioned to CTO to focus on the technology vision behind products like this analysis engine.
The product is scheduled for release in the first quarter of 2025, positioning Sepasoft to address a concrete pain point: manufacturers have the data but lack the accessible tools to extract and act on it. The connectors are proven—MQTT, APIs, Kafka have become standards—so the differentiation lies in the analysis layer and its ease of deployment.
For operators and plant managers, the value is immediate. One detail worth exploring in the full episode is how Sepasoft's tight integration with Inductive Automation's Ignition platform reduces friction further, allowing analysis modules to live alongside HMI and SCADA on a single system rather than requiring yet another standalone tool.
See also
What challenges do manufacturers face when preparing MES data for artificial intelligence applications?
Many manufacturers struggle with getting clean MES data into AI systems. MES data isn't static after initial recording—quality results change, tests run, and datasets require continuous validation before feeding into machine learning models.
How did Stone Brewing improve operational efficiency after implementing MES software for brew house control?
Stone Brewing, a Southern California brewer recently acquired by Sapuro, rolled out Sepasoft software to control their entire brewing process and beer production, enabling operators to modify recipes with minimal training and improve consistency across batches.
What are the key benefits of integrating MES directly with HMI and SCADA platforms on a single architecture?
By running modules on Ignition, both HMI SCADA and MES exist on the same platform, enabling simple controls like interlocks without requiring separate systems, reducing complexity and improving data flow between operator interfaces and manufacturing execution.