Welcome to Episode 3 of our special podcast series on Industrial DataOps. Today, we’re excited to sit down with Andrew Waycott, President and Co-founder of TwinThread, to explore how AI and Digital Twins can transform manufacturing operations.
Andrew has been working with industrial data for over 30 years, from building MES and historian solutions to developing real-time AI-driven optimization at TwinThread. In this episode, we discuss the state of industrial data, the role of AI, and why closed-loop automation is the future of AI in manufacturing.
What is TwinThread?
TwinThread was founded with a simple but powerful mission: Make AI accessible to non-technical engineers in manufacturing.
As Andrew explains:
"Most engineers in manufacturing shouldn’t have to become data scientists to solve industrial problems. TwinThread is about giving them AI-powered tools they can actually use."
The platform covers data ingestion, contextualization, AI analytics, and closed-loop optimization, all while allowing manufacturers to start small, scale fast, and operationalize AI without massive IT overhead.
Mapping TwinThread to the Industrial Data Platform Capability Model
For those following our podcast series, you know we’ve been refining our Industrial Data Platform Capability Map—a framework to understand how different vendors fit into the industrial data ecosystem. Andrew breaks it down step by step:
Connectivity: TwinThread ingests data from a wide range of industrial systems—Historians, OPC, MES, databases, IoT platforms, and MQTT.
Digital Twin & Contextualization: The platform structures data into Digital Twins, modeling not just assets, but also maintenance, production, and process relationships.
Data Cleaning & Quality: TwinThread automates the process of cleaning, organizing, and adding context to industrial data.
Data Storage: While TwinThread functions as a cloud historian, it doesn’t require companies to replace existing on-prem historians.
Analytics: The core strength of TwinThread is its ability to analyze and optimize processes using AI, applying predictive models to industrial operations.
Data Sharing: The platform generates curated datasets—ready for BI tools like PowerBI, Snowflake, or Databricks—allowing manufacturers to turn raw data into actionable insights.
Visualization & Dashboards: Unlike traditional generic dashboards, TwinThread provides visual tools optimized for operational decision-making.
As Andrew puts it:
"We don’t just show data. We help you solve problems—whether that’s quality optimization, energy efficiency, or predictive maintenance."
A Real-World Use Case: Quality Optimization at Hills Pet Food
One of TwinThread’s most successful deployments is with Hill’s Pet Food (a Colgate company), where they’ve transformed quality control across all global production lines.
The Challenge:
Dog and cat food requires strict control of moisture, fat, and protein levels to ensure product consistency and compliance.
Manual adjustments led to variability, waste, and inefficiencies.
Traditional sampling-based quality control meant problems were discovered too late—after bad batches were already produced.
The Solution:
TwinThread integrates with Hill’s existing infrastructure, pulling data from historians and process control systems.
Their Perfect Quality AI Module predicts final product quality in real time—before production is complete.
The system automatically optimizes setpoints at the beginning of the line, ensuring the process always stays within ideal quality parameters.
The Results:
No more bad batches—quality issues are detected and corrected before they occur.
Maximized yield & cost efficiency, as AI continuously fine-tunes production to hit quality targets at the lowest possible cost.
Scalability—The system is now running on 18 production lines worldwide.
And perhaps most impressively:
"We implemented a fully closed-loop, AI-powered quality control system—probably the first of its kind in the food industry."
Closed-Loop AI: The Key to Scalable Industrial Automation
Many companies struggle to move beyond pilot projects because AI-driven insights still require manual intervention. TwinThread changes that with closed-loop AI.
Instead of just providing insights, the system automatically adjusts process parameters to maintain optimal performance.
Andrew explains:
"A lot of people think closed-loop automation means making adjustments every millisecond. But in reality, most industrial processes don’t need real-time micro-adjustments—what they need is the ability to make controlled, intelligent changes at regular intervals."
At Hills Pet Food, AI-generated adjustments are sent directly to the control system, where operators can:
Manually review recommendations before applying them.
Auto-accept adjustments within pre-set limits.
Why Closed-Loop AI Matters:
Eliminates the risk of “shelfware”—AI models that aren’t actively used often get abandoned.
Ensures long-term impact—AI insights become part of daily operations, not just a one-time report.
Frees up operators—Instead of constantly tweaking processes, they focus on higher-value tasks.
The IT/OT Divide: What Makes AI Projects Succeed?
One of the biggest barriers to AI adoption in manufacturing is organizational silos between IT and OT.
Red flags in AI projects?
No IT/OT collaboration—When IT and OT teams don’t align, AI solutions often fail to scale beyond pilots.
No senior-level sponsorship—Without executive buy-in, projects get stuck in proof-of-concept mode.
Lack of automation maturity—Companies still manually tracking process variables on paper aren’t ready for advanced AI-driven optimization.
Andrew sees a major shift happening:
"Nine years ago, getting buy-in for AI in manufacturing was nearly impossible. Today, leadership teams actively want AI solutions—but they need a clear roadmap to operationalize them."
Standardization: The Next Big Challenge for Industrial AI
Despite advances in AI and cloud data storage, the industrial world still lacks standardized ways to store and structure data.
Andrew warns:
"Every company is reinventing the wheel—creating their own custom data lakes with unique structures. That makes it nearly impossible to build scalable, interoperable AI solutions."
Andrew suggests the industry needs a standardized approach to cloud-based industrial data storage—similar to how Sparkplug B standardized MQTT architectures.
Final Thoughts
We had a fantastic conversation with Andrew Waycott, who shared insights on AI, Digital Twins, and scaling industrial automation.
If you’re interested in learning more about TwinThread, check out their website: www.twinthread.com.
Or visit them at the Hannover Messe at the AWS Booth, Hall 15, Stand D76. More information can be found on the HMI website.
Stay Tuned for More!
Subscribe to our podcast and blog to stay updated on the latest trends in Industrial Data, AI, and IT/OT convergence.
🚀 See you in the next episode!
Youtube: https://www.youtube.com/@TheITOTInsider
Apple Podcasts:
Spotify Podcasts:
Disclaimer: The views and opinions expressed in this interview are those of the interviewee and do not necessarily reflect the official policy or position of The IT/OT Insider. This content is provided for informational purposes only and should not be seen as an endorsement by The IT/OT Insider of any products, services, or strategies discussed. We encourage our readers and listeners to consider the information presented and make their own informed decisions.
Share this post