ABOUT THIS EPISODE
What happens when you combine shift-level insights, OT data, and predictive analytics? You stop burning buns, and start scaling profits.
In this episode of the Zeroed In Podcast, Marvin Tansley (Thing-Zero) sits down with Aaron Pfeifer (Atlas OT) to unpack how manufacturers can:
• Pull signal from noisy operations data
• Spot patterns in downtime, waste, and inefficiency
• Use successful shifts as live playbooks for struggling ones
• Bridge the gap between OT realities and IT-driven AI models
Whether you’re leading a plant, an OT security program, or a digital transformation initiative, this conversation gives you a frontline look at how to turn data into behaviors that improve reliability, safety, and profitability.
⏱ Chapters:
00:00 Start / Framing ROI: Maintenance math, partial valve testing (SIL mindset)
02:45 Real-time + historical data: Compare assets, vendors, and cycles for gains
05:30 Prep the data estate for AI: “You already own the apartment” (cloud)
08:50 Secure democratization: Read-only, one-way flows; exec dashboards that matter
12:10 Crawl → Walk → Run: Models, thresholds, and cost-of-action decisions
16:20 Edge models safely: Optimize offline; push anomalies up, not commands down
19:30 Brownfield realities: Knowledge loss, out-of-date prints, safety interlocks
24:30 The cost of NOT modernizing: Unplanned downtime vs. proactive budgets
28:50 Heterogeneous control stacks: Standardizing without standardizing
33:10 Agile POCs on your cloud: Measure waste first; fast wins prove value
37:10 Close: Better business → better world (and better jobs)
🎧 Brought to you by Thing-Zero — Secure. Predict. Augment. Automate.
📌 Follow Marvin Tansley on LinkedIn - https://www.linkedin.com/in/marvin-tansley-a914551/
📌 Follow Aaron Pfiefer on LinkedIn - https://www.linkedin.com/in/adpfeifer/
Follow Thing-Zero on:
- YouTube: YouTube.com/@Thing-Zero
- Linkedin: linkedin.com/company/thing-zero
- 𝕏: https://x.com/ThingZ96954
- Web: https://www.Thing-Zero.com
#IoT #OT #AI #PredictiveAnalytics #manufacturing
