We sat down with IRIS Industrial Analytics Director and lead architect Michael Lynch for a wide‑ranging conversation on data, emerging technologies, and the future of manufacturing—uncovering practical insights manufacturers can apply today. With more than 30 years of experience in the field, from onsite service calls to engineering machines to run bigger, faster, and stronger, Michael has seen firsthand what truly drives performance on the plant floor. His career has taught him that data is the single most powerful tool in modern manufacturing—essential for clarity, efficiency, and meaningful innovation.
For Michael, data replaces assumptions with evidence, turning complexity into clarity and revealing the real opportunities to improve performance. It's the foundation that unlocks digital transformation, empowers teams to make informed decisions, and ultimately delivers longer, more efficient production runs with measurable ROI.
Data Turns “Art” Into Science.
Early in his career, Michael Lynch saw firsthand how many operators relied on instinct in the absence of reliable machine data—skilled “artists” making judgment calls based on experience, not evidence. But as he quickly learned, intuition has limits. What he calls the “Mythbusters” side of manufacturing requires separating bias and anecdotal knowledge from the real story behind machine behavior.
When data is clean, contextual, and collected at the right moments, it replaces guesswork with clarity. It answers critical questions with confidence: Where are we losing efficiency? What’s driving waste? Do we truly need more equipment or capital, or can we get more out of what we already have?
And often, the data tells a very different story than what teams expect. Many plants assume they’re running at 80% OEE—only to discover through empirical data that their true performance sits closer to 25–30%. Pencil whipped numbers don’t stand up to machine learning’s truth. You simply can’t optimize what you’ve never accurately measured.
That's why Michael encourages manufacturers to start small and scale smart. It's more than a memorable phrase—it's sound economic practice. Digital transformation can feel complex, but a focused pilot on a single line creates quick wins, proves ROI, and builds organizational confidence. When IRIS conducts an OEE pilot on one line, manufacturers often see a 25% bump in efficiency, a result strong enough to justify broader investment—such as the ~$80,000 needed for a segmented network to scale the solution across additional lines.
These early insights also help leaders avoid unnecessary capital spend. Once objective data is on the table, many discover they don't need new equipment or facility expansions—they simply need to run existing assets more efficiently, improve underperforming shifts, or reduce hidden waste.
Pilot projects aren't just a starting point; they're the bridge that protects budgets, aligns teams, and paves the way for transformation that grows sustainably and delivers measurable results.
Once a pilot establishes real baselines, the natural next step is understanding how every part of the plant contributes to—or constrains—performance.
Manufacturing Data Lives Inside an Ecosystem.
Solving modern manufacturing challenges requires more than isolated fixes—it calls for connected, bespoke solutions. Through decades on the plant floor, Michael Lynch learned that a single machine rarely tells the whole story. Adjacent equipment, utilities like compressed air and power, environmental conditions, and both upstream and downstream processes all influence performance. These aren’t separate pieces—they form an ecosystem. To diagnose issues accurately and protect uptime, manufacturers need to model and understand how those elements work together. As more variables are tracked, machine learning and generative AI take on the heavy lifting, churning through large datasets to surface defect patterns, anomalies, and correlations that human intuition would never catch.
For Michael, data is the key to tackling manufacturing's biggest and most costly problem: waste. Scrap, downtime, and energy loss affect every facility, and the strongest ROI comes from pinpointing the true sources of those inefficiencies. Clean, contextualized data helps leaders see beyond symptoms and get to root causes across machines, processes, and conditions.
But collecting data alone isn't enough. When alarms and events are tied to precise time slices, teams know exactly when and where to look, dramatically accelerating root cause discovery.
With the ecosystem mapped, the next step is strengthening the data foundation that makes AI truly effective.
Data Is the Foundation for AI.
Years of hands‑on troubleshooting taught Michael a foundational truth: not all data is created equal. For data to be genuinely useful, it must be validated, cleaned, normalized, and captured when machines are in the right operating state. Otherwise, teams risk making decisions based on noise instead of insight. High quality data isn’t optional—it’s the prerequisite for any meaningful digital transformation. Without it, AI can’t deliver reliable value. And in today’s environment, AI‑enabled operations consistently outperform intuition‑driven ones.
To future‑proof an operation, Michael encourages manufacturers to adopt open, connected architectures that make it easy to evolve with new technologies and emerging opportunities. IRIS is built with this philosophy at its core. Platforms integrate seamlessly with modern industrial protocols—including MQTT, OPC UA (via FactoryTalk Optix), and relational SQL databases—ensuring that data can flow freely across the enterprise. IRIS can publish into a Unified Namespace (UNS) and “put data into any open system manufacturers want,” while its agentic approach allows it to query source systems and APIs to pull ground truth information directly from manuals, SOPs, and information software.
In this era of unprecedented connectivity, network infrastructure becomes mission‑critical. As more devices and systems come online, so do the potential attack vectors—making strong networking and cybersecurity practices essential to both uptime and safety. IRIS’s open yet secure platforms help eliminate information silos while supporting the foundational controls manufacturers need as they scale.
By combining interoperability with disciplined network design, manufacturers gain the flexibility to adopt new tools without disruption—and the confidence that their data is being moved, accessed, and protected in a way that minimizes risk and maximizes longterm value.
While automation and AI are often framed as threats to labor, Michael sees their true potential: empowering frontline workers. By personifying and democratizing data through AI agents, manufacturers can place real‑time expertise directly in the hands of their teams.
Frontline operators—especially those who are newer, younger, or working off‑shift—gain immediate, contextual support through:
- Tailored, multilingual instructions
- 3D visual models
- Guided troubleshooting steps
- Realtime insight into what's happening and why
This gives operators the confidence and clarity to solve issues the moment they arise—reducing downtime, improving consistency, and strengthening the overall capability of the workforce.
With a reliable data foundation in place, manufacturers can finally translate AI insights into real improvements on the plant floor.
Better Data. Better Machines. Better Processes.
Data isn't just king—it's here to stay, and its value will only continue to grow. Throughout his career, Michael has learned that data is the single most powerful tool in modern manufacturing. It unlocks the full potential of AI, reduces waste, guides smarter investments, and empowers every level of the workforce.
Backed by The Kendall Group's deep manufacturing expertise, IRIS Industrial Analytics turns complex plant data into clear, actionable insights. If you're ready to harness that power to strengthen your operation, we'll help you take manageable, low risk steps that drive efficiency, spark innovation, and build long term success.
