Edge AI in the Digitalization of Industrial Process: Rainer Maidel - Business Development Manger, BE.Services

SA-95 is a standard that’s often misunderstood, but incredibly powerful. While many think ISA-95 is rigid or overly complex, it actually enables flexibility by: ⇨ 𝐃𝐞𝐟𝐢𝐧𝐢𝐧𝐠 𝐚 𝐬𝐡𝐚𝐫𝐞𝐝 𝐯𝐨𝐜𝐚𝐛𝐮𝐥𝐚𝐫𝐲 for manufacturing concepts, creating a true ontology for your data. ⇨ 𝐂𝐫𝐞𝐚𝐭𝐢𝐧𝐠 𝐬𝐜𝐚𝐥𝐚𝐛𝐥𝐞 𝐩𝐥𝐚𝐜𝐞𝐡𝐨𝐥𝐝𝐞𝐫𝐬 for every type of data, so you can start small and add new use cases later without rebuilding everything. ⇨ 𝐏𝐫𝐨𝐯𝐢𝐝𝐢𝐧𝐠 𝐭𝐡𝐞 "𝐰𝐡𝐲" 𝐛𝐞𝐡𝐢𝐧𝐝 𝐞𝐯𝐞𝐧𝐭𝐬, not just the "what," giving crucial context to your analytics and AI models. But how do you move from theory to a practical, modern implementation? In our latest AI in Manufacturing podcast episode, we explore exactly that with ISA-95 expert Jeroen Janssen, who is an MES/MOM Consultant at Rhize Manufacturing Data Hub. In the episode, you’ll learn: ✅ How to overcome a data culture that creates so many silos. ✅ The "use case stacking" method for a phased, value-driven implementation. ✅ What a native ISA-95 data hub looks like and how a graph database can bring it to life. ✅ Why this standardized approach is the key to unlo

Is the Timebase free historian getting an AI-Native DataOps component with Knowledge Graphs capability? You’ll hear it here first. In the latest episode of the AI in Manufacturing podcast, I sit down with Jeff Knepper, President at Flow Software Inc., to discuss the intersection of Information Management and AI in modern manufacturing, plus the exciting announcement of Timebase Atlas launch.   Here’s some of what we cover in this episode:   ✅ Why manufacturers struggle to make use of their data ✅ Building reliable pipelines for AI-driven use cases ✅ AI Agents in Manufacturing – Where they fit and what they need ✅ Unified Analytics Framework vs. Unified Namespace ✅ Historization Strategies – Best practices from edge to cloud ✅ Timebase Atlas Launch Announcement: Data Modeling, Pipelines, Knowledge Graphs, and AI interfaces ✅ MCP and Flow AI Gateway: Beyond APIs to Context-Aware Agent Interfaces

Most data-quality initiatives focus on things like freshness or schema. That works for IT data, but not for sensor data. Sensor data is different. It reflects physics. To trust it, you need contextual, physics-aware checks. That means spotting: → Impossible jumps → Flatlines (long quiet periods) → Oscillations → Broken causal patterns (e.g., valve opens → flow should increase) It’s no surprise that poor data quality is one of the biggest reasons manufacturers struggle to scale AI initiatives. This isn’t just data science, it’s operations science. Think of data quality as infrastructure: a trust layer between your OT data sources and your AI tools. Making that real requires four building blocks: 1. 𝐒𝐜𝐨𝐫𝐢𝐧𝐠 – Physics-aware anomaly rules, baselines 2. 𝐌𝐨𝐧𝐢𝐭𝐨𝐫𝐢𝐧𝐠 – Continuous validation at the right cadence (real-time or daily) 3. 𝐂𝐥𝐞𝐚𝐧𝐢𝐧𝐠 & 𝐕𝐚𝐥𝐢𝐝𝐚𝐭𝐢𝐨𝐧 – Auto-fix what you can; escalate what you can’t 4. 𝐔𝐧𝐢𝐟𝐨𝐫𝐦𝐢𝐳𝐚𝐭𝐢𝐨𝐧 & 𝐒𝐋𝐀𝐬 – Define “good enough” and enforce it before data is consumed Why it matters: ✅ Data teams – Less cleansing, faster delivery ✅ AI models – Reliable inputs = repeatable results ✅ Ops teams – Catch failing sensors before downtime ✅ Business – Avoid safety incidents, billing errors, bad decisions In the latest episode of the AI in Manufacturing podcast, I sat down with Bert Baeck, Co-Founder of Timeseer.AI, to discuss time-series data quality and reliability strategies for AI in manufacturing applications.

What really makes data and AI innovation teams succeed in manufacturing? In this episode of the AI in Manufacturing Podcast, I speak with Van Tucker, VP of Harbor Lockers by Luxer One, a company that develops and manufactures smart public lockers. We discuss the challenges and strategies for building effective innovation teams in manufacturing. Here are some of the  insights that Van shared: 𝐂𝐮𝐥𝐭𝐮𝐫𝐞 𝐢𝐬 𝐭𝐡𝐞 𝐟𝐨𝐮𝐧𝐝𝐚𝐭𝐢𝐨𝐧 Innovation thrives when people, from the boardroom to the factory floor, believe in the mission. Core values must be lived daily, not just written on posters. 𝐀𝐠𝐢𝐥𝐢𝐭𝐲 𝐨𝐯𝐞𝐫 𝐩𝐞𝐫𝐟𝐞𝐜𝐭𝐢𝐨𝐧 Instead of waiting months for a polished rollout, start simple. Test small ideas quickly, gather feedback, and iterate. Even in hardware manufacturing, lightweight R&D “sandboxes” allow experimentation without disrupting core production. 𝐌𝐚𝐧𝐚𝐠𝐢𝐧𝐠 𝐩𝐫𝐞𝐬𝐬𝐮𝐫𝐞 𝐚𝐧𝐝 𝐛𝐮𝐫𝐧𝐨𝐮𝐭 Burnout shows up in declining quality and disengagement. The best leaders don’t wait, they stay close to their teams, recognize early warning signs, and act before problems escalate. 𝐁𝐫𝐢𝐝𝐠𝐢𝐧𝐠 𝐈𝐓 𝐚𝐧𝐝 𝐎𝐓 The old silos are gone. Effective leaders create environments where engineers from IT and OT collaborate, not compete. Quick collaborative wins build trust and momentum across functions.

Most industrial processes still run on the same foundation: - Hard-coded logic in PLCs that follows predefined rules. - The intuition of process and plant engineers, built from years of experience. This combination has powered industry for decades, but it has limits. When the challenge involves many interacting variables, unknown relationships, and non-linear effects, traditional control starts to strain. Why? Because fixed rules can’t adapt fast enough to changing conditions, and even the best human intuition can only process so much complexity at once. Instead of relying on fixed instructions, RL agents learn directly from real-time feedback. They can: ✅ Adapt continuously to new conditions. ✅ Handle high-dimensional problems with countless variables. ✅ Uncover novel, more efficient strategies that humans might overlook. The result? An optimization layer that works alongside your existing control system, making it smarter, more adaptive, and capable of delivering gains where complexity used to be a roadblock In the latest episode of the AI in Manufacturing podcast, I sat down with Dr. Kyrill Schmid, the Lead AI Engineer at MaibornWolff GmbH, to discuss the application of reinforcement learning agents for optimizing industrial plants.

Can AI agents really make decisions in high-stakes industrial environments? Generative AI agents, on their own, do not have a robust understanding of cause-and-effect for real-world decision-making. However, when combined with Deep Reinforcement Learning, AI agents gain the ability to reason, learn from interaction, and make decisions that solve operational problems in complex, real-world environments, like the plant floor. Case in point. Bryan DeBois and his team at RoviSys developed an Autonomous AI agent to manage a notoriously difficult glass bottle production process, where small disruptions like temperature fluctuations can quickly push the process out of specification. Here’s how they approached it: ✅ 𝐒𝐭𝐞𝐩 1 - 𝐌𝐚𝐜𝐡𝐢𝐧𝐞 𝐓𝐞𝐚𝐜𝐡𝐢𝐧𝐠 They captured the knowledge and decision-making strategies of expert human operators and used this to train the AI agent, essentially teaching it how to respond to different operating conditions. ✅ 𝐒𝐭𝐞𝐩 2 - 𝐃𝐞𝐜𝐢𝐬𝐢𝐨𝐧 𝐒𝐮𝐩𝐩𝐨𝐫𝐭 𝐌𝐨𝐝𝐞 Initially, the agent didn’t control the process directly. It simply made recommendations. Operators reviewed the suggestions and gave feedback using a simple green/red button system. This built trust and allowed the team to validate the AI’s decisions without risk. ✅ 𝐒𝐭𝐞𝐩 3 - 𝐂𝐥𝐨𝐬𝐞𝐝 𝐋𝐨𝐨𝐩 𝐂𝐨𝐧𝐭𝐫𝐨𝐥 Only after months of successful operation in support mode did they enable full automation. Even then, strict safety measures were in place: ⇨ Limited control authority ⇨ Clearly defined operating boundaries ⇨ Automatic handover to human operators if conditions exceeded the agent’s training The Results: ⇨ Human operators typically needed 7–20 minutes to bring the process back into spec ⇨ The AI agent consistently did it in under 5 minutes ⇨ And it maintained safety by operating strictly within validated limits In the latest episode of the AI in Manufacturing podcast, I sat down with Bryan, Director of Industrial AI at RoviSys, to dive deeper into how manufacturers can leverage AI and autonomous agents to optimize manufacturing operations and improve efficiency

The industrial data stack was never built for enterprise-wide intelligence. It was built in silos, optimized for local decisions. As a result, it is not designed to support unified, contextualized, and scalable data management across an organization. And that’s why Industrial Data Platforms are essential for scaling digital transformation.  To help organizations understand what makes such a platform effective, David Ariens and The IT/OT Insider team created the Industrial Data Platform Capability Map, outlining the seven key capabilities every platform should have: 1. 𝐂𝐨𝐧𝐧𝐞𝐜𝐭𝐢𝐯𝐢𝐭𝐲 – a secure and scalable connectivity layer to integrate different data sources into the Industrial Data Platform. 2. 𝐂𝐨𝐧𝐭𝐞𝐱𝐭𝐮𝐚𝐥𝐢𝐳𝐚𝐭𝐢𝐨𝐧 & 𝐃𝐚𝐭𝐚 𝐌𝐚𝐧𝐚𝐠𝐞𝐦𝐞𝐧𝐭 – delivering data enriched with the right context, so users don’t have to gather and piece together information from multiple sources manually. 3. 𝐃𝐚𝐭𝐚 𝐐𝐮𝐚𝐥𝐢𝐭𝐲 – Detecting and fixing data issues in your pipeline, from sensor to final report. 4. 𝐃𝐚𝐭𝐚 𝐁𝐫𝐨𝐤𝐞𝐫 𝐚𝐧𝐝 𝐒𝐭𝐨𝐫𝐞 – The ability to ingest, store, and manage contextualized data at scale, enabling efficient data subscription and large-scale querying.  5. 𝐄𝐝𝐠𝐞 𝐀𝐧𝐚𝐥𝐲𝐭𝐢𝐜𝐬 – The capability to perform analytics and machine learning within the data platform, or at the edge, close to where the data is generated. 6. 𝐕𝐢𝐬𝐮𝐚𝐥𝐢𝐳𝐚𝐭𝐢𝐨𝐧 – The capability to deliver high-quality, contextualized data to users through intuitive and accessible interfaces for fast, informed decision-making. 7. 𝐃𝐚𝐭𝐚 𝐒𝐡𝐚𝐫𝐢𝐧𝐠 – The capability to openly expose platform data to external users and applications through standard interfaces and integrations.   In the latest episode of the AI in Manufacturing podcast, I sat down with David, Founder of IT/OT Insider, to dive deeper into these capabilities and how organizations can implement them. We also discussed the IT/OT Academy, an online training program designed to help IT and OT professionals build a shared vocabulary, framework, and collaboration strategy to move digital initiatives beyond pilot projects and into full-scale plant deployment.

Instead of sending data to the cloud for processing, Edge AI analyzes data right where it’s generated, on the machine, in the plant, in real time. It’s the difference between reacting later and responding now. What Happens When You Keep Intelligence at the Source? 𝐑𝐞𝐚𝐥-𝐭𝐢𝐦𝐞 𝐃𝐞𝐜𝐢𝐬𝐢𝐨𝐧𝐬 A conveyor motor vibrates abnormally. Edge AI detects the anomaly instantly and slows the line before damage occurs. 𝐒𝐦𝐚𝐫𝐭𝐞𝐫 𝐌𝐚𝐢𝐧𝐭𝐞𝐧𝐚𝐧𝐜𝐞 Time-series models forecast when a press will wear out, so teams fix it during scheduled downtime, not after it fails. 𝐐𝐮𝐚𝐥𝐢𝐭𝐲 𝐂𝐨𝐧𝐭𝐫𝐨𝐥 𝐚𝐭 𝐭𝐡𝐞 𝐄𝐝𝐠𝐞 Cameras inspect every product. Edge AI flags visual defects without ever uploading a frame to the cloud. In the latest episode of the AI in Manufacturing podcast, I sat down with Rainer Maidel, Business Development Manager at BE.services GmbH, the creators of Coligo Edge AIoT Software. We had an in-depth discussion about the application of Edge AI in the digitalization of industrial processes.

The first foundation model purpose-built for refining and petrochemicals? Here's the thing. The oil, gas, and petrochemical industry is under pressure like never before. ⇨ Demand is set to double in 15 years ⇨ Facilities are shutting down ⇨ Energy transition is colliding with operational cost realities At the same time, companies are being told AI will solve it all. But here’s the truth. Most AI was built for the internet, not industrial plants. ❌ It can’t explain its decisions ❌ It hallucinates ❌ It’s fragile with messy, real-world data ❌ It struggles with incomplete time series and unstructured reports Now apply that to a refinery running 24/7, filled with volatile compounds and extreme conditions. And you start to see the problem. AI that can’t be trusted is worse than no AI at all. That’s why Callum Adamson and his team built Orbital. The first foundation model designed specifically for refining and petrochemicals. Instead of trying to stretch general-purpose AI into high-consequence environments, Orbital is: ✅ Purpose-built ✅ Physics-aware ✅ Production-grade But more importantly, it takes a Tri-Modal Architecture that combines the following into federated intelligence: 1. 𝐓𝐢𝐦𝐞 𝐒𝐞𝐫𝐢𝐞𝐬 𝐌𝐨𝐝𝐞𝐥 – For signals and sensor data 2. 𝐏𝐡𝐲𝐬𝐢𝐜𝐬-𝐁𝐚𝐬𝐞𝐝 𝐌𝐨𝐝𝐞𝐥 – For real-world grounding 3. 𝐋𝐚𝐧𝐠𝐮𝐚𝐠𝐞 𝐌𝐨𝐝𝐞𝐥 – For intuitive interaction and explanation In the latest episode of the AI in Manufacturing podcast, I sat down with Callum, who is the Co-Founder and CEO of Applied Computing, to discuss the application of Superintelligenece in Oil, Gas, and Petrochemicals.

Part traceability in manufacturing has long relied on traditional barcodes that fail where it matters most: under heat, blasting, and coating, e.t.c.   As a result, manufacturers normally place barcodes after key part transformations.   That means, for 70%+ of the production process, you're flying blind. You're guessing which parts went through which treatments.   And when something fails? You're looking at massive recalls, supplier penalties, and lost time.   What if we could code physical parts in a way that never fades?   That’s exactly what Serra Tuzcuoglu and her Co-Founder invented. CDOT AI Code, a frequency-based, AI-readable identifier that can survive the harshest industrial conditions.   Unlike traditional visual patterns, it uses signal recognition instead of contrast, enabling it to be read even when surfaces are scratched, painted, or distorted.   A global appliance manufacturer gave them a challenge: “If your code can survive enamelling and high heat, we’ll use it.”   It did. That was the beginning. CDOT AI Code Code is now deployed globally, from Renault crankshafts to Ford EV battery trays, military tanks to printing cylinders.   In the latest episode of the AI in Manufacturing podcast, I sat down with Serra Tuzcuoglu, the CEO and Co-Founder of COSMODOT, to discuss:   ✅ How CDOT AI Code works ✅ Solution Architecture & Integration into the factory network ✅ Readiness for AI-based quality analytics and creation of digital twins. ✅ Real-world examples and Case Studies