Ep. 261 - Change Management in Manufacturing: Operators, Tribal Knowledge, and the Industrial Elder

The Siemens pop up factory is a modular CPG plant where virtual PLCs, real controllers, and a live digital twin run small batches side by side. Most factories are built for long runs of the same product. Davide Picciotto explains why that model is being stretched. COVID supply chain shocks pushed brands to add flexible areas inside existing plants, and events like the World Cup raised a sharper question: what if a factory could be stood up close to demand, run personalized products for a season, and then move? Siemens is validating the idea with PepsiCo through what Davide calls a hybrid immersive prototype: half the factory is virtual, half is real, and both run as one system. Pininfarina, the design house behind decades of Ferraris, designed the modular shell. Andrea Sassetti walks through the live environment that debuted at Hannover Messe. Four PLCs, some physical and some virtual on Industrial Edge, feed the Industrial Information Hub, which adds a semantic layer and pushes data to the cloud. Virtual PLCs also simplify firmware and security patching. A capability matching engine reads each recipe step, checks which equipment is available and stocked with the right material, and routes the batch in real time. The same engine builds rainbow pallets of mixed product layers, using the Simatic robot library over SRCI so the same code runs on a Universal Robots cobot or any compatible arm. The most interesting idea is in the mixer. Traditionally, mix time and speed are hard coded into the recipe. Siemens ran thousands of Simcenter fluid dynamics simulations across viscosities, particle sizes, and impeller types, trained a model in AI Studio from the Altair RapidMiner acquisition, and deployed it on Industrial Edge. The recipe now specifies material properties, and the model infers setpoints on the fly, even when an ingredient comes from a new supplier. Davide closes on what is still missing: digital threads between engineering and execution remain incomplete, with one CPG company counting thousands of clicks to launch a single product. PepsiCo has reported a 20 percent throughput gain at a US Gatorade plant from its digital twin work with Siemens and NVIDIA. About the Guests Davide Picciotto is Digital Enterprise Industry Director for CPG at Siemens Digital Industries Software. In 20 years at Siemens he moved from MES in Genoa to food and beverage presales in the US, helped found the Digital Enterprise team in 2015, and ran the Siemens showcase at MxD in Chicago. Andrea Sassetti is an Innovation Manager on the Siemens Digital Enterprise team. He joined Siemens in 2009 as an MES developer on what is now Opcenter, then moved to the US in 2018 as solution architect for the Siemens showcase at MxD. Timestamps 0:00 Welcome and PACK EXPO details 2:10 Meet Davide Picciotto and Andrea Sassetti 5:00 What Siemens Digital Enterprise covers 8:20 What is the pop up factory 11:20 The business case for modular factories 17:50 Pininfarina designs the modular factory 24:20 PepsiCo and the validation stage 28:30 Live demo: Industrial Edge and virtual PLCs 35:00 Rainbow pallets and SRCI robots 40:00 AI mixer setpoints trained on simulation 45:00 What visitors will see at PACK EXPO 51:10 The future of pop up factories References Siemens Industrial Information Hub: https://www.siemens.com/en-us/products/simatic-apps/industrial-information-hub This episode is sponsored by Siemens is a global technology company operating across industrial automation, digital software, smart infrastructure, and mobility. Siemens Xcelerator is its open digital business platform and marketplace. https://www.siemens.com About Your Hosts Vladimir Romanov is a co-host of The Manufacturing Hub Podcast and the founder of Joltek, an independent manufacturing and industrial automation consulting firm specializing in modernization strategy, digital transformation, and workforce development. Joltek works with manufacturers and investors to de-risk modernization and build the int

There is no pyramid anywhere in the ISA 95 standard. David Schultz explains what is actually inside it and how to turn it into a working data model. Most engineers meet ISA 95 as a layered triangle on LinkedIn and never get past it. David Schultz of Amárach StackWorks takes that image apart. The pyramid appears nowhere in the standard, which exists to define how business systems such as ERP exchange information with manufacturing and control systems. Eight parts are published today, Part 1 refreshed in 2025, with Part 9 still in committee. Start with Parts 1 and 2. Part 1 sets the scope, Part 2 delivers the resource, information and process models. Parts 3 and 4 add the activity models and the work master. Parts 5 and 6 supply the verbs and exchange patterns. From there it collapses into a frame you can hold in your head: four resources of equipment, physical asset, personnel and material, four domains of production, maintenance, quality and inventory, and four exchanges covering definition, capability, schedule and performance. Use what works and leave the rest. On adoption David is candid. The full set runs roughly $1,400, and the paywall keeps it out of most LLM workflows because ISA would forfeit copyright by opening it up. Where it breaks is data governance. When every site builds bespoke models, stitching them together means inventing a fourth abstraction. Vlad offers a first hand version from Procter and Gamble, where nobody at plant level owned the data structure moving from Rockwell controls into Proficy. For brownfield sites the advice is the Strangler fig pattern: leave what runs and replace functionality over time. The second half demystifies knowledge graphs using a family tree. A graph stores a subject, a predicate and an object, and the relationship is a first class citizen, not a foreign key. RDF defines the things, RDFS adds schema, OWL adds inference so a system can derive siblings from a shared parent, and SHACL constrains the shapes. Applied to a tank, that means volume and level as properties plus the resource relationship network to map which pump feeds which vessel. Rhize implemented the entire 2018 standard as a graph. David stays cool on AI. A shared ontology means a model never guesses whether two names match, but he expects it to underwhelm. His book pick is The Semantic Web for the Working Ontologist. About David Schultz David Schultz is a principal consultant and co founder of Amárach StackWorks, a consultancy describing itself as standards based manufacturing operations management built on a modern technology stack. He has spent close to 30 years in process control and factory automation, starting at the device layer before moving through SCADA into MES and ERP integration. He was previously a senior consultant at Rhize Manufacturing Data Hub, where ISA 95 was implemented as a knowledge graph. He last appeared on Manufacturing Hub for Episode 85. https://www.amarach.io Timestamps 0:00 Introduction and ICC preview 2:05 David Schultz and Amárach StackWorks 4:10 What ISA 95 actually is and why the pyramid is not in it 7:20 Where to start with the standard 9:40 The ROI of a shared data model 15:00 Which parts of the standard to read first 20:40 The adoption gap: cost and misconceptions 24:20 Who inside a plant should own the model 32:50 Knowledge graphs: RDF, RDFS, OWL and SHACL 43:30 The standard is not prescriptive about protocols 47:40 Knowledge graphs, ontologies and AI 53:40 Future of standards, career advice and book pick References ISA 95 Standard and Training: https://www.isa.org Rhize ISA 95 Learning Resources: https://docs.rhize.com/isa-95/ About Your Hosts Vladimir Romanov is a co-host of The Manufacturing Hub Podcast and the founder of Joltek, an independent manufacturing and industrial automation consulting firm specializing in modernization strategy, digital transformation, and workforce development. Joltek works with manufacturers and investors to de-risk modernization and build the int

Most conversations about starting a systems integrator skip the parts that sink people. Dave and Vlad spend this one on insurance, licenses and procurement. After a month of founder interviews, Dave and Vlad close the series with their own experience. Vlad started at plant level with Procter and Gamble and Kraft Heinz, deciding how outside integrators would execute, before crossing over to work for an integrator across food and beverage and CPG packaging. Dave came from aviation and aerospace through a German machine builder and a distributor, then ran a small integration company doing MES, OEE and track and trace. Both reached the same conclusion. Integration starts after the requirements are written, so you quote to a spec and a budget, build the thing, and real handover rarely follows. Blame does not sit in one place. Vlad puts it plainly: people resent a solution that has not actually solved something for them. But end users say they do not need training and ask for the lowest number, and training is the first line cut. Dave counters that if a customer wants five dashboards on an ugly six inch HMI that has sat in the plant for fifteen years, that is the job. Vlad's answer is that your time is finite, and a year on work you cannot show anyone positions you as the person who does that work. The technical centerpiece is a migration that sounds nearly free. Moving from RSView 32 to FactoryTalk View gets justified because conversion tools exist. Then come the undocumented Visual Basic scripts written around tag limits on old controllers, which no conversion tool touches. Then the PLCs will not support the new terminals and servers, so the controllers go too. Then the networks underneath, often Data Highway Plus. Then servers, remote racks, IO, sensors and drives. Vlad argues these decisions get made by people who never price the second and third order effects. An integration firm is a service firm, billing time and materials or a fixed fee, with a markup on hardware and software. Vlad quotes roughly $200 a month for $2 million of liability coverage, a capable laptop, and software licenses reaching tens of thousands of dollars per seat without preferential pricing, usually several across PLC and HMI programming. Then procurement, which nobody warns new integrators about. Dave has cleared it in twenty minutes and has waited seven months, three or four of those with lawyers arguing over a master service agreement. Vlad closes on complexity. Industrial equipment no longer ships with a full schematic and a six dollar replacement sensor, so ramp up time climbs while the population who understand electrical, mechanical and process together shrinks as the best of them retire. Dave sees it from the other side. Entry now takes an LLC, a laptop and a phone, and he expects private equity to keep buying medium and large integrators over the next five years. Timestamps 0:00 Introduction 10:30 Vlad and Dave's paths into integration 18:40 Why integrators start after the requirements are written 23:40 Who is to blame when training gets cut 30:00 RSView 32 migrations and the hidden cascade 36:40 Choosing where you sit on the integration spectrum 42:20 How integrators make money and what it costs to start 48:40 Procurement and getting paid 53:10 The future of integration and rising complexity 1:01:40 Career advice 1:13:20 The industrial app marketplace References The Little Red Book of Selling by Jeffrey Gitomer: https://www.amazon.com/Little-Red-Book-Selling-Principles/dp/0743572548 Influence, New and Expanded by Robert Cialdini: https://www.amazon.com/Influence-New-Expanded-Psychology-Persuasion/dp/0062937650 About Your Hosts Vladimir Romanov is a co-host of The Manufacturing Hub Podcast and the founder of Joltek, an independent manufacturing and industrial automation consulting firm specializing in modernization strategy, digital transformation, and workforce development. Joltek works with manufacturers and investors to de-risk modernization a

Most people ask how to start a systems integrator. Rich Kuechenmeister has effectively built one five times, and the lesson is that the parent business decides what you become. Rich is owner and partner at Fishbone Technical Services. Before that he built automation groups inside a semiconductor equipment OEM, an automation components manufacturer, a products distributor, and a custom machine builder where he stood up a UL 508A panel shop from nothing. Each of those businesses sold something different, and that changed the engineering. At a machine builder the machine is the dog and the controls are the tail. At a distributor you build toolkits so somebody else can build the solution, handcuffed to the lines you carry. Rich argues that constraint is what made him good, because a single vendor solution is easy while four component areas from four vendors is where real integration lives. His definition of a control systems integrator stays traditional. SCADA and below, field instrumentation through PLC and HMI, bridging a little into MES. Fishbone works line control, process control on boilers, pumps and valves, and material handling where suppliers ship a motor and a few sensors and nothing more. The business advice is specific. He would rather have one $100,000 project than ten $10,000 projects, because that is one project management cycle instead of ten. He cautions against mixing maintenance work and capital projects without a plan, since context switching between a vision system callout and a pump control program burns budget on both. Dave adds the pattern he keeps seeing: come in at the bottom and you get saved in the phone as the maintenance guy, never the capital project guy. Rich hires on desire, attitude and aptitude, and admits he was once the bottleneck himself. Looking ahead, he expects acquisitions to keep rolling up integrators and leaving a void underneath, since a twenty person firm will not get out of bed for the projects a five person firm lived on. On AI he is practical. He turned three day estimates into eight estimates in a single day, but insists code writing is a subset of the job. He is also still doing a PLC-5 conversion and a DeviceNet upgrade. About Rich Kuechenmeister Rich Kuechenmeister is owner and partner at Fishbone Technical Services, a nationwide remote technical services company he co-founded in 2021 with business partner Robert Summers. Fishbone is based in Texas and covers plant engineering and control systems integration from field instrumentation through PLC, HMI and SCADA, with some MES. Rich began his career in the United States Navy working on radars, servos and guidance systems, then spent three decades in industrial automation building engineering groups before starting Fishbone. https://fishbonetechnical.com Timestamps 0:00 Introduction 2:20 Inside Fishbone Technical Services 5:20 From the Navy to OEMs, distributors and machine builders 9:00 Becoming unafraid and starting a company 15:40 What a control systems integrator actually is 20:40 How the parent business shapes the work 31:40 Filling the sales hopper and qualifying projects 36:40 Maintenance work versus capital projects 43:00 Hiring for desire, attitude and aptitude 48:20 Letting go of the client relationship 56:20 Distributors, acquisitions and the void they leave 1:04:40 AI, estimating and reverse engineering PLC code References PrOTect IT All by Aaron Crow: https://www.amazon.com/PrOTect-All-Plain-Spoken-Cybersecurity-Protecting/dp/B0HBM3F4XS ISA/IEC 62443 Series of Standards: https://www.isa.org/standards-and-publications/isa-standards/isa-iec-62443-series-of-standards About Your Hosts Vladimir Romanov is a co-host of The Manufacturing Hub Podcast and the founder of Joltek, an independent manufacturing and industrial automation consulting firm specializing in modernization strategy, digital transformation, and workforce development. Joltek works with manufacturers and investors to de-risk modernization and build the inter

Oil and gas automation runs on royalty accounting as much as process control. Dustin Symes explains how that shapes every SCADA decision JPI Solutions makes. Canadian oil and gas is structured differently than most assume. In the US mineral rights usually sit with individuals, so custody transfer draws a hard line between upstream and midstream. In Canada the provinces hold most rights, so a producer can gather 100 wells into one plant and sell to a pipeline company there. That lands on the automation: much of the SCADA data collection exists to answer who is paying whom and how much royalty is owed. Once a well is drilled, automation might find 5 to 10 percent off nameplate, so the work protects a long production tail rather than chasing throughput. Mining behaves nothing like it. Thinner margins, $100 million capital projects, an hour long elevator ride down, underground space running 20 miles across. Those are small cities: control systems cover substations, air monitoring and lighting as well as conveyors and pumps. Mining clients want programmers on site rather than remote, and they buy for continuity: a major vendor PLC paired with their DCS, meant to last 20 years. Horizontal drilling has rescaled oil and gas, from one well and a dozen sensors to a 32 well pad carrying 300 IO, pushing designs toward redundant PLCs and ring networks. The genuinely new technology lands on the data layer, where MQTT and report by exception are displacing polled Modbus. Dustin set two gates before his first hire: enough cash banked to make that person whole for a month or two, and enough overload to have work worth delegating. He hired out of the instrumentation program at SAIT rather than a 15 year instrument mechanic. When he weighed taking on partners, dozens of people advised against it. He brought in Dan and Brent Lawther anyway in 2018, went from five people that spring to ten by the fall, and says JPI would not exist without them. His hardest won lesson is unglamorous. Communication is the only consistent complaint he has ever had from a customer. Nobody complains about the code. They complain that nobody told them it was finished. On AI, JPI is decompiling Rockwell files into shutdown keys and landing 85 to 90 percent of the way there on a good prompt, and running log analysis across FactoryTalk directory, HMI and historian logs to find real root causes. He expects PLC code to become text so it can be version controlled like software, and customers to treat that as baseline within five years. About Dustin Symes Dustin Symes is Chief Technology Officer at JPI Solutions, a Calgary based systems integrator serving oil and gas and mining clients across North America from four offices in Alberta, BC and Saskatchewan. A Red Seal journeyman electrician who moved into automation, he founded JPI in 2017. JPI covers process control from the sensor to the HMI, then up into SCADA and corporate data. https://www.jpisolutions.ca Timestamps 0:00 Introduction 5:50 Why he started JPI Solutions 10:50 Oil and gas primer: upstream, midstream, downstream 12:50 What makes mining automation different 16:10 Where JPI sits in the automation stack 28:30 Bigger well pads, redundant PLCs and MQTT 41:20 Bringing on partners when nobody recommended it 51:10 Communication as the only consistent complaint 1:04:40 The future of integration, AI and knowledge graphs References Radical Candor by Kim Scott: https://www.amazon.com/Radical-Candor-Kick-Ass-Without-Humanity/dp/1250103509 Creativity, Inc. by Ed Catmull: https://www.amazon.com/Creativity-Inc-Overcoming-Unseen-Inspiration/dp/0812993012 ThredCloud: https://www.thredcloud.com About Your Hosts Vladimir Romanov is a co-host of The Manufacturing Hub Podcast and the founder of Joltek, an independent manufacturing and industrial automation consulting firm specializing in modernization strategy, digital transformation, and workforce development. Joltek works with manufacturers and investors to de-risk modernization a