by Gary Mintchell | Apr 10, 2026 | Manufacturing IT, Operations Management
During a recent conversation, the concept of the definition of Smart Manufacturing was broached. I began to think about just what is Smart Manufacturing. I hoped it wasn’t like Clark Griswald’s definition of Christmas talking with Ruby Sue (National Lampoon’s Christmas Vacation)—well, I guess it means something different to everyone. I thought there must be more enterprise strategy to the concept.
Claude.ai is my LLM chat researcher of choice. It’s generally pretty good. I’ve never used it to write before. I guided it through a series of questions. Each answer suggested another question.
If Smart Manufacturing is simply a continuation of digitizing manufacturing data, that is something I worked on in 1976. Surely there is more. So I asked about integrating with Lean. That led to Systems of Systems. I have been fascinated with Arthur Koestler’s concept of holons ever since reading Ken Wilber’s A Brief History of Everything. Perhaps a holarchy describes a governance model or overall management model for Smart Manufacturing.
The following essay was written mostly by Claude. I gave it my website and told it to write an essay in my style. I thought I might rewrite it to a degree, but that was too much. This is pretty good. What do you think? (I think I won’t ask it to do this again!)
###
From sensor to enterprise, today’s factories are not simply automated — they are structured as nested, self-governing wholes. A philosophical and engineering framework for understanding what smart manufacturing really is.
KEY INSIGHT
Smart manufacturing is not just a technology upgrade. It is an architectural and philosophical shift — one that was anticipated by a philosopher in 1967 and formalized by engineers in 1994, long before Industry 4.0 was coined.
Smart manufacturing has become one of the most used — and most loosely defined — terms in industrial discourse. Vendors apply it to any product with a sensor. Consultants apply it to any initiative with a digital roadmap. Yet beneath the marketing noise, something genuinely coherent and intellectually rich is taking shape on factory floors around the world. This essay argues that smart manufacturing is best understood not merely as a technology platform, but as a hierarchically integrated, emergent system — one that finds its clearest theoretical expression in the engineering concept of a system of systems and the philosophical concept of a holarchy.
I. Defining Smart Manufacturing
The difficulty with smart manufacturing begins with its definition. No single institution owns the term, and the leading definitions reflect the different vantage points of their authors.
NIST — NATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY
“Fully-integrated, collaborative manufacturing systems that respond in real time to meet changing demands and conditions in the factory, in the supply network, and in customer needs.”
SMART MANUFACTURING LEADERSHIP CONSORTIUM (SMLC)
“The ability to solve existing and future problems via an open infrastructure that allows solutions to be implemented at the speed of business while creating advantaged value.”
MITTAL ET AL. (2019) — PROCEEDINGS OF THE INSTITUTION OF MECHANICAL ENGINEERS
“A set of manufacturing practices that use networked data and information and communication technologies (ICTs) for governing manufacturing operations.”
DELOITTE — PAUL WELLENER, VICE CHAIRMAN, US INDUSTRIAL PRODUCTS
The widespread digitization of all manufacturing practices — from the factory floor to all aspects of business, including product design, supply chain, production, distribution, and sales.
What these definitions share is a common architecture: real-time data integration, cross-system connectivity, and intelligent responsiveness to dynamic conditions. What they lack is a unified conceptual framework for understanding why those properties, taken together, produce something qualitatively different from traditional automation. The system-of-systems and holarchy frameworks, introduced below, supply that missing frame.
II. Smart Manufacturing and Lean
Lean manufacturing — rooted in the Toyota Production System — is a philosophy focused on eliminating waste (muda), reducing variability, and maximizing customer value through continuous improvement (kaizen). Its relationship to smart manufacturing is complementary and synergistic, though the two are conceptually distinct.
Lean provides the operational philosophy; smart manufacturing provides the technological infrastructure that extends and amplifies it. While lean reduces complexity by eliminating waste and non-value-adding activities, smart manufacturing uses sensors, robotics, cloud computing, and analytics to increase transparency, improve communication, and adapt to change — capabilities that lean alone cannot provide at scale.
When lean or smart manufacturing is applied alone, each can reduce conversion costs by roughly 15%. The integrated approach reduces them by as much as 40% — and can simultaneously reduce poor-quality costs by 20% and work-in-process inventory by 30%.
BOSTON CONSULTING GROUP
The missing ingredient in lean-only implementations is visibility. Lean principles help improve production speed, but they do not inherently integrate processes to provide visibility across the entire value chain. Failures are difficult to detect and correct at scale when monitoring depends on manual observation and paper-based documentation. Smart manufacturing fills this gap precisely.
Empirical research confirms that the two paradigms tend to co-evolve. A 2022 study published in the International Journal of Production Economics found that while lean is applied without smart technologies in some facilities, smart technologies are almost always applied in conjunction with lean — suggesting that lean principles serve as a necessary precondition for effective smart implementation. The researchers found that Manufacturing Execution Systems (MES) had the strongest such dependency on lean of all smart technologies studied.
Hybrid frameworks — sometimes called “Lean Industry 4.0” or “Dynamic Lean 4.0” — have formalized this integration, developing digitized extensions of classic lean instruments such as Sustainable Value Stream Mapping, Extended SMED, and Digital Poka-Yoke. These tools automate waste detection, accelerate improvement cycles with real-time data, and sustain gains that lean alone may struggle to maintain over time.
III. Smart Manufacturing and Industry 4.0
Smart manufacturing and Industry 4.0 (Industrie 4.0) are closely related but not identical. Industry 4.0 is a specific initiative originating in Germany; smart manufacturing is a broader, global umbrella concept. In the United States the latest industrial evolution has been called “smart manufacturing”; in Europe, “Industry 4.0.” Both label what is effectively the Fourth Industrial Revolution.
The German government launched Industry 4.0 as part of its high-tech national strategy, with the goal of creating the intelligent “Smart Factory” — characterized by adaptability, resource efficiency, ergonomic design, and integration of customers and partners in value processes. Its technological foundation consists of cyber-physical systems and the Internet of Things. The Reference Architecture Model Industry 4.0 (RAMI 4.0) was developed to provide structural guidance for implementation.
Some definitions of smart manufacturing explicitly treat Industry 4.0 as its technological substrate: smart manufacturing deploys the technologies of Industry 4.0 to enhance manufacturing performance and optimize energy and workforce utilization. The distinction, when drawn, is primarily one of scope. Industry 4.0 specifies a set of enabling technologies and a broad conceptual framework; smart manufacturing describes the operational and strategic outcome of applying those technologies across the full enterprise.
That enterprise scope is significant. Smart manufacturing includes Industry 4.0 digital technologies, but it also digitizes activities outside the factory floor — product design, supply chain, distribution, and sales. Industry 4.0 is the engine; smart manufacturing is the vehicle, the road, and the destination.
IV. The Role of MES
If Industry 4.0 provides the technological substrate of smart manufacturing and lean provides its operational philosophy, the Manufacturing Execution System (MES) provides its operational nervous system.
MES is a computerized system used in manufacturing to track and document the transformation of raw materials to finished goods. It provides the information that helps manufacturing decision-makers understand how current conditions on the plant floor can be optimized to improve production output, working as a real-time monitoring system to enable control of multiple elements of the production process.
In manufacturing operations management, MES serves as a bridge between the planning and control systems of the enterprise — such as the ERP — and actual manufacturing operations. It captures data from machines, sensors, and operators in real time to provide accurate, up-to-date information about production activities. In smart manufacturing terms, MES is the layer that makes the factory’s data coherent and actionable.
Smart factories depend on MES as the “hidden engine” that connects systems, optimizes processes, and delivers the real-time insights needed to maintain competitive advantage.
INDUSTRY ANALYSIS, 2025
The integration of MES with the Industrial Internet of Things (IIoT) is now among the most important trends in manufacturing technology. MES systems use IIoT technologies to collect data from sensors, machines, and connected devices in real time, enabling enhanced visibility, predictive analytics, remote monitoring, and process optimization. Machine learning techniques are applied to identify patterns, enabling predictive maintenance and autonomous quality control.
As of 2025, manufacturing execution systems are shifting from rigid, monolithic on-premise software to modular, cloud-connected platforms — integrating AI to move from descriptive reporting toward predictive and prescriptive capabilities. This architectural evolution is directly relevant to the framework developed below.
V. Smart Manufacturing as a System of Systems
With this landscape established, we can now ask a more structural question: what kind of entity is smart manufacturing? The answer, this essay argues, is that it is a system of systems — and understanding it as such explains many of its most characteristic properties.
The foundational definition comes from Mark Maier’s landmark 1998 paper, “Architecting Principles for Systems-of-Systems” (Systems Engineering). Maier identified five key characteristics of any system of systems:
Operational independence of component systems. Managerial independence of component systems. Geographical distribution. Emergent behavior — the system performs functions that do not reside in any single component. Evolutionary development — the system is never “delivered” in a final state but evolves continuously.
Each of these maps directly onto smart manufacturing. ERP, MES, SCADA, PLM, IIoT platforms, robotics, and quality systems each operate independently and are typically owned by different organizational units and sourced from different vendors — yet they interoperate to produce smart manufacturing. Their integration spans geographies, from shop floor to cloud to supply chain partner. Their most valued capabilities — real-time responsiveness, predictive maintenance, autonomous scheduling — do not reside in any single constituent system but emerge from integration. And smart manufacturing is never finished: technologies evolve asynchronously and are adopted incrementally across the enterprise.
The ISO/IEC/IEEE 21839 definition is instructive: a system of systems is “a set of systems or system elements that interact to provide a unique capability that none of the constituent systems can accomplish on its own.” This is precisely why the BCG research shows that the value of smart manufacturing exceeds the sum of its parts — 40% conversion cost reduction versus 15% for any single approach. Emergent value is not a marketing claim; it is a structural property of system-of-systems architecture.
VI. The Holon: A Deeper Framework
The system-of-systems framing is primarily an engineering concept — useful for architecture, interoperability, and governance. But it leaves a philosophical question unanswered: what is the nature of the entities being integrated? Here, the concept of the holon, developed by Arthur Koestler in his 1967 work The Ghost in the Machine, provides a richer and more generative answer.
Koestler coined the term holon to describe natural organisms as composed of semi-autonomous sub-wholes that are linked in a form of hierarchy — a holarchy — to form a whole. A holon is something that is simultaneously a whole and a part. It possesses a degree of independence and can handle contingencies without instruction from higher authorities, while simultaneously being subject to control from those authorities. The first property ensures stability; the second ensures integration.
“Everything is composed of holons, which are simultaneously parts and wholes. Each holon is a constituent of a larger entity while also containing smaller holons within.”
ARTHUR KOESTLER, THE GHOST IN THE MACHINE, 1967
Koestler identified two fundamental and opposed tendencies in every holon: a self-assertive tendency (the holon asserts its own identity, autonomy, and integrity) and an integrative tendency (the holon subordinates itself to the larger whole). He used the Janus metaphor: every holon faces simultaneously inward — representing the whole — and outward — representing the part.
His goal was explicitly to transcend the interminable debate between reductionism (only the parts matter) and holism (only the whole matters). Holarchy, he argued, renders the debate obsolete: every entity is simultaneously a self-contained whole and a dependent part, and the system is understood only when both dimensions are seen together.
VII. The Holarchy of Smart Manufacturing
Manufacturing engineers recognized the power of Koestler’s framework in the early 1990s. The Holonic Manufacturing Systems (HMS) Consortium, established in 1994 as part of the Intelligent Manufacturing Systems (IMS) initiative, formalized the application. In an HMS, key elements — machines, cells, factories, parts, products, operators, and teams — are modeled as holons with autonomous and cooperative properties. This is not a metaphor; it is an engineering architecture.
Smart manufacturing, viewed through the holon lens, forms a coherent five-level holarchy:
THE HOLARCHY OF SMART MANUFACTURING
L5
Smart Manufacturing encompassing holon · strategy & value chain
L4
Enterprise / ERP business strategy · supply chain
L3
Factory / MES production execution · OEE · quality
L2
Cell / Workstation scheduling · local coordination
L1
Machine / Sensor / Robot autonomous operation · IIoT data
At each level, the holon is simultaneously a self-contained, functioning whole and a dependent part of the level above. A CNC machine continues operating when the cell controller loses connectivity. A manufacturing cell continues producing when the MES is in maintenance mode. An MES continues managing the shop floor when the ERP is offline. Each level asserts its autonomy (self-assertive tendency) while remaining integrated into the larger whole (integrative tendency).
What makes this more than an organizational chart is the concept of emergent capability. The most valued properties of smart manufacturing — predictive maintenance, autonomous quality control, real-time supply chain adaptation — do not reside at any single holon level. They emerge from the relationships between levels. No sensor knows what a delivery delay means. No MES knows what a margin squeeze means. But a holarchy that connects sensor to cell to factory to enterprise can generate responses to both, in real time, without human intervention at each step.
The Structural Homology with Industry 4.0
One of the more striking findings in recent research is the structural homology between the holonic architecture developed in the 1990s and the I4.0 Component — the fundamental unit of RAMI 4.0, Germany’s reference architecture model for Industry 4.0. Researchers have implemented OPC-UA and AutomationML in combination with holonic approaches to develop digital manufacturing systems explicitly aligned with RAMI 4.0. The I4.0 Component, with its asset shell and information model, is structurally isomorphic with the holon’s dual nature: it is both an autonomous operational entity and an interoperating node in a larger industrial network.
This convergence is not coincidental. It suggests that the holonic architecture represents something close to a natural structure for complex adaptive manufacturing systems — one that engineers in Germany in 2011 and engineers in Japan in 1994 arrived at independently, because the logic of the problem demands it.
MES as the Acknowledged Integration Layer
Maier identified several types of system of systems, including Acknowledged SoS — those with central management and resources, but where constituent systems retain their independence. Smart manufacturing fits this type well, and MES is the acknowledged integration layer that makes it work. MES provides the coordination, visibility, and real-time data flow that connect the machine/sensor holons at Level 1 to the enterprise holons at Level 4, without eliminating the autonomous operational identity of any level.
This also explains the empirical finding that MES has the strongest dependency on lean principles of all smart manufacturing technologies. Lean provides the operational discipline — the waste elimination and continuous improvement logic — that gives MES its vocabulary of meaningful metrics: OEE, cycle time, first-pass yield, changeover time. Without lean, MES is a data collector. With lean, it is a performance management system. Together, within the holarchy, they constitute the operational intelligence of the smart factory.
VIII. Implications for Practitioners
This framework is not merely philosophical. It has direct practical implications for organizations implementing or evaluating smart manufacturing initiatives.
Interoperability standards are the interface architecture of the holarchy. Standards such as OPC-UA (for machine-to-system communication), ISA-95 (for MES-to-ERP integration), and RAMI 4.0 (for the I4.0 Component) are not bureaucratic formalities — they are the interfaces that allow holons at different levels to maintain both their autonomy and their integration. Organizations that invest in standards-compliant architecture are building a genuine holarchy; those that rely on point-to-point custom integrations are building a fragile network that will resist evolution.
Implementation must be holarchy-aware, not component-wise. A persistent failure mode in smart manufacturing projects is treating constituent technologies — MES, IIoT platforms, AI analytics — as independent installations. The holonic framework makes clear why this fails: the value of smart manufacturing is emergent, residing in the relationships between levels, not in any individual component. Implementation strategy must be designed to cultivate those relationships from the outset.
Governance must match the holonic structure. In a holarchy, each level retains managerial independence. In practice, this means that the IT organization, the OT organization, the supply chain organization, and the business units will each have legitimate ownership interests in their respective holons. Smart manufacturing governance frameworks must accommodate this — not by subordinating all levels to a single authority, but by establishing the integrative interfaces through which autonomous holons coordinate voluntarily.
Maturity should be assessed at every holon level. A factory that has a sophisticated MES but primitive machine-level data collection is a holarchy with a missing level. A factory with excellent IIoT sensor coverage but no integration to ERP is a holarchy that stops at Level 3. Maturity models for smart manufacturing should be structured holarchically — assessing capability at each level and the quality of integration between levels.
The human holon is not optional. Koestler’s holarchy includes human beings at every level. The Human-Cyber-Physical System (HCPS) literature is explicit that workers, teams, and organizations are holons in the smart manufacturing system, not merely operators of it. Implementations that treat the human role as residual — what remains after automation — miss a structural component and underperform accordingly.
Conclusion
Smart manufacturing is one of the most consequential transformations in the history of industrial production. But its significance is obscured when it is described primarily as a technology deployment — a list of sensors, platforms, and AI applications. The system-of-systems framework reveals its architectural logic: a set of operationally and managerially independent systems, geographically distributed, producing emergent capabilities that none could achieve alone, evolving continuously rather than arriving complete.
The holarchy framework goes further. It reveals that smart manufacturing is not just a way of connecting systems — it is a way of understanding what those systems are. Every constituent, from a single sensor to a global enterprise, is simultaneously a self-contained whole and a dependent part. The value of smart manufacturing — the 40% cost reduction, the predictive maintenance, the autonomous quality control — is not the sum of its parts. It is a property of the whole that emerges from, and only from, the integrity of every level of the holarchy.
Arthur Koestler formulated the holon in 1967 to resolve a debate in philosophy of science. Manufacturing engineers formalized it in 1994 to solve problems of shop floor complexity. German industry embedded its logic in RAMI 4.0 in 2015. The convergence is not coincidence. It is evidence that the holarchy is the natural structure of any complex adaptive system that must be simultaneously stable and responsive, autonomous and integrated, local and global.
That is precisely what a smart factory must be. And that is what smart manufacturing, at its best, becomes.
by Gary Mintchell | Apr 9, 2026 | Generative AI
Cal Newport, Computer Science PhD and Professor at Georgetown University, explains AI, LLMs, and the like better than anyone else I follow. His newsletter from last month, Why Hasn’t AI Made Work Easier?, explains some of the reports we’ve begun hearing through the media noise.
He writes:
I’ve been studying the intersection of digital technology and office work for quite some time. (I find it hard to believe that my book, Deep Work, just passed its ten-year anniversary!?) Here’s a pattern I’ve observed again and again:
(And, yes, I’ve lived through these…and more.)
- A new technology promises to speed up some annoying aspects of our jobs.
- Everyone gets excited about freeing up more time for deep work and leisure.
- We end up busier than before without producing more of the high-value output that actually moves the needle.
- This happened with the front-office IT revolution, and email, and mobile computing, and once again with video-conferencing.
Will AI be anything different?
I’m now starting to fear that we’re beginning to encounter the same thing with AI as well.
My worries were stoked, in part, by a recent article in the Wall Street Journal, titled “AI Isn’t Lightening Workloads. It’s Making Them More Intense.”
Based on some actual research:
The piece cites new research from the software company ActivTrak, which analyzed the digital activity of 164,000 workers across more than 1,000 employers. What makes the study notable is its methodology: it tracked individual AI users for 180 days before and after they began using these tools, providing clear insight into what changed. The results?
“ActivTrak found AI intensified activity across nearly every category: The time they spent on email, messaging and chat apps more than doubled, while their use of business-management tools, such as human-resources or accounting software, rose 94%.“
Ah, not everything was affected.
The one category where activity was not intensified, however, was deep work:
“[T]he amount of time AI users devoted to focused, uninterrupted work—the kind of concentration often required for figuring out complex problems, writing formulas, creating and strategizing—fell 9%, compared with nearly no change for nonusers.”
Why?
It’s not quite clear why AI tools are having this impact. One tantalizing clue, however, comes from Berkeley professor Aruna Ranganathan, who is quoted in the article saying: “AI makes additional tasks feel easy and accessible, creating a sense of momentum.”
I lived through these changes and concur:
This points toward a pattern similar to what happened when email first arrived. It was undeniably true that sending emails was more efficient than wrangling fax machines and voicemail. But once workers gained access to low-friction communication, they transformed their days into a furious flurry of back-and-forth messaging that felt “productive” in the abstract, activity-centric sense of that term, but ultimately hurt almost every other aspect of their jobs and made everyone miserable.
Click on the Follow button at the bottom of the page to subscribe to a weekly email update of posts. Click on the mail icon to subscribe to additional email thoughts.
by Gary Mintchell | Apr 8, 2026 | Generative AI
I have cued up a couple of analyses on AI. Referring to Om Malik’s observation about generating news velocity, it’s hard to keep up. But the two I have remain relevant.
The first from an analyst I’ve followed for many years, Benedict Evans, End of Network Effect essay looks at the business model (or lack thereof) of OpenAI.
He opens:
OpenAI has some big questions. It doesn’t have unique tech. It has a big user base, but with limited engagement and stickiness and no network effect. The incumbents have matched the tech and are leveraging their product and distribution. And a lot of the value and leverage will come from new experiences that haven’t been invented yet, and it can’t invent all of those itself. What’s the plan?
He compares an OpenAI executive with Steve Jobs—where do you start when developing technology and a product?
“Jakub and Mark set the research direction for the long run. Then after months of work, something incredible emerges and I get a researcher pinging me saying: “I have something pretty cool. How are you going to use it in chat? How are you going to use it for our enterprise products?”
– Fidji Simo, head of Product at OpenAI, 2026
“You’ve got to start with the customer experience and work backwards to the technology. You can’t start with the technology and try to figure out where you’re going to try to sell it”
– Steve Jobs, 1997
Pretty damning.
Evans isolates four fundamental strategic questions.
Where’s the unique selling proposition?
First, the business as we see it today doesn’t have a strong, clear competitive lead. It doesn’t have a unique technology or product. The models have a very large user base, but very narrow engagement and stickiness, and no network effect or any other winner-takes-all effect so far that provides a clear path to turning that user base into something broader and durable. Nor does OpenAI have consumer products on top of the models themselves that have product-market fit.
It’s very early in the market cycle.
Second, the experience, product, value capture and strategic leverage in AI will all change an enormous amount in the next couple of years as the market develops. Big aggressive incumbents and thousands of entrepreneurs are trying to create new features, experiences and business models, and in the process try to turn foundation models themselves into commodity infrastructure sold at marginal cost. Having kicked off the LLM boom, OpenAI now has to invent a whole other set of new things as well, or at least fend off, co-opt and absorb the thousands of other people who are trying to do that.
They are all in the same boat.
Third, while much of this applies to everyone else in the field as well, OpenAI, like Anthropic, has to ‘cross the chasm’ across the ‘messy middle’ (insert your favourite startup book title here) without existing products that can act as distribution and make all of this a feature, and to compete in one of the most capital-intensive industries in history without cashflows from existing businesses to lean on. Of course, companies that do have all of that need to be able to disrupt themselves, but we’re well past the point that people said Google couldn’t do AI.
Things are moving quickly right now.
The fourth problem is expressed in the quotes I used above. Mike Krieger and Kevin Weil made similar points last year: when you’re head of product at an AI lab, you don’t control your roadmap. You have very limited ability to set product strategy. You open your email in the morning and discover that the labs have worked something out, and your job is to turn that into a button. The strategy happens somewhere else. But where?
The current market.
This means that most people don’t see the differences between model personality and emphasis that you might see, and most people aren’t benefiting from ‘memory’ or the other features that the product teams at each company copy from each other in the hope of building stickiness (and memory is stickiness, not a network effect). Meanwhile, usage data from a larger (for now) user base itself might be an advantage, but how big an advantage, if 80% of users are only using this a couple of times a week at most?
Result?
In the meantime, when you have an undifferentiated product, early leads in adoption tend not to be durable, and competition tends to shift to brand and distribution. We can see this today in the rapid market share gains for Gemini and Meta AI: the products look much the same to the typical user (though people in tech wrote off Llama 4 as a fiasco, Meta’s numbers seem to be good), and Google and Meta have distribution to leverage. Conversely, Anthropic’s Claude models are regularly at the top of the benchmarks but it has no consumer strategy or product (Claude Cowork asks you to install Git!) and close to zero consumer awareness.
There is much more to his analysis. Definitely worth a read—and some thinking.
Click on the Follow button at the bottom of the page to subscribe to a weekly email update of posts. Click on the mail icon to subscribe to additional email thoughts.
